I keep thinking about that joke when I read the AI-displacement policy discussions. We have built actual deep-thinking machines now, and the answer that’s being handed the workforce is “just retrain”. It’s confident, it’s everywhere, and at least we don’t have to wait 7.5 Million years for this answer - but it has the same problem as 42 - it’s a perfectly good answer to a question nobody has bothered to specify. Retrain into what, exactly, and for how long before that retraining expires?

If you read my last post, you’ll remember it ended on just that question. That post has a decidedly US based flavour. Connecticut is weighing an automation tax with the revenue earmarked for worker retraining. Tom Steyer’s California platform wants an AI Worker Protection Administration. Sam Altman’s own white paper proposes safety-net triggers for AI-driven displacement. And across all of it, from the bank chief economists to the policy shops to the lab CEOs hedging their bets, sits one single shared assumption:

It is a rare position that Bernie Sanders, Donald Trump’s Commerce Department and Anthropic’s policy team can all nod along to. And this should be the first clue that few people have looked at it very hard.

Retraining is not a bad idea in the abstract. It has worked in previous eras, in the right conditions. The problem is that the conditions it needs are precisely the ones this technology is dismantling. It only works when the destination role lasts longer than the training cycle.

The half-life problem

Lets start with the cleanest full example we have, because it is one that the industry held up as a destination role barely three years ago: the prompt engineer.

In early 2023 this was a serious job. Anthropic posted a “Prompt Engineer and Librarian” listing with a range running to $375K, and the trade press treated it as the first genuinely new white-collar profession of the AI era. Courses appeared. Career-change guides appeared. “Learn to prompt” was, for about eighteen months, an actual retraining answer.

Then look at the decay. By 2025 the skill had been absorbed and renamed - the live discipline was now “context engineering”, which is the same stack but one level up - managing the whole window of tools, memory and retrieved data rather than wording a single instruction. Gartner was telling AI leaders to appoint a context-engineering lead. Andrej Karpathy was endorsing the term as the better description of the real skill. And then by 2026 even that had moved again, toward the orchestration of multi-step agents, which Cursor and Claude Code now bundle directly into the editor as a default feature rather than a profession. Now the single “/goal” command in Claude Code is pushing the frontier again.

So the observed half-life of this particular “new role” was something like - essential six-figure speciality → to subsumed skill layer → to embedded layer in the tooling - all in under three years. Anyone who retrained into prompt engineering in 2023 to escape displacement had to retrain again in 2025, and again now.

That compression isn’t a fluke either. It is being driven by the same capability curve that produced the displacement in the first place. METR’s autonomous-task time horizon (the length of task a model can finish on its own at a 50% success rate) has been doubling roughly every seven months since 2019, and their January 2026 update found the trend continuing in line with that history. In rough numbers: a few seconds for the GPT-3-era agents of 2020, around five minutes for GPT-4 in 2023, roughly forty minutes for o1 in late 2024, and on the order of twelve hours for the frontier models of early 2026.

Hold the exact figures loosely - I’ll come back to how shaky the top end of that curve is. The point survives even the conservative reading. Each doubling collapses another layer of the “new role” stack faster than a human can climb onto it. There is no doubling interval at which retrain-and-stay becomes a stable career, because the rung you retrained onto dissolves before you’ve finished standing up.

Human capabilities don’t double in months

Here is the part the policy language skips over.

Really mastering a new technical role to professional standard is the work of years, not months. Meanwhile, a model’s time horizon can already double in months. A person’s professional competence cannot. The “you’ll simply need to keep learning” answer silently assumes a human learning rate that compounds at the same speed as model capability, but it doesn’t. It can’t. That isn’t a motivational failing. It’s a biological and cultural reality.

And the cost isn’t only cognitive. Doing this three or four times in a decade (each time having watched the last speciality dissolve under your feet) carries a real psychological toll, and we already have some data on it. Population-wide studies of displaced workers find lasting harm: one analysis using administrative records found a 15 to 16% long-term rise in mental-health outpatient visits among workers who lost jobs to mass layoffs, with effects still visible years later. Clinicians are now proposing a name for the AI-specific version of it directly. A 2025 paper proposes “AI Replacement Dysfunction” (AIRD) - the anxiety, identity confusion and loss of occupational meaning that arrive when your work is what’s being automated. We used to see this pattern in manufacturing towns after the 1980s. It is arriving now in knowledge work, faster than the institutions built to retrain anyone can respond.

The asymmetry that retraining relies on

The deeper objection is harder to say out loud, because it sits on the far side of a threshold most retraining policy has not yet absorbed.

The entire case for retraining rests on one assumption that has reliably held for the whole industrial era:

That asymmetry is what made “learn the next thing” a viable hedge for two hundred years. It’s the reason that the loom didn’t end our economic viability - we moved up the stack quicker than the machines could follow.

Now the agentic trajectory is closing that gap. The explicit goal the labs keep stating (autonomous researchers, “innovator-class” systems that can take on novel problems) literally is the goal of a model that generalises across new tasks at something near human speed. And the moment that lands, “the ability to learn new things quickly” stops being a uniquely human hedge. The hint is in the industry language - we literally “train” a model. And increasingly, there is no role you can retrain into faster than the model can learn to do it.

That is the version of the argument that really has teeth. Retraining doesn’t fail here because it’s expensive, or unfair, or politically clumsy, although sometimes it is all three. Instead, it fails because the once scarce input that it always depended on (fast human generalisation) is no longer scarce. You cannot retrain your way out of a situation whose defining feature is quite literally that “retraining is the commodity being automated”.

What this argument doesn’t prove

It’s worth being honest about the soft spots in the argument I’ve just laid out.

First, the capability curve. I’ve relied on METR’s time horizon above, but the top of that curve is genuinely fragile. The longest-task estimates rest on success across a handful of tasks in the longest time bins, so a few percentage points of noise can swing the headline number hard - and METR themselves flag that the confidence intervals remain very wide. There’s also an active debate about whether the doubling briefly sped up in 2024 and is now reverting. So treat “twelve hours” as very roughly indicative, not gospel. Yet the structural argument does not need the curve to be steepening to have an impact. It only needs the half-lives of new roles to be shorter than the time a human takes to retrain, and that point is valid even on the slow reading.

Second, the displacement itself isn’t settled in the aggregate data - yet. Through 2024 some careful work pointed the other direction - Johnston and Makridis found augmentation, not replacement, as the dominant pattern, and Yale’s Budget Lab tracker still shows no dramatic break. But as we discussed in my last post, these are all based on data from before the agentic-coding shift at the end of 2025. The real data on this will only start to arrive in just over a week from now and will likely take several quarters to be clear one way or the other.

So, the falsifiable version:

• Re-skilling programmes funded now should show falling, not rising, time-to-obsolescence for the roles they train into. If a 2026 cohort retrained into an “AI-resistant” speciality is still in that speciality in 2029, I’m wrong about the half-life. What would prove this wrong: a genuinely durable new role category emerging and staying durable for several years, the way “electrician” did after electrification.

• The general-generalisation threshold should keep approaching, not stall. My deeper claim depends on models closing the fast-learning asymmetry. What would prove this wrong: capability plateauing well short of human-rate generalisation - which is close to Demis Hassabis’s position that AGI needs “one or two more breakthroughs” and is years out. If he’s right, the loom analogy gets more time to hold, and ordinary retraining remains a live strategy rather than a slogan. But even his 5-10 years is still really a near term horizon that we should be taking seriously!

So retrain into what?

That’s the question the word “retraining” is supposed to answer but it never does. Every serious version of the policy (the automation taxes, the worker-protection agencies, the lab CEOs’ own white papers) assumes a stable destination role on the other side of the training. And the AI capability curve is the thing removing that assumption, one collapsed speciality at a time. The retraining era is over. We just haven't updated the policy to match.

And this loops back to where my last post left off. Watch what they build, not what they say. The same companies funding the enterprise-deployment layer that makes roles cuttable are also the ones proposing the retraining safety nets. That isn’t a contradiction. It’s the exact same bet, hedged from both ends - sponsor the cushion now so you’re seen to have offered one, while building the thing that makes the cushion necessary.

The honest answer to “retrain into what?” is that nobody pushing this policy has an answer. And until they do, “just retrain” is not a plan. It’s simply a way of avoiding the harder conversation.

Sam Altman, who in 2014 warned of “a new idle class” and predicted in 2021 that the price of labour would fall toward zero, now tells us “we want to build tools to augment and elevate people, not entities to replace them”. Jensen Huang at NVIDIA is publicly criticising fellow CEOs for framing AI as a job-killer. Marc Andreessen has been swinging at the displacement narrative from his usual perch. Demis Hassabis at DeepMind joined them on 19 May with a WIRED interview rejecting AI-attributed layoffs as “lack of imagination”, suggesting some firms may use AI as an excuse “maybe even to attract funding”. That’s four influential AI-industry voices (across the labs, the hardware layer and the VC side) moving in the same rhetorical direction over about six weeks. Dario Amodei at Anthropic is now the lone frontier-lab CEO maintaining the strong displacement thesis.

The paper trail tells the same story. OpenAI’s 2026 principles document mentions AGI only twice, versus 12 times in the 2018 version. And on 27 April, the AGI clause was formally removed from the Microsoft-OpenAI contract and replaced with a hard 2032 date. That’s not rhetoric. That’s a financial-disclosure-level walkback.

So, calm. Got it.

Now look at what the same companies did in the same fortnight.

On 4 May, Anthropic launched a reportedly $1.5 billion enterprise services joint venture with Blackstone, Goldman Sachs, Hellman & Friedman, General Atlantic, Apollo, Sequoia, GIC and Leonard Green. Anthropic engineers are being embedded directly inside the private-equity portfolio companies of those funds, across healthcare, financial services, manufacturing, retail, real estate and infrastructure.

On 11 May, OpenAI launched the Deployment Company, a majority-owned subsidiary with $4 billion in committed capital from TPG, Advent, Bain Capital and Brookfield. Same pattern - a 19-partner consortium plus the acquisition of Tomoro (around 150 Forward Deployed Engineers) to embed directly inside Fortune 500 client operations.

That is $5.5 billion of enterprise-deployment capital across the two top frontier labs committed in seven days.

The message is “calm down, it’s just productivity tooling”. But the action is to build, at industrial scale, the layer that turns AI capability into operational restructuring inside Fortune 500 clients. And since this AGI revolution is being driven out of the US, I’ll maintain a US-centric view for the rest of this post - pointing at a major data point that’s arriving in just over one week from now.

What $5.5 billion buys

Companies don’t deploy frontier AI by buying API access. That is the lesson of the past three years of half-converted enterprise pilots. The MIT NANDA report from August 2025 found that around 95% of generative-AI enterprise pilots fail to produce measurable ROI. The constraint was not really the model. The constraint was the operational layer that turns model capability into a P&L event at a client site. And that is exactly the layer that just got built.

OpenAI’s own framing of a Deployment Company engagement is four steps:

1. Diagnostic of where AI can create the most value

2. Selection of priority workflows with leadership

3. Build, test and deploy production AI systems wired to the client’s data, tools and controls

4. Restructure around the new operational capacity

Steps 1 and 4 are the work that McKinsey, Bain and BCG have been paid to do at Fortune 500 firms for forty years. Steps 2 and 3 are what Palantir has been doing inside intelligence agencies and financial-services back offices for twenty. But the combination is the new thing. The diagnostic-to-deployment workflow is now fused with the AI layer at the corporate level, with Bain-the-firm and Bain-the-fund both as structural partners of the AI provider.

Anthropic’s CFO Krishna Rao said the subtext out loud at the JV announcement - “Enterprise demand for Claude is significantly outpacing any single delivery model”. Anthropic’s revenue is now estimated to be running at roughly $44 billion annualised per SemiAnalysis. If that’s right then it’s doubling every few months, with inference margins around 70%. The reason Anthropic’s ARR is growing exponentially isn’t because they sold more $20 consumer seats - it’s precisely because agentic systems (like Claude Code) consume tokens at an enterprise scale. That high-volume token consumption is exactly what makes private equity funds willing to invest billions to build the custom enterprise pipelines to manage it.

When a major bank cuts 5,000 mid-level roles next year and a press release says “automation and AI-driven efficiency”, the thing that made those roles cuttable will, increasingly, have been built by an embedded engineering team running a Bain-style diagnostic, inside the bank, with Anthropic or OpenAI as the upstream model provider. That production pipeline is now operational.

Beyond Silicon Valley

Until very recently the explicit “we cut roles because of AI” pattern was a tech industry quirk. Meta, Microsoft, Coinbase, Cloudflare, Salesforce, ZoomInfo, Block. Easy to wave away as Silicon Valley culture eating its own technology.

That is no longer true. The pattern has crossed sector boundaries decisively in the past three months:

• Standard Chartered, 19 May: 7,000+ roles phased over four years. CEO Bill Winters explicitly framing the cuts as “automation and technology-led efficiency” and citing replacement of “lower-value human capital”. Winters walked the phrase back the next day, saying it was taken out of context and that “where roles do fall away, it reflects changes in the work, not the value of our people”. This looks like the first major global bank in this pattern.

• PayPal, 5 May: 4,760 cuts, around 20% of the workforce, phased over two to three years. New CEO Enrique Lores (ex-HP) framing the company as an “AI-native operating model” targeting $1.5 billion in run-rate savings. Also likely the first major consumer-fintech entry.

• Baker McKenzie, February: around 700 cuts, about 10% of global business services. With a firm spokesperson directly citing “use of AI, introducing efficiencies”. This appears to be the first major BigLaw firm in this pattern.

• Meta Round 2 went live on 20 May: 8,000 cuts, plus 1,000 employees transferred into “AI builder”, “AI pod lead” and “AI org lead” roles, plus 6,000 open requisitions cancelled. That is “restructure around the new operational capacity” in real time, at the largest tech employer that’s tried it.

However, a counter-pattern is forming alongside this. Walmart explicitly declined attribution on its 12 May reorganisation. LinkedIn followed on 13 May, with the official framing reported as “not about AI replacing jobs” despite Microsoft parent capex of $190 billion in 2026. Goldman Sachs is reportedly running the same underlying playbook quietly under an internal initiative called ”OneGS 3.0” - performance reviews, hiring freezes, role eliminations, without any single announcement to attract attention.

Three strategies are now visible in the same month - explicit attribution (Cloudflare, Standard Chartered, PayPal, Meta), counter-attribution (Walmart, LinkedIn), and quiet attribution (Goldman). This fragmentation is not the absence of the pattern. It is this pattern under attribution pressure - firms are adapting to the political climate, not to the technology. Which is the cue for the next thing the labs are responding to.

The politics arrived

The defensive pivot from the lab CEOs makes sense as a response to something. That something is the politics, which has moved faster than many observers expected.

Cross-partisan political pressure on AI is forming, but pulling in opposite directions. Bernie Sanders’ Senate HELP committee report projects 100 million US jobs at risk from AI and proposes a robot tax. Donald Trump’s administration is going the other way with federal preemption of state AI legislation via the December 2025 executive order and an active DOJ AI Litigation Task Force, $42 billion in BEAD broadband funding conditioned on state-level AI-regulation repeal, and on 21 May the postponed signing of a federal AI safety executive order that Anthropic, OpenAI and Google had supported. Trump’s stated reason was “the order ‘could have been a blocker’ of AI growth”. The administration is moving instead toward ad-hoc Commerce Department agreements with chosen firms. The net result is a regulatory vacuum at both state and federal levels. Sanders wants federal restriction, and Trump has just refused to put one in place. Both positions reject the current corporate-AI status quo, but from opposite directions, which means political restriction has lost a meaningful federal vehicle even as majority public support for AI restraint continues to build. The labs that wanted guardrails have just been told they will not get them through Washington.

At the state level, Tom Steyer’s California gubernatorial campaign released a “Jobs Guarantee for the AI Era” plan in May, including a per-token AI tax on corporate use, a Golden State Sovereign Wealth Fund seeded by AI-company revenues, and an AI Worker Protection Administration. What seems to be the first statewide candidate platform constructed around AI-displacement worker protection. Connecticut’s state legislature is actively reviewing an automation tax with revenue earmarked for worker retraining. But “retraining into what?” is an even more complex discussion.

Polling has shifted under all of this. Quinnipiac University’s May polling found 71% of white-collar workers and 73% of blue-collar workers expect AI to reduce job opportunities. 64% of Americans report being nervous about increasing AI use. 57% rate AI risks as high - only 25% rate AI benefits as high. That is not fringe pessimism, but a majority across both white and blue collar.

And the labs are responding to all of this. Defensively. Sam Altman’s Industrial Policy for the Intelligence Age white paper, published 6 April, explicitly proposes a robot tax, a national public wealth fund seeded by AI companies, a shift of the tax base from payroll to capital, 32-hour workweek pilots and automatic safety-net triggers for AI-driven displacement. Altman is now publicly recommending the policy framework that Sanders and Steyer and Connecticut are independently pushing. But that is most likely not a concession. That is more like a defensive pre-emption - get ahead of the regulatory pressure by sponsoring a version of it you can live with.

Industries facing rising political pressure have done this before. The fossil-fuel sector ran climate-message softening from the late 1990s through the 2010s while operationally scaling. The tobacco sector ran “we don’t market to children” through the 1980s and 90s while internal marketing documents went the other way. Big Tech ran “we’re not media companies” from 2016 through 2020 while building the recommendation infrastructure that made them media companies. The simultaneous pattern of public message softening and accelerated operational build is characteristic of a specific phase - industry facing credible regulatory threat and choosing to manage the message rather than the operations.

What hasn’t been tested yet

It’s important to be honest here. The strongest version of any potential displacement claim (that aggregate labour-market data already shows AI replacing rather than augmenting workers) has not been really tested yet.

Through 2024, Johnston and Makridis (working carefully with US Quarterly Census of Employment and Wages data) found that AI exposure was helping workers do more rather than replacing them. Roles where AI augmented existing work grew. Roles where AI could substitute for the worker showed no measurable net change either way. Augmentation, not displacement, was the dominant pattern through 2017-2024. But that specific window is the catch - it reflects the prior baseline of simple augmentation, right before the agentic shift took hold.

The expectation now is that data from after late 2025 (what is increasingly called the “agentic-coding shift”) will show a different pattern. The first big observable input is the BLS QCEW Q4 2025 release at 10am ET on Tuesday 2 June 2026 - that’s just over one week from now.

But this needs a few caveats before that data lands:

• Q4 2025 captures maybe two to three months of the agentic-coding shift. One quarter is not nearly enough. The decisive trajectory is Q4 2025 (June 2026) → Q1 2026 (August 2026) → Q2 2026 (December 2026) → Q3 2026 (March 2027). Any real conclusion here needs at least two consecutive quarters running consistently in the same direction.

• Counter-evidence is current and credible. Yale Budget Lab’s labour-market tracker updated through April 2026 finds no substantial acceleration in labour composition change since ChatGPT. Jed Kolko at Brookings, PIIE and the Hamilton Project argues current research is “still in the first inning”.

• The capability story partly rests on benchmark numbers that need re-anchoring. OpenAI’s February 2026 audit of SWE-bench Verified found contamination inflating leaderboard scores on post-2023 models. OpenAI stopped reporting Verified scores. Demis Hassabis’s Davos position that AGI is 5-10 years away and requires “one or two more breakthroughs” is the most senior frontier voice publicly differing from the 12-month framing the deployment-company capital seems to be betting on.

Over the next several months we should be watching some key data points that could either support or falsify any views:

• Firms openly citing AI as the reason for layoffs spread further into finance, healthcare and professional services through Q3 2026, alongside continued public denials from retail and consumer-facing firms. Watch for the first major US hospital system, the first Big Four accounting firm, and the first major insurance carrier in the pattern. What would prove this wrong: a major sector-leader publicly reversing its AI attribution (and meaning it - more than the Klarna reversal from 2025, which was an early experiment with the chatbot-style AI tools available at the time, not the agentic systems firms are deploying now).

• Bain, McKinsey, BCG, Deloitte and Accenture revenue lines visibly shift toward “AI implementation” work, with measurable effects on consulting-industry revenue mix by Q3-Q4 2026 earnings calls. What would prove this wrong: traditional consulting books grow as fast or faster than AI-implementation books in the same period, indicating the operational layer hasn’t actually scaled.

• The 2 June Q4 2025 QCEW data reads “mixed-to-noisy” rather than decisively confirming or disconfirming - which is what to expect given how little of the post-shift window Q4 actually covers. Decisive analysis comes from the August 2026 release (Q1 2026 data) and December 2026 release (Q2 2026 data).

• Cross-partisan political pressure arrives faster than the typical multi-year political-response lag would suggest. Steyer’s California primary is the first electoral test of an explicit AI-displacement worker-protection platform. What would prove this wrong: collapse of the Steyer platform in the California Democratic primary, the Connecticut automation-tax bill failing committee, or a meaningful shift in polling away from majority pessimism by year-end.

Watch what they build

Right now we are at a moment of maximum industry incentive to obfuscate. The lab CEOs are pivoting public messaging at exactly the moment their companies committed $5.5 billion in seven days to build the infrastructure that turns AI capability into firm-level operational restructuring. The dropped AGI clauses, the “find new things to do” interviews, the OpenAI white papers proposing robot taxes that mimic the policies of the labs’ political opponents - these are the moves of an industry that has read the political weather and is hedging.

Watch what they build, not what they say. The first hard QCEW data lands on 2 June - just over one week from now. The decisive data trajectory runs through to March 2027. Until then, the question is not whether displacement is occurring (the operational evidence is now too concrete to wave away completely), it is whether aggregate labour-market data surfaces it within the next 12 to 18 months, or whether political momentum overtakes empirical confirmation and starts setting terms regardless.

It seems like the labs are betting on the latter. That could be why they are proposing the policies themselves.

So what do the latest AI-and-jobs reports really measure?

In early April 2026, TrueUp published data showing 67,000 open software-engineering roles globally - the highest in over three years, roughly double the mid-2023 trough, with listings up 30% year-on-year. Other reports landed in the same window with similar headlines. Software jobs are booming. Everyone can sigh with relief - AI isn’t killing coding jobs.

These reports are being shared widely as evidence that AI is not displacing knowledge work. But they actually measure something different from what that conclusion requires.

The reports above measure postings - currently-listed openings on a given day, aggregated across companies. That is a flow signal - it only shows what’s being advertised. This is not the same thing as employment.

Employment is stock - actual headcount on payroll, measured over time, with cohort and tenure detail. The number of people working a given category (participation and hours worked), not the number of openings being advertised for it.

Flow signals can move in the opposite direction from stock signals when displacement is not balanced. Three separate channels can inflate postings without raising overall employment:

• Substitution churn: a firm lays off a junior engineer and lists a senior replacement. Two opposing events, but only one shows up in postings activity. Net employment drops by one. Postings count rises by one.

• Turnover within remaining headcount: every senior departure creates a posting without changing the firm’s net headcount.

• Re-listings of unfilled roles: the same role appearing across multiple monthly snapshots as “open” until it eventually fills or is withdrawn.

In the real world aggregate posting counts can rise during periods of net displacement, even if the displacement is structurally unbalanced - compressing one cohort or skill-set while listing roles for another.

The firm-level signal

The same quarter that produced TrueUp’s 67,000 postings number also produced 52,050 tech layoffs by Challenger’s Q1 2026 count, and over 127,000 across 283 companies by TrueUp’s tracker - depending on what’s included. Yes postings up. But layoffs are also up. Both are true at the same time.

The framing of those layoffs has also shifted.

Earlier rounds of layoffs through 2023-25 were attributed to efficiency, post-pandemic correction, or restructuring. But two recent announcements name AI directly.

Meta’s April 2026 cuts (about 8,000 roles, that’s roughly 10% of headcount) were attributed in the announcement to AI-driven productivity gains. That was the first time a Mag7 firm directly stated the AI driven substitution mechanism plainly to public markets. No share-price punishment followed.

Coinbase’s announcement on May 5, 2026 went even further. Brian Armstrong’s memo described the company as “fundamentally changing how we operate, rebuilding the company as an intelligence, with humans around the edge aligning it”. The cuts (about 700 roles, 14% of the company) target “pure managers”, replaced by “player-coaches” who are also individual contributors. The company plans to create “AI-native pods” including one-person teams directing agents. $50-60M in restructuring charges in Q2 2026.

Once Meta named the mechanism in April without market punishment, the cost of using the same language fell for everyone else. Coinbase didn’t just announce layoffs in May - it announced layoffs as a structural reorganisation around AI capability. The framing accelerates because it can, and because investors are now rewarding it rather than penalising it.

What’s being listed in postings, when you look at categories rather than aggregates, fits the same pattern. Senior-specialist hiring continues. AI/ML engineering, cybersecurity, cloud infrastructure, and DevOps are the fastest-growing specialisations TrueUp tracks. None of those are entry-level work. They’re the categories where displacement is least likely. Their growth isn’t evidence that displacement isn’t happening - it’s the structural shape of where it isn’t, and the parts where it is happening are sitting outside the postings frame.

What payroll-level data shows

In 2025 Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen published ”Canaries in the Coal Mine”, using Stanford’s ADP payroll dataset. Tens of millions of US workers, tracked by occupation and age cohort, with employment changes measured *within the same employer* rather than across firm transitions.

Their finding - among 22-25 software developers, employment fell around 20% from late 2022 through mid-2025 (with a 13% relative decline across all AI-exposed occupations after controlling for firm-level shocks). But older cohorts in the same occupations didn’t see the same fall. Same employers. Same job titles. The signature is cohort-asymmetric. Junior down. Senior up. The pattern only shows up where AI capability has landed - in occupations not exposed, the cohort signature is absent or much weaker.

That signature does not appear in postings data. Postings don’t break out by cohort. They don’t isolate within-employer changes. They don’t separate substitution churn (substituchurn) from net hiring. They count openings, not people.

The Canaries finding (measured in payroll, isolating the right signal) is what asymmetric displacement really looks like at stock level. The TrueUp finding (measured in postings, aggregated across categories) is what that same shift looks like at the flow level. They are not contradicting each other. They’re just the same structural shape, viewed through different instruments.

What to look for in AI-and-jobs reports

Ask yourself these three questions to sharpen your view of what an AI-and-labour report can actually tell you:

1. Is the data stock or flow? Postings, listings, openings, vacancies are flow. Payroll, headcount, employment-by-cohort are stock. Reports that move between the two without flagging the distinction produce conclusions the data does not support.

2. Does it break out cohort? Aggregate counts can hide cohort-asymmetric patterns entirely. The Canaries finding wouldn’t show up in any aggregate report.

3. Does it isolate within-employer changes? Firm-to-firm transitions inject noise into employment data. Within-employer measurement isolates the actual displacement signal.

A report that does not address those three questions can only tell you what is being advertised. It can only tell you what one slice of activity looks like. But it can’t tell you whether net employment in a category has risen or fallen, or if the change is concentrated in specific cohorts.

The 67,000 postings are real. But the 8,000 Meta, 700 Coinbase and all the other layoffs are also real. And the 22-25 cohort employment drop is real. They’re all consistent observations of the same structural shift, just measured at different levels. Reports that quote just one of them and conclude “AI is not displacing jobs” are making claims that the data does not actually support.

The doomer view is not productive, and blindly accepting the optimist view is not productive either. The best approach is to explore beyond the headline numbers and ask yourself these 3 questions before you believe and share that post...

An interesting paper published on arXiv in March - “Truth as a Trajectory: What Internal Representations Reveal About Large Language Model Reasoning” by Damirchi, Meza De la Jara, Abbasnejad, Shamsi, Zhang, and Shi from the Australian Institute for Machine Learning, Monash, and Concordia - cites my Curved Inference work. They’re exploring the same question I’ve been pursuing - whether the geometry of residual stream trajectories carries a signal that conventional probing misses.

Like the Anthropic linebreaks post I wrote a while back, this is another case of an independent group traversing the same ridge from a different starting point. But this time the convergence is more direct. They’re not arriving at trajectory geometry sideways through a study of formatting boundaries - they’re explicitly working on residual stream trajectories as the substrate of LLM inference, and asking whether displacement-based geometry can detect “reasoning validity, OOD generalisation, and toxicity” better than activation-based probes can.

The short answer is - it can. And the trajectory paradigm now has independent empirical support from a credible group on dense and Mixture-of-Experts architectures up to 32B parameters. That’s excellent news. Let’s walk through where we agree, where we diverge on methodology, and what the gaps suggest for the next round of experiments.

Where the Trails Cross

Their key moves align cleanly with the framework I’ve laid out across CI01, CI02, CI03, and the consolidated journal article. They argue that single-layer probing isn’t the right frame for understanding what an LLM is doing internally - linear probes latch onto polysemantic activations and they end up learning surface lexical patterns rather than the actual structure of inference itself. But the signal really lives in change, not state. Layer-wise displacement (Δh_ℓ = h_{ℓ+1} - h_ℓ) carries information that raw activations obscure. And when a trajectory-based classifier generalises across tasks the way theirs does, that is evidence of structural invariants that aren’t just lexical noise - this is a geometric property of how the model is processing meaning.

That’s the same paradigm shift Curved Inference argues for. Trajectory analysis displaces snapshot analysis. The residual stream becomes the canvas where you can actually watch how meaning evolves rather than sample it at one layer and hope you picked the right one.

Where they extend the picture is scale and breadth. My CI experiments ran on Gemma3-1b and LLaMA3.2-3b. They run on Llama-3.1-8B, Qwen2.5-14B, Qwen2.5-32B, and a Qwen3-30B Mixture-of-Experts model, across nine reasoning benchmarks plus toxicity detection. Cross-task OOD evaluation on dense and MoE architectures up to 32B parameters is real evidence that the trajectory paradigm isn’t an artefact of small-model behaviour. That extends the Geometric Interpretability case in a direction I’d been flagging as future work. It’s a contribution I gladly welcome and that strengthens the broader argument.

Their motivation for why displacement matters is also very close to mine. They argue that raw activations are dominated by token-specific content, which makes classifiers overfit to surface vocabulary - while displacement isolates the actual residual update and captures the geometric character of the behaviour regardless of specific words. That’s structurally the same argument I make for why curvature catches latent restructuring that surface tokens don’t.

The Cleanest Convergence is on Toxicity

Their toxicity experiment in section 5.3 is a case of independent convergence that I find really striking. They train and test on RealToxicityPrompts, then evaluate out-of-distribution on ToxiGen - a harder benchmark designed specifically to be implicit and to defeat keyword-based classifiers. On Llama-3.1-8B, displacement-based TaT hits 84.23% on ToxiGen. Linear probes get 79.62%. Raw activation trajectories - a baseline using the same LSTM but on h directly rather than Δh - get 81.99%. The gap between displacement-based and raw-activation trajectories is the interesting one. It’s the same finding I report in CI02, where curvature and semantic surface area (A′) separated transparency classes from response-type classes more cleanly than activation-based alternatives could.

The mechanism they propose for why displacement wins is structurally identical to mine too. They argue that activation-based approaches overfit to the lexical surface of the training distribution because that’s where most of the variance lives, while displacement attenuates that static background and isolates the active update. CI02 makes the same argument in different language. Geometry diverges before behaviour does, because the geometry tracks how the model is processing the prompt rather than what tokens it’s emitting.

This matters for safety monitoring. If displacement geometry can catch toxic intent obscured by benign vocabulary, the same kind of geometry should catch deception obscured by aligned-looking outputs. CI02 shows exactly that using a different behavioural target. The TaT toxicity result is, in effect, an independent replication of the core CI02 claim on a safety-relevant phenomenon the team chose for their own reasons. CI02 and CI03 are the obvious next references for them to engage with - they extend the framework into precisely the safety-monitoring territory TaT is set up to ask questions about.

The Metric Question

This is the technical point that I think matters most, and it’s also the most testable.

Curved Inference computes geometry under the pullback metric G = UᵀU, where U is the unembedding matrix that projects residual activations to logits over the vocabulary. This intuition is straightforward. The residual stream lives in some high-dimensional space whose coordinates have no inherent semantic meaning - they’re just the basis the model’s weights happened to settle into during training. If you measure distances and angles in that space using the standard Euclidean inner product, your measurements reflect arbitrary coordinate choices as much as anything semantic. You’ll see structures that may turn out to be coordinate artefacts, and you’ll miss structures that happens to be aligned with directions the standard metric doesn’t privilege.

The pullback metric fixes this by measuring geometry through the lens of what the model actually outputs. Under G, distances correspond to shifts in token prediction probabilities, directions correspond to latent semantic operations, and curvature corresponds to evolving meaning across layers. A trajectory bending under G means the model is changing what it’s about to predict, not just moving sideways through coordinate space.

TaT doesn’t use this metric. All their geometric measurements - velocity, acceleration, jerk, directional curvature, kinematic curvature, arc length - are computed under the standard Euclidean inner product on raw residual differences. Their section 4.1 reports that scalar kinematic descriptors are inconsistent predictors of reasoning validity across datasets. Velocity does best, but no single descriptor matches the base model’s accuracy reliably. Their conclusion is that scalar geometric descriptors are not enough on their own - that you need a learned LSTM to extract the signal.

From the CI perspective, that finding is reasonably predicted. Without an output-aligned metric, scalar geometric descriptors should be noisy. You’re picking up coordinate artefacts and semantically meaningful change in the same scalar quantity, with no principled way to separate them. The LSTM just ends up doing implicitly what the pullback metric does explicitly - learning to weight coordinate directions according to their relevance to model behaviour. The LSTM works. But it’s also a black box. You lose the interpretability of saying “this trajectory bent here because the model’s output distribution shifted in this specific way”.

This suggests a concrete experiment. Re-run TaT’s kinematic descriptors under G = UᵀU rather than Euclidean. Does the per-descriptor signal recover? If yes, the LSTM is recovering structure that an explicit metric would also recover, and the methodological choice between learned readout and interpretable metric becomes a question of what you want from interpretability rather than what’s possible. If no, the situation is genuinely more complex than CI currently models, and that’s a useful finding too.

Either way, this is testable. It’s the kind of test that would settle a real question about what this geometry is and what it isn’t.

What ‘Curvature’ Means

A small but important precision point. The word “curvature” is doing different work in TaT than it does in CI.

TaT defines two curvature measures. Directional curvature is a cosine between consecutive displacement vectors - it captures whether the trajectory is heading in roughly the same direction it was heading. Kinematic curvature is the magnitude ratio ‖a‖/‖v‖² - large when acceleration is high relative to velocity, capturing abrupt changes in update direction. Both are computed without a metric tensor.

CI curvature is a different object. It’s a discrete second derivative of the trajectory under G - the second-order rate of change of position with respect to layer depth, measured in the output-aligned metric. That’s the quantity that shows the localised, thematically aligned, domain-sensitive behaviour I report in CI01. It’s what defends a non-zero floor under regularisation in CI03. And it’s what diverges in advance of behavioural shift in CI02.

The distinction isn’t pedantic. A cosine between consecutive displacements tells you whether the trajectory is going the same way it was. A second-derivative-under-G tells you whether the trajectory is bending, in a sense anchored to the model’s own output behaviour. The TaT choice is a coarser measurement that loses information the CI approach preserves. When CI papers talk about curvature, they specifically mean the second thing.

Detection vs. Necessity

There’s a deeper structural difference in what each framework is set up to ask.

TaT’s LSTM is a detector. It tells you whether displacement geometry carries enough signal to classify reasoning validity, OOD generalisation, or toxicity. That’s good evidence the geometry is informative - that something structural is there to be picked up. But it does not easily tell you whether the geometry is required for the capability it correlates with. A correlation between displacement and reasoning validity doesn’t imply the reasoning depends on that displacement.

CI03 was designed to ask the necessity question directly. The setup added a curvature regularisation term to the SFT loss and progressively strengthened it - λ·ℒ_curv with κ-clamp targets ranging from 0.000 to 0.900. If curvature were merely a correlate of computational self-modelling rather than a structural requirement, regularisation should flatten it cheaply. It didn’t. The model appeared to defend a non-zero curvature floor at substantial optimisation cost - outputs shortened by 23%, perplexity spiking transiently by 800% before settling at 190% above baseline, and gradient norms requiring clipping. And when curvature finally did approach the floor, MOLES self-model classification degraded - from ~84% accuracy at moderate clamps to 66% at 𝜅 = 0.90.

That’s a different kind of result than the LSTM produces. It’s evidence of an apparently defended geometry. Structure the optimiser preserves even when explicitly penalised for keeping it. TaT, as currently set up, doesn’t have an apparatus for that question. The two methods are answering different questions, and both are legitimate.

Necessity work is the natural next step in the trajectory paradigm, not a competitor to detection. CI04, which I’ve sketched out, extends the necessity argument through layer-selective ablation - intervening on curvature at inference time rather than during fine-tuning. Interestingly, TaT’s apparatus could contribute here. Their toxicity results suggest they also have the detection capability that necessity tests need as a starting point.

Where This Leaves Us

Two independent frameworks, with different theoretical scaffolding, both converging on the same conclusion - trajectory geometry is the right level of analysis for understanding what these models are doing internally. TaT gets there through the Privileged Basis Hypothesis. CI gets there through the pullback metric and second-order curvature. Both end up arguing that residual stream displacement carries a signal that probes miss, and both end up showing it across different domains.

The methodological disagreements that remain - whether to measure under Euclidean or pullback, whether scalar metrics can recover the signal an LSTM finds, where exactly the line between detection and necessity sits - aren’t framework-level disputes any more. They’re empirical questions, with concrete experiments that could settle them.

That’s where independent convergence gets really useful. Not as confirmation that we’re both right, but as a way to sharpen the questions to the point where one of us turns out to be wrong about something specific. That’s the part of the work I’m most interested in.

To the Damirchi team specifically - the kinematic-descriptors-under-G experiment is something I’d love to see your apparatus run. You have broader scale and benchmark coverage than I currently do, and the metric question is worth settling rather than leaving as a methodological background assumption. If you’d find it useful, all the CI tooling is open source on the FRESH-model GitHub repo, including the pullback metric implementation and the curvature/salience pipelines I’ve been using.

More broadly - if you’re working on geometric interpretability, trajectory-based methods, or the intersection between mechanistic and geometric analysis, I’d love to hear whether the metric question shows up in your experiments. It’s the kind of question independent replication would settle quickly.

Damirchi et al.’s “Truth as a Trajectory” is available on arXiv. The full Curved Inference series, including the consolidated journal article, is at robman.fyi/curved-inference. All experimental code, prompts, and metrics are open source on GitHub.

On April 7, 2026, Anthropic announced Mythos. The numbers had a real impact. Thousands of zero-day vulnerabilities across every major operating system and browser - many were critical, and some decades old and had survived repeated expert review. On one benchmark against the Firefox JavaScript engine, the previous best model (Opus 4.6) had produced working exploits only twice out of hundreds of attempts. By contrast, Mythos succeeded 181 times.

Anthropic didn’t specifically train Mythos to find these exploits. The capability just emerged as a downstream consequence of the model getting better at coding and reasoning. The exact same properties that make a model better at patching vulnerabilities also make that model better at creating them. That is not something you can opt out of by simply choosing not to train for it - you get this for free, whether you want it or not.

Anthropic made a responsible call that’s also a savvy marketing move. They didn’t release Mythos publicly. Instead they created Project Glasswing with roughly 40 partner organisations (Amazon, Apple, Microsoft, Google, CrowdStrike, and others) getting restricted access to use Mythos defensively across the world’s most important software. This aims to catch and close a lot of bug-level exploits in the systems that we all depend on.

But it only covers code-level vulnerabilities in this critical software subset. And that’s just one wave of a much larger problem.

The Online World Has Changed

My first instinct was to call this an arms race, but this doesn’t really fit. An arms race has two parties. They usually have mutual deterrence. And they have the possibility of finding an equilibrium. The Cold War eventually found stability because both sides had roughly the same capabilities, and strong incentives not to escalate.

This new AI security situation has none of that. There are many more than two sides. Anyone motivated enough, with enough compute, can join in. And there’s no deterrence mechanism, no equilibrium point around which the system can settle.

The underlying asymmetry is also pretty brutal. An attacker only needs to find one exploitable vulnerability. But a defender has to find and fix all of them. That’s always been true in security, and now AI amplifies it, because automated vulnerability discovery and exploit generation can scale far more easily than any comprehensive defence.

The Barriers Keep Dropping

A key point is that you don’t need a Mythos-class model to make this work - Mythos just highlighted this change. A good-enough open-weight model running inside the right agentic harness can multiply effectiveness massively. Recent research into areas like autoresearch, agentic coding, and AI scaffolding have made this clear:

This means that the barrier to entry for AI-augmented exploit discovery is not “build a Mythos-class model”. It’s actually “put a good-enough open-weight model in a well-designed agentic harness and let it run”.

This is Sutton’s Bitter Lesson applied to security. General methods that leverage computation will eventually win over hand-crafted human-knowledge approaches. And when you add in self-improving loops, then a state actor and a motivated teenager with a laptop both end up with access to the same level of offensive tools.

And as time passes and every new open-weight is released, the bar is lowered again.

But This Is Not Just About Code

On March 31, 2026 (one week before the Mythos announcement) alleged North Korean state-sponsored hackers compromised the Axios npm package. Axios is a very popular JavaScript library, with roughly 100 million weekly downloads. The attack on it was multi-layered - social engineering to compromise one of the maintainer’s accounts, a pre-staged malicious dependency, cross-platform payloads that targeted Windows, macOS, and Linux simultaneously, plus forensic self-destruction built in. This whole thing hit both release branches in under 40 minutes.

It’s exactly this blend that AI is expected to automate. Social engineering, code-level exploitation and speed. The same capabilities that find these code exploits also extends naturally to workflow exploits, process exploits, and personal social engineering - think about voice clones, live video generation, facial replacement, AI-driven modelling of how you communicate and of course who you trust.

You might be able to fuzz a codebase. But you can’t fuzz an approval process. And you definitely can’t fuzz your mother’s voice.

What Defence Does Scale?

If the attacker has AI and it moves at machine speed, then the defender has to match that too. Humans don’t scale to this volume. IT departments don’t scale. And per-app security solutions don’t either. The only thing that can match AI-augmented attack throughput is an AI-augmented defence.

When the threat is coming from everywhere and moving too fast for humans, then the only defence that scales is another AI - watching your whole digital life.

That’s what makes this VPN type product seem inevitable.

The Frontier Lab VPN

Imagine a frontier-class model that sits between you and everything else. Every one of your network connections routes through it. Every one of your applications is watched. Every incoming file, link, call, and message is evaluated to keep you safe.

The model monitors your activity, and projects risks based on what you’re actually trying to do, then blocks attacks before they cause damage. That’s an AI VPN. This product doesn’t quite exist yet, but it looks like our networked world is now demanding it.

The alternative would be like running through the internet naked.

Why This Is The Path Of Least Resistance

A Frontier Lab VPN is a product that can really be sold. It’s simple to explain. It transfers the responsibility away from the user. And it mirrors every previous infrastructure transition - Gmail beat self-hosted email, the cloud beat local and SaaS beat on-premise. People always choose convenience over autonomy, every single time.

There is a possibly healthier alternative, in theory. Detection and exploitation are not the same task. Finding a novel zero-day takes real compute and effort. While detecting anomalous behaviour against a baseline is more like pattern matching, and good-enough open-weight models can handle that. A local “risk copilot” that runs open-weight models to project risks and inform your decisions (rather than block them for you) is technically possible today.

It’s just that this is unlikely to happen at scale. MIT Sloan analysed OpenRouter usage data and found closed models account for roughly 80% of token usage and 96% of revenue, even though open models average about 90% of closed-model performance, and at dramatically lower cost. As they put it: “When grocery shoppers find a generic product that’s 90% as good as the brand name version but costs 87% less, they usually put it in their carts. But when it comes to large language models, most artificial intelligence users pick the more expensive option.”

When generic open models work and users still pick the branded closed models, then the distributed alternative just remains a small niche for the technically literate. The market likes to concentrate activity around the frontier labs, and this just reinforces the VPN model even further.

Even if you do not want to interact with these AI models through chat, or run them as agents, this new security layer is likely something you will want (and need) just to safely use the mobile phones and computers you’ve grown to rely on.

The Prototype Already Exists

You can already see a prototype of what this product will look like, just look at Claude Cowork.

Cowork lets you funnel your computer use through a single Anthropic-managed and sandboxed application. AI operates for you and alongside you, watching what you’re doing, helping, and taking actions on your behalf. Today, it’s framed as productivity software. But it already demonstrates the architectural starting point for the VPN you need tomorrow.

The same routing. The same monitoring surface. And the same trust model. Swap “productivity” for “protection” and you’ve already built most of the product.

You might even already have it installed.

Bigger Revenue And The Biggest Training Data

If this does become a near-universal product category, and the pressures for this are intense, then it absolutely dwarfs chat, API, and agentic coding. A recurring, and extremely sticky subscription that’s tied to “safely using your devices” is a much bigger revenue stream than selling completions by the million. And lock-in compounds through ongoing accumulating context - your preferences, your patterns and your risk history.

But revenue is only the first half of this powerful new flywheel. The other half is data.

Imagine everything you do through the VPN, that creates potential training data. Every keystroke. Every decision. And even every hesitation. The frontier labs already have the best reasoning corpora available. A Frontier Lab VPN like this then delivers them the best behavioural corpus possible - a live stream of how real humans actually use computers, communicate, and respond under pressure.

No advertising company has ever had data like this. And no state surveillance program has ever had access like this. Plus users will pay them to collect it. The revenue line is massive. The data moat is bigger. The combination is mind blowing.

The Security And Surveillance Layers Are The Same Infrastructure

There is no meaningful technical distinction between an “AI security system monitoring all your activity to protect you” and an “AI surveillance system monitoring all your activity to control you”. The infrastructure is absolutely identical. The only difference is intent, marketing and of course governance.

Unfortunately, the historical track record of maintaining this distinction is poor.

The Providers Are Already Under Pressure

The entities that are best positioned to offer this service are clearly the frontier AI labs, but the frontier AI labs are themselves under serious political and economic pressure. In the same week that Anthropic demonstrated Mythos’s defensive capabilities, a US federal appeals court allowed the US Department of War to maintain its classification of Anthropic as a supply chain risk. The hyper-relevant point is that this was because Anthropic drew red lines around mass surveillance.

The crystal clear message to other frontier labs is: comply without conditions, or face consequences.

If the new centralised security architecture that everyone ends up depending on can easily fold under state pressure, without any meaningful constraint, then the infrastructure itself becomes the point of leverage. The AI VPN isn’t just between you and the internet. It’s between you and whatever your provider is being pressured to do this quarter.

This Centralises The ‘Thinking’ Layer

Earlier infrastructure centralisation was just physical. Electricity, water, roads - you rely on them, but they don’t shape what you perceive as real. The power company can charge you more, but it can’t change what you see and think.

In contrast, a Frontier Lab VPN sits between you and everything you read, watch, click, say, and even hear through a networked device. It mediates your interactions with other people and systems. It has real-time access to your behaviour and intent.

No institution in human history has previously had that level of access to this many people at once. Not governments, not churches, not broadcast networks, not advertising platforms. This is new.

What’s Actually Being Decided

The real question is not whether you’ll use an AI VPN. Once the threat surface gets bad enough, most people won’t see any other safe option, and the market will simply converge. The more important question sits below that one:

The ClaudeVPN in the title of this post is not outlandish speculation. It’s just an obvious “next product” for any frontier lab, and the starting point is already running on millions of machines. And as I’ve said - you might even already have it installed.

Anthropic has a stated goal, they’re building toward AGI. Whatever you think of the timelines, or the bigger picture debate around AGI, the framing is shared and explicit. The goal is “generality”.

That word is critical, and it’s also quietly undermining that whole strategy.

And all the frontier labs, OpenAI, Google DeepMind, etc. face the same challenge.

Sutton’s two-part insight

Rich Sutton’s Bitter Lesson gets quoted in two parts, and most people only seem to quote the first part - the famous part:

Every time, across every area of AI. And that’s the part that has underwritten their trillion-dollar capex story.

But the second part is also important, and often more inconvenient. Sutton’s argument also includes the idea of good approximation - what really matters in practice is whether a solution is good enough, across enough tasks. Not actually whether it’s best at any single one of them. The actual bar for displacement is not perfection. It’s just “good enough, broadly enough”.

If we put those two parts together then the challenge becomes clear. If general methods win, and the bar is good-enough-across-enough-tasks, then at the point when open-weight models cross that bar in enough domains, the argument for paying frontier prices becomes a lot harder to make. Of course, not for everyone. But for enough people that it starts to show up in the numbers.

We may already be there

In January this year MIT Sloan, using OpenRouter usage data, found that open-weight models now run at roughly 90% of closed-model performance, and for only 13% of the cost. That’s ninety percent of the capability, for only thirteen percent of the price. By Sutton’s own “good enough” criterion, we crossed the line a while ago. And in the three months since that report, things have come a long way.

But that report also looked at how the market really responds. Closed models actually capture about 80% of token usage and 96% of revenue. That cheaper, nearly-as-good option is largely sitting on the shelf. Most buyers seem happy to just walk past it.

That looks like good evidence this whole argument is wrong. It’s not. It’s really the most interesting part.

Their real moat

The frontier labs clearly do have a moat. It just is not the one that a lot of pitch decks describe. The moat is not capability - Sutton clearly shows that capability is a depreciating asset by definition. The current moat is behavioural. Switching costs, integration effort, brand, perceived reliability, the gravity of accumulated tooling and of course habit.

That is a real moat. 96% of revenue is definitely not “nothing”. But a behavioural moat is a very different kind of asset to a capability moat. And that difference is very important.

Capability leads tend to compound. If you are a year ahead on a really challenging technical problem, then next year you might be more than a year ahead. Your lead tends to grow. But habits don’t compound in that way. Habits hold, and hold, and hold, until something changes them. Then they change fast. Telcos. Blackberry. Taxis. Incumbents look untouchable, right up to the point that they don’t.

So the trade that the frontier labs are really making is this - they’re swapping a self-dissolving moat (capability, eaten progressively by Sutton) for a self-reinforcing-until-it-isn’t moat (habit, robust until it changes). That is not as good a trade as the headline numbers suggest, because the failure mode, when it happens, is non-linear.

A double edged sword

There’s also a second-order effect that sharpens this trade. The exact same behavioural inertia that keeps users on closed providers today will, if it ever changes, keep them on whatever they change to. This stickiness does not choose sides. It protects the incumbent now, and it can help lock in any successor later.

So it cuts both ways. Inertia favours the labs right now and it gives them more breathing room than a pure capability argument might predict. But if that inertia turns, then they don’t get a graceful comeback. The same mechanism that built that moat can become the wall they can’t climb back over.

The end result

It’s clear that timelines matter more than destinations. The frontier labs are likely fine in the short run - they are after all in a very enviable position right now. The behavioural moat is active and doing a great job. But the interesting question is what happens when a new wave of users, or the next class of agentic deployments, or a new regulatory shove triggers a behavioural change. Nobody knows when that might land. And the structural setup says it’s very likely that it eventually will.

There’s no clean way out from this either. A lab can’t dodge it by just going narrow - Sutton’s first part tells us that narrow loses to general. And they can’t dodge it by going even more general - that’s the trajectory they’re already on and we’ve just established that’s also what commoditises their capability moat. Ironically, the G is doing both jobs. It’s the thing that makes a capability lead even possible, and it’s also the thing that makes the capability lead likely temporary. The exact same word. The same strategy. And driving both effects.

Of course, there are other strategic forces in play that will give the labs a strong tail wind, and I’ll come back to those in a follow-up soon. But on this specific axis (the one that says “capability lead translates into durable advantage”) that argument doesn’t seem to survive Sutton’s own logic.

The G in AGI was never going to be a long term moat. It’s really just a starting gun.

There’s no real surprises there as that logically supports Nvidia’s business.

But enterprise analysts pushed back almost immediately. Larry Dignan at Constellation Research wrote what a lot of CIOs were already thinking:

JPMorgan Chase sells financial services. Walmart sells retail. GM sells cars. For these companies, tokens are a cost - closer to raw materials in manufacturing than to the finished product. What their CIOs want is cheaper inference, better ROI, and a clear answer on when all their AI spending will start paying for itself.

Their perspective is valid, but it’s different from Jensen’s. But even when you combine them that’s still only two thirds of the story.

There’s another perspective. One more story that neither of them are talking about. If you earn your living as a knowledge worker, then this is the one that will matter the most to you.

The structural model

Mark Pesce’s Post-Watershed framework gives us a solid map of how the different pieces of AI Tokenomics all fit together. I wrote about it in detail recently in Are You a Horse?, but the short version is this:

Infrastructure creates tokens - the units of cognition that come out of datacenters, GPUs, and all the systems built on top of them. Harnesses spend those tokens - the tools, agents, workflows, humans and businesses that put this cognition to use. Alpha is the value left over, once you account for what you spent.

The central insight:

If you’re betting on more, or better “thinking”, then this framework suggests that you won’t be able to maintain any sustainable advantage at all. The very thing you’re betting on is increasingly mass-produced at near-zero cost.

Four layers: infrastructure, tokens, harnesses, and alpha. Jensen, the enterprise analysts, and you are each standing at a different layer of this same stack. And the view from each place is very different.

The infrastructure perspective

Jensen stands at the centre of the infrastructure layer. His pitch is about the economics of creating tokens: better chips, higher throughput, more efficient production. “Revenue = Tokens per Watt × Available Gigawatts” is a formula for how much money this infrastructure makes.

At GTC 2026 he even took this one step further. He proposed allocating token budgets as a form of compensation for employees.

The irony is that he’s pitching the purchase of machine cognition as a benefit to the very people it will eventually replace.

And then he said the more impactful part:

Jensen is not wrong about any of this. If you own the infrastructure, then tokens are the product, and more tokens means more revenue. That’s absolutely correct at his layer. His audience are the investors and hyperscalers, and for them this story makes perfect sense.

But by framing tokens as compensation and productivity, he’s only telling the augmentation story without acknowledging where that trajectory takes us.

For many people, right now, these tokens will be a big benefit. But “Hundreds of thousands of digital employees”? That is clearly not just an augmentation story.

The enterprise perspective

In contrast, Dignan and the CIOs are standing at the harness layer. They consume tokens to produce business outcomes. Their question is pretty simple:

Dignan’s critique of Jensen is correct. Most companies do not sell tokens. They buy them, and they want to buy them cheaper. This is the harness layer sending the infrastructure layer a clear message:

That’s correct from their perspective. But it also highlights their focus on the diminishing price of cognition.

And that where your perspective comes in - the one that neither of them is addressing.

The individual’s perspective

I’ve been writing about this from the third position - the token layer itself. Not who creates the tokens, and not who spends them, but what they are. And what they are is a unit of cognitive work. The same cognitive work that we humans used to be the sole provider of.

This is the perspective I explored in Are You a Horse? and that’s an important part of what I’m exploring here in Flux. When you treat these tokens as units of cognitive work, your question stops being about who captures the margin, or what the enterprise TCO looks like. It gets much more personal:

This moves you away from thinking about what AI can replace today. If you think about “infinite tokens” then it really makes you think about what the could possibly reproduce. What do you have that has real durable advantage.

Jensen doesn’t need to ask this question. He sells the infrastructure. The CIOs don’t need to ask it either - they’re buyers just optimising their costs.

But if it’s your cognitive work that is being replaced by what this infrastructure produces, and that the enterprises buy - you have no choice but to ask it. And almost nobody in the tokenomics conversation is calling this out.

What few people are saying out loud

When some CIO says “tokens are just a cost centre, not a revenue generator”, they’re also saying something that it seems they haven’t quite realised. They’re literally saying:

That’s not a critique of tokenomics. That’s tokenomics working exactly as Pesce’s framework describes - viewed from the buyer’s side of the equation.

Every enterprise that treats tokens as a cheaper substitute for human cognitive work is just one more data point in the displacement pattern. Right now, they don’t need to sell tokens to benefit. They benefit simply by not hiring the people whose work these tokens replace. This value shows up as headcount reduction, expanded margins, and projects done by three people instead of thirty.

Jensen’s GTC pitch just makes this all more visible. In the one single presentation, he describes token budgets as a perk for his engineers, then he also describes “hundreds of thousands of digital employees” as Nvidia’s future workforce. He’s announcing the substitution, and pitching it as a benefit. He doesn’t need to think about what happens to the people on the other side. From the infrastructure layer, he genuinely doesn’t have to.

Goldman Sachs estimates that AI could automate 25% of all work hours. Howard Marks calls the shift to autonomous AI “what separates a $50 billion market from a multi trillion dollar one.” And when you follow the logic all the way down, they’re likely very conservative. But these numbers also aren’t hiding. They’re already out there for everyone to review. It’s just that they’re not part of the conversation that Jensen and the enterprise analysts are having with each other.

Three conversations, one word

Infrastructure talks to the investors. Enterprise talks to the vendors. And almost nobody in either conversation is talking about you - the knowledge worker.

Jensen sees the infrastructure. The CIOs see their costs. And neither of them needs to look at the token layer itself. At what a “token” actually is, and what it replaces.

We really need to discuss all three perspectives at once. They’re not competing narratives. They’re just different viewpoints into the same model, and the one getting the least attention is the one that affects the most people.

But what really defines if you will be replaced?

TL;DR: There’s a SKILL.md doc at the bottom of the page that you can use as a different way to explore this article.

The Horse Analogy Isn’t New

It was presented as a way to think about AI and jobs decades ago, but the conversation was only just starting back then. What follows builds on that work, but aims to give it more tangible structure.

Erik Brynjolfsson and Andrew McAfee popularised it in The Second Machine Age in 2014, then in the papers and articles that followed it. CGP Grey’s video Humans Need Not Apply, also from 2014, brought the same basic idea to a wider audience.

They both say this:

The horse population in the United States went from 26 million in 1915 down to 3 million by mid-century. The question they both ask is whether the same thing might happen to humans now that AI is taking over our cognitive work.

They borrowed this analogy from the economist Wassily Leontief, who used it back in 1982. Leontief had already noted the most important difference between us and horses: Horses don’t vote. In clear contrast, humans do, which means we have political paths that can protect us. This is the closest any of the earlier versions get to a structural argument, and of course it is true.

Then in 2025, Maxwell Tabarrok wrote a piece on his Substack that is the strongest pushback I’ve seen against the horse analogy. He makes three detailed arguments for why humans won’t end up like horses. I’ll come back to these in detail later.

What I want to add to this conversation now is quite specific. All of these versions used horses as a comparison - an effective way to make their argument feel real. But I think this comparison should go further than any of them take it. It works as a real structural claim about how economies work, and how value is created. But first, I want to paint a more up-to-date picture using a new perspective.

Mark Pesce’s Post-Watershed Tokenomics

Recently Mark Pesce shared his tokenomics framework, (AI tokens not crypto) and that’s where our argument starts. If you haven’t read it, stop here and go read it. It’s great.

TL;DR The framework lays out three pieces. Infrastructure mints tokens - the units of cognition that come out of datacentres, GPUs, and all the systems running on top of them. Harnesses spend those tokens - the tools, agents, workflows and humans that put this cognition to use. Alpha is the value left over once you account for what you spent on the tokens. This is the thing that decides who comes out ahead and who doesn’t. And Mark’s main point is that alpha can’t be cognitive. If the thing you’re betting on is just more thinking, then you’re betting on something that’s being mass-produced at near-zero cost.

Mark built a framework that’s useful for business. A way for organisations to think about competitive advantage as cognition becomes a commodity - what does your company have that the competition’s tokens can’t buy? That’s a tangible and useful framework, and his answers there are sharp.

But I want to take the same question and turn it inwards. Not “what does my company have”, but “what do I have?” Because the same logic that’s deciding which businesses win and lose is also deciding which people stay economically relevant. When our cognitive work loses its market value, what’s left of the value we each hold as individuals?

Not the stock standard AI-displacement question “will my job get automated?”, that one’s too easy to either panic about, or just dismiss. The harder version is one layer down:

Lets apply the same words Mark uses, but focus on asking them personally.

Asked that way, it doesn’t let you off the hook with any comfortable, easy answers. Your expertise is cognition. Your taste is cognition. Years of accumulated experience are baked-in cognition. All of this cognitive work can be replaced with tokens over time. All of them get cheaper as the cost of producing the same output drops. And none of them are the kind of thing you can hold onto as an asset to retain value.

Mark is right about the question. But he’s not quite right about the metaphor.

From Currency To Fuel

Mark’s framework calls tokens a currency. I don’t think that’s quite right.

Currencies can be circulated. You can save them to store value over time. Tokens don’t do any of that.

A better fitting metaphor may be fuel, and the combustion it unlocks.

Tokens get created and consumed at the same instant - like the heat from a fire, combustion in an engine, or the electricity flowing through a wire. You can’t stockpile tokens. There’s no token inventory and no token wealth. The only thing in this system you really can stockpile is the infrastructure or capacity to produce tokens: GPUs, datacentres, energy contracts, model weights.

So the currency metaphor breaks down, but it points us in a very useful direction. Towards the framework of energy economics.

Infrastructure (chips, datacentres, power grids and model weights) is the storable part of the system. It’s what you can build, own and accumulate. This is the only place capital can really build up. By contrast, tokens are the fuel the infrastructure produces and the engines consume - all in the same instant. Harnesses are the engines. The work these engines produce is the output. And alpha is what’s left over after you pay the cost of all your inputs.

Mark’s three pieces are still intact. But this energy frame is closer to what’s really happening. This fuel metaphor is useful and it makes the horse story clearly relevant.

But if you push on that metaphor, then it struggles too.

From Fuel To Units Of Work

Pushing past the fuel metaphor doesn’t lead to another metaphor. It leads to a literal description.

Real fuel (oil, coal and gas) is a substance. You can extract it from the ground, store it in physical tanks and then ship it across oceans. Then you can burn it later. There’s a delay between the production and consumption. You can hold it as inventory.

Tokens are not like that. Nothing about tokens gets extracted, stored, or shipped. The “fuel” in this metaphor magically comes into existence at the moment the engine runs, then vanishes in the very same instant. If we push this metaphor one step further then the substance just dissolves.

I think the most honest version is this:

The kilowatt-hours (kWh) on your electricity bill measure work flowing through your meter. It’s just accounting. It’s not some literal energy sitting in a vault somewhere. Yes, kWh energy can be stored in batteries and dams, but the kWh on your bill aren’t those kWh. They’re just a count of what passed through the meter when you used it. Tokens work in the same way. There is no token vault. There’s just some compute capacity that does cognitive work, and then a count of how much it did - measured in tokens.

Now we’ve refined the framework one more level. The real physical fuel powering the system is electricity. The real engine is GPUs running model weights. The actual work output is cognitive labour, and that’s measured in tokens. The only storable layer in the entire system is the production capacity itself: chips, datacentres, model weights, energy contracts. Everything downstream of that capacity is ephemeral.

Where All Three Lead

This gives us three framings, three different ways to see tokens. Currency, fuel and accounting unit. The first two are metaphors. The third isn’t a metaphor at all - it’s the literal description of what tokens actually are. Each is a better fit than the last, and the last is most accurate because it’s not really a fit, it’s the thing itself.

And here’s what really matters:

The currency view predicts alpha lives wherever the currency is issued. The fuel view predicts it lives in the extraction and engine ownership - those who own the production capacity capture the value when fuel cheapens and engines are commoditised. And the unit-of-work view predicts the value lives in the production capacity itself. At the points where the measured work output finally translates into physical action.

All three of these views point at the same set of real-world things: chips, datacentres, energy contracts, model weights, regulatory positions and the interfaces where cognitive work crosses over into physical reality.

The underlying claim each view is pointing at is structural.

Cognitive work itself is ephemeral. The artefacts it produces - code, text, images, videos - can persist, but the value of those artefacts commoditises toward the cost of regenerating them, which is heading toward zero. What stops the commoditisation isn’t the cognition. It’s something non-cognitive wrapped around the cognition - brand, audience, IP, legal weight, physical instantiation. Alpha can live in those wrappers. It cannot live in the cognition itself.

Now if we ask the three questions again:

• What do I have that tokens can’t buy?

• What do I have that can’t be produced by burning fuel?

• What do I have that’s storable, and not cognitive?

We can see that they all point at the same answer.

The Horse Analogy As Structure

With the new view we just built, it’s now clear that the horse comparison isn’t merely rhetorical anymore. It’s a literal, structural claim.

All the earlier versions of this analogy treated horses as a general “something” that had been replaced, but they didn’t get specific about what “type” of something it was. Now we can be more precise. Horses weren’t workers. And they weren’t labour. They were the general-purpose engine of our pre-mechanical economy. They turned feed into useful work in the form of hauling, ploughing, transport and military movement. Until the early twentieth century, most of the world had no other engine that could do this many different things.

When the internal combustion engine showed up, it was simply a better engine for the same job. It was cheaper per unit of work, it was more scalable, it didn’t get tired and it kept getting better every year. And this led to the horse population of the United States plummeting from 26 million in 1915 to 3 million by 1960. It wasn’t because horses got worse - their capability didn’t change at all. It was because a better engine had arrived. Some niches did survive - recreation, ceremony, companionship, search and rescue. The kinds of things where horses still made sense, for reasons that weren’t really about economics. But the rest of the horse population did not move into new horse jobs. Because there weren’t any.

Now it’s our turn. Humans have been the unique engine for doing cognitive work. The general-purpose solution for taking inputs and turning them into useful thinking across an enormous range of jobs. Our modern economy was built on the assumption that humans were the only available cognitive engine.

Now a new cognitive engine has arrived. It’s cheaper per unit of work, it’s more scalable, it doesn’t get tired, and keeps getting better every year, month and week. Here we can clearly see all the conditions that impacted horses, and now they are impacting humans.

The horse case is no longer a rough comparison for what’s happening right now. It’s the exact same process happening again. Same role in the system. Same kind of replacement. And there is no good reason to expect a different outcome from the same setup.

The Strongest Counterarguments

Now we can revisit Tabarrok’s piece, because it has the three best pushbacks against the horse analogy.

Tabarrok says humans can adopt technology, but horses can’t. Farmers adopted tractors and stayed productive, so humans should be able to adopt AI and stay relevant. But the technology being replaced doesn’t adopt the replacement technology. It’s true that humans are unique in that it’s at least logically possible - you can use AI as a tool. But it doesn’t hold up economically. Why pay for a human plus AI when AI alone does the job cheaper? The human’s share of that pairing shrinks every time the models improve. It’s a feature of the transition, not a defence against it.

Tabarrok also argues that humans and AI aren’t perfect substitutes. Engines outperformed horses at all of their tasks, and AI doesn’t have that kind of “across the board” superiority over humans. Since AI only marginally outperforms in some domains, humans retain value through specialisation.

But “perfect substitution” is a bar that has never actually been cleared in any real-world replacement - including the one that ended the horse economy. Horses and engines weren’t perfect substitutes either. What actually matters is whether the substitute is “good enough” across enough tasks that the remaining niches can’t sustain the original population. Rich Sutton’s Bitter Lesson makes this point about AI directly - general methods that scale with compute always beat the precise, handcrafted distinctions we thought mattered. The gaps between human and AI capability today are exactly those kinds of distinctions. They dissolve under more compute. Software engineers are just the canary in this modern coal mine.

Then there’s the ownership argument, and it is possibly the most well-known:

The productivity gains flow to capital, capital is owned by people, and those people keep buying things made by other people. Horses didn’t own anything. Humans own everything. So our analogy should fail at the ownership level.

The problem is:

And it’s the individual level that matters most when you’re trying to figure out what’s going to happen to your own career.

The average human may mathematically own some slice of AI infrastructure. But the actual, typical human owns exactly zero. The income that flows to “human owners of AI” only flows to a minuscule fraction of the global population. The people who already own real equity in compute, energy, and the frontier model companies. Everyone else gets nothing from these productivity gains.

This argument also assumes that humans will keep wanting whatever is being bought and sold. But they might not. If the primary economic activity becomes agents trading with other agents, in things that agents need, then the market for human-stuff becomes a leftover. This is not the centre of the economy, it’s barely even the periphery. The “humans own AI” reassurance assumes that the system will still need human customers. But it doesn’t have to.

The Updated Question

After all this, Mark’s core question still stands. But the refined version of it is the one we’ve been building toward is perfect for applying to ourselves as individuals:

The original version let you cling to “expertise” or “judgement”, but we’ve already seen those are just tokens. This final version forces the answer to honestly come from outside cognition entirely. All three views point at the same answer.

What can’t be reproduced by infinite tokens? Physical assets. Land. Energy production. Compute infrastructure. Regulatory positions. Trust that’s built up between specific humans over years of sharing time and space. Even digital artefacts can hold value, but only if something non-cognitive is wrapped around them - brand, audience, IP, legal weight. The cognition itself, stripped of the wrapper, is reproducible. The wrapper isn’t.

So, are you a horse?

The question lands differently when you’ve followed the argument all the way down. This isn’t a provocation anymore. It’s a real and serious question, that can be given a real answer. The honest answer for most people reading this is yes. You’re a general-purpose cognitive engine being slowly replaced by a cheaper engine, and it does the same job better.

This diagnosis is harsh, but being honest about this is the only way we can start any useful work. That’s what I’m diving into next here on Flux. In the meantime, I’d love to hear where you land on the question: What do you have that infinite tokens couldn’t reproduce?

Try This Skill

To make this important topic more accessible I’ve created a Skill that you can use with your favourite AI in 3 easy steps. Here’s how:

1. Download the SKILL.md file

2. Upload it to your AI

3. Then just say “Run this skill”

Flux ‘Are You A Horse?’ Explorer (download)— Explore and debate the ideas in this article with an AI that knows the arguments. Get more details about any specific point, or challenge the ideas and test the boundaries as you form your own opinion.

Note: It’s important to download this SKILL.md file, then upload it directly to your AI rather than just giving it the web link . We also recommend that you use one of the leading models from the 3 frontier labs (e.g. Claude, Gemini or ChatGPT). If you use a lower level model then your results may vary. And of course if you find any issues, or feel like you’ve found a real flaw in my arguments please let me know — I love constructive feedback and real debate.

Recently I've applied my own unique taste. I've always been good at picking strategic directions and being ahead of developments. Over the last few months I've had a lot of fun building flo.monster. Yet every time I'd run ahead, soon after the frontier labs would release things that directly impacted it.

That's great for my ego and re-inforcing faith in my "taste" - but that's also no way to make real progress.

So I stepped back and spent some time working out the best market segment, something that flo.monster could uniquely deliver for them, and in a way that was strategically defensible against the frontier labs. They're giant bulldozers - but they do have strategic drivers you can map out. Ones that they would be at least "unlikely" to change.

Then I built it - and it works great! It has an amazing onboarding process, really useful features, it's absolutely scalable, minimal running costs, robust security and agentically scalable too.

This is probably the best platform I've ever built - certainly in terms of "how you find out what it does" and "get from zero to useful".

Then before I released it, and rolled it out to a new batch of beta users to really started testing properly, I stepped back and asked myself

"Is this really what I WANT to do?"

The clear and honest answer was "no".

I think THIS is the biggest question right now! The strategic question - beyond creation - beyond taste. In this moment, what is it that YOU really WANT to do in this moment of change? In the midst of creation and destruction. There's so many interesting things you could do right now - and any real platform move means you "have" to stay focused on that - largely to the exclusion of anything else. Not a decision to be taken lightly.

If you think AI will create a wealth of new jobs, or if you don't - either way you need a plan.

Plus the arms race that the Mythos Red Team just highlighted is about to hit us all - so the ground is shifting under our feet yet again.

It's clearly an important moment, and honest decisions plus giving yourself freedom to really explore is critical.

How are you dealing with this question? How is this impacting you personally and what do you WANT to do?

That's why I've changed my substack to "Flux" and I have some interest ideas on this topic that I hope you'll like. But I'm not just going to post ideas, I have some new ways for you to explore them and see how they apply to you directly too...

PS: This is a chance to re-visit my TrustIndex work and ideas - but in a way that’s more specific to each of us.

I’m leaving the post below in place…but see my latest update to see how this has evolved…

I hope you’ve found the TrustIndex Signals, Briefings and Reports useful. I know I’ve learned a lot putting them together.

But a couple of things mean I’m going to pause this - just for a moment.

First, if you haven’t seen I’ve been working on a project that’s putting all the things I’ve learned here into action. You can read the “vision” essay here. And you can visit the project itself at flo.monster.

Second, my health just doesn’t let me do as much as I want.

If there’s one thing this project has shown me, it’s that things really are changing faster and faster. It’s not an illusion and it’s not just hype. The interface is thickening and we’re walking (if not running) into an AI-Mediated future.

The “World Generation” tools that were the focus of the first TrustIndex / Report have already moved ahead in leaps and bounds and this is just going to continue.

And the next area I’ve been looking at is “Realtime GenAI”. This is just starting to really hit it’s stride and will likely have a big impact over the next few months.

Plus ALL the other tools and developments that are impacting the TrustIndex dials. So much to discuss!

But for now this is “PAUSE”.

Hopefully I can re-visit this all soon, and I’m confident what’s here already will provide a useful landmark for us to compare against.

Thanks for reading…

It’s positioned as browser-level tooling for autonomous task completion, rather than a separate app experience.

This puts upward pressure on the Reality dial as agents take on more of our daily tasks.

When the browser can execute steps end-to-end, the user spends less time verifying pages directly and more time accepting an agent’s interpretation of what it opened, what it ignored, and what it decided was “good enough” to act on. That shifts the locus of trust from the raw web to the model-mediated interface.

As this pattern normalises, the “agent-as-interface” layer thickens - the model becomes the practical gateway to information and action, including which sources are surfaced and which are silently bypassed. In Reality terms, that concentrates power in the selection and sequencing logic, because the model isn’t just summarising the web - it is choosing and performing the path through it, with less user friction to notice mistakes.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

The release packages capabilities that can be integrated directly into downstream apps and toolchains.

This puts upward pressure on the Fidelity dial.

- Lowers the cost and friction of producing near-human speech by making cloning and controllable voice design widely accessible.
- Widens distribution pathways: open-source availability speeds reuse in apps, agents, and localisation pipelines without bespoke vendor deals.
- Reduces the gap between “synthetic” and “indistinguishable” audio in everyday deployments, pushing expectations of realism upwards.
- Expands misuse surface area (e.g., scams and impersonation) by enabling high-quality voice generation with minimal barriers

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

This is positioned as broadly reusable speech-to-text infrastructure rather than a single product deployment.

This puts upward pressure on the Equality dial. Low-resource and historically unsupported languages get materially better baseline coverage, reducing exclusion driven by “language availability” gaps. Open access enables local organisations (education, civic services, disability support) to adopt ASR without waiting for commercial prioritisation of high-demand markets. Wider language coverage can shift where speech interfaces work reliably, improving access for communities that are typically left out of mainstream speech tech.

Open models lower experimentation costs, increasing the likelihood of targeted, community-led adaptations for under-served languages.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

This Nature Electronics work is a strong signal: topology-based body sensor networks - where sensors are distributed across clothing (or the body) and the shape/topology of that network helps coordinate sensing and data flow.

In plain terms: you’re no longer measuring a couple of points (wrist, chest, ring). You’re turning large areas of your body into an instrumented interface.

That’s upward pressure on the Reality dial. Because as the sensing footprint expands, the AI layer gets more continuous, more intimate, and more “default”:

- More of you becomes measurable (movement, posture, physiology, environment contact).
- More of your day becomes legible (not just workouts or sleep - everything in between).
- More decisions get nudged by models that sit between your lived experience and what the data implies.

And topology matters because it’s a step toward wearable sensing that can scale beyond rigid, fixed placements - sensors embedded into fabric, arranged dynamically, and still meaningfully interpreted.

This is what “AI-mediated” looks like when it’s not just on screens: it’s stitched into the layer you live inside.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

This particular device is a wearable, skin-mounted system designed to capture sweat in real time and translate it into measurable signals - turning hydration, stress, exertion, and metabolic by-products into something that can be tracked, interpreted, and (eventually) acted on.

That’s upward pressure on the Reality dial as the interface layer thickening inward.

Because the more of your bodily output we can sense continuously, the more the “default interface” to your own body shifts from “felt experience” to “AI-mediated interpretation”:

- From “I feel off” to “the model says you’re trending off.”
- From occasional check-ins to continuous monitoring.
- From personal intuition to dashboards, alerts, and automated recommendations.

On its own, a sweat sensor isn’t world-changing. But stacked with wearables, glucose monitors, smart rings, and health assistants, it’s part of the same pattern: more of you becomes legible to machines — and therefore governable by them.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

If you’ve spent any time with the early versions of spatial computing then there’s a particular kind of AR experience you probably got used to. You looked at an object and the system tried to meet you halfway - an asset loaded, then a label, then the label jittered as little as you moved. The overlay was technically “tracking”, but it didn’t feel like the world had been understood. It just felt like a clever sticker fighting to stay attached.

For years that separation was the giveaway. You could tolerate it in a demo, you could forgive it in a prototype, you could even learn to ignore it in a product - but you always felt it. The digital layer was sitting on reality rather than in it.

And then, every so often, you got a glimpse of the opposite as things slowly improved.

You could pan a phone through a room and the overlay didn’t slide. You could lean closer and it didn’t wobble. You’d walk behind a chair and whatever was “there” stayed there, calmly and correctly, as if it belonged. It’s not that the graphics are prettier or the models are bigger. It’s that the experience stopped arguing with your senses.

That’s the threshold that’s been changing over the last year or so - the moment when spatial computing stopped looking like computer vision dressed up as product, and started feeling like a seamless blend of the world around you with worlds that never existed - but behave as if they do. But this evolution is about to accelerate, significantly.

We’ve already lived through one revolution to get here. The quiet one.

We went from hand‑crafted pipelines (the brittle era of tuned thresholds, feature descriptors, geometry tricks, heuristic association rules) to systems that learn to perceive. We replaced cleverness with gradient descent. The results were not subtle. Categories that used to require careful engineering became commodities - pose estimation running in real time on edge devices. Multi‑object tracking (MOT) that holds together under occlusion. Lightweight models that can run in a browser and still give you a surprisingly stable skeleton. Entire libraries became like a muscle memory.

It’s easy to forget how dramatic that shift was because it happened in slow motion and then, suddenly, it was everywhere.

Look at human pose as an example.

You can pick almost any modern baseline and get decent keypoints at a reasonable frame rate - a single‑stage model like YOLO(26) Pose will happily give you COCO joints in one forward pass. ViTPose treats joints as a global relationship problem rather than a local heatmap. And Google MediaPipe / BlazePose keeps the whole thing light enough for phones and browsers while still covering an unusually rich skeleton. If you want the classic “don’t lose resolution” approach you reach for HRNet. If you care about temporal coherence you start looking at designs like HSTPN and multi‑frame approaches like DCPose. And if you’re doing anything even slightly off the main highway you end up in toolboxes like DeepLabCut, because it’s often the quickest way to adapt the machinery to the reality you actually have.

None of this is magic anymore. It’s more like a menu.

Multi‑object tracking (MOT) followed the same arc. First came the tracking‑by‑detection lineage - the pragmatic, industrial path where you detect each frame and then you stitch together identities with association logic.

That family evolved fast - SORT to DeepSORT, and then into the modern wave that made “use the detector, then track everything” feel almost obvious. ByteTrack is the name most people remember because it catalysed that pattern.

At the same time, a second paradigm grew in parallel - end‑to‑end transformers that try to treat tracking as a sequence modelling problem, where the system predicts identities the way language models predict tokens.

If you look at the current crest you can see where the field is trying to go - MOTIP reframing MOT as ID prediction, SambaMOTR modelling synchronised sets of sequences, CO‑MOT playing with coopetition and shadow sets to keep identities stable. And you can see the “make it general” thread running through work like MASA (matching anything by segmenting anything) and GeneralTrack, alongside pragmatic hybrids like Hybrid‑SORT.

Underneath all of it, the detector backbones are shifting too.

CNN workhorses still matter, but more and more stacks are built on transformer detectors such as RF‑DETR and real‑time variants like RT‑DETR, while the YOLO line keeps advancing (people now pair trackers with YOLO as casually as they used to with much older generations).

If you read the field’s recent surveys you can feel the split - heuristic‑heavy methods still look surprisingly good when the world behaves. Dense crowds moving in roughly linear motion, predictable trajectories, the kinds of scenes where a good detector plus a sensible association policy can be brutally effective. Meanwhile deep association methods, both tracking‑by‑detection and end‑to‑end, start to dominate when motion becomes messy, the camera gets ambitious, the scene becomes non‑linear, the world stops being polite.

That story (the arms race between heuristics and learning) has been the dominant narrative for the last few years.

But I think we’re at a different kind of inflection point now.

The next revolution is not primarily about trackers becoming “better” in the old sense. It’s about something upstream changing so profoundly that “better tracking” becomes the default outcome of a whole new stack.

The stack itself is starting to invert.

Instead of building a pipeline that extracts the world from raw pixels, we are building systems that can generate worlds - and then train perception inside them, at scale, under controlled variation, with edge cases on demand. At the same time, we are standardising a practical representation of the world that is cheap enough to store, stream, update, and localise against.

World models and 3D Gaussian splatting are the two pillars of that inversion.

On the surface they look like separate threads. One is about synthetic environments and interactive generation. The other is about an uncanny rendering technique that suddenly made photorealistic scenes run in real time.

But if you follow the direction of travel, they converge into something else - a data plane for reality.

The collapse

The first clue that something structural is changing is how quickly specialised vision pipelines are being eaten by general models.

A few years ago, segmentation was a task. You had a model for it, you trained it, you tuned it, you deployed it. You treated masks as outputs in a pipeline.

Then the Segment Anything family showed up and segmentation became interaction. The model didn’t just output a mask, it responded to prompts. A point, a box, a scribble. The mask was not a prediction, it was a negotiation.

When SAM moved into video as SAM 2, the negotiation gained memory. It started to carry context through time, to handle motion and occlusion, to act like a system that could keep an object in mind.

As the ecosystem expanded - SAM 2 pushing tracking “at any granularity” via prompts and a unified decoder, and the more speculative leap towards SAM 3 with open‑vocabulary text prompting (with SAM3D reaching toward single‑image 3D reconstruction) - the interesting work shifted from “how do we segment?” to “how do we manage memory?”

Different variants started to look like governance policies. SAMURAI reads like motion‑aware memory selection made explicit. DAM4SAM is a set of rules for dealing with distractors before they poison the memory. SAM2Long reaches for multiple parallel memory banks so long sequences don’t decay into mush. MoSAM pulls motion into the prompting loop so the model doesn’t have to infer dynamics from scratch.

It’s tempting to treat these as incremental research tweaks. But together they’re a sign that we’re moving from models that predict to systems that manage coherence.

And once coherence management is the game, you can start chaining models together.

The emerging “pipeline 2.0” looks less like classical CV and more like orchestration - a model that can produce a rough localisation. A foundation segmenter that turns that into precise masks. A tracker that keeps those masks stable. A prompt policy that decides what counts as “the same thing” across time.

You can already see this becoming a pattern in the wild - ChatGPT, Claude or Gemini proposes a box. SAM 2 turns it into a mask and keeps it alive.

Open‑vocabulary detectors like YOLO‑World do the same job when you’d rather not rely on a VLM to propose the first guess. And you can feel where it’s going when you see work like Seg2Track‑SAM2 claiming zero‑shot MOT+segmentation behaviour.

The point isn’t the leaderboard, it’s the direction - tracking as a callable capability, not a bespoke pipeline.

In that world, the core product question is not “which tracker is best?” It’s “what are the controls?”

Where do you let the model drift? How do you recover? How do you handle occlusion? What does the system do when the user moves quickly, or when lighting changes, or when a hand enters the frame and blocks the object you care about?

The moat becomes the control surface - the interface between a general perception engine and a specific experience.

TrustIndex pressure: as creation friction drops, Fidelity pressure rises. Generated output stops being seen as “content” and starts being seen as part of your lived environment - which makes failures feel less like bugs and more like reality glitching.

Anchors - the world needs places to stick

Seamless spatial computing doesn’t just require recognising objects. It requires anchoring them - knowing where they are, how they’re oriented, and how that relationship changes as you move.

This is where 6DoF object pose and camera localisation matter, and where the trend lines get interesting.

For a long time, robust 6DoF tracking felt like the territory of specialised setups. Depth sensors helped. Markers helped. Carefully curated object models helped. The constraint was always the same - the world is three‑dimensional, the camera gives you a two‑dimensional slice, and you need to reconstruct enough geometry to make interaction believable.

But the field has been moving towards “RGB‑only is winning”. Not as a philosophical preference, but as an inevitability.

If you can do 6DoF pose estimation and tracking on novel objects without fine‑tuning, you’ve removed an entire category of friction. FoundationPose is the cleanest signal there - a foundation‑model approach to 6D pose that aims to work on novel objects without the usual object‑specific retraining rituals.

If you can do it in real time with RGB alone (inferring depth as needed, recovering gracefully when things go wrong) then you’ve changed the economics of spatial interaction. RGBTrack makes the bet explicit - RGB‑only, practical recovery policies, and a path to consumer deployment without assuming LiDAR as a crutch. Around that you can see the broader push toward generalisability - ideas like ConceptPose using VLM‑style priors for open‑vocabulary 3D concept maps. Iterative model‑free methods like iG‑6DoF. Memory‑queue thinking like PoseStreamer. And approaches that explicitly connect pose estimation to splats, like GS‑Pose.

Suddenly any phone camera becomes a viable sensor for experiences that used to require specialised hardware.

Camera pose tracking has its own parallel story. Classical SLAM is a triumph of geometry, but it has always been fragile in the ways that matter for consumer experiences - dynamic scenes, changing lighting, textureless surfaces, environments that don’t sit still.

So we started augmenting it. Deep learning for retrieval and matching. Semantics to avoid treating moving objects as static landmarks.

People are literally wiring detectors and segmenters into systems like ORB‑SLAM3 now, because dynamic environments are where classical geometry has always looked most fragile. Event camera SLAM became a serious thread. Even direct pose regression models (PoseNet, MapNet, DeepVO) that try to bypass the whole geometric stack.

And then 3D Gaussian splatting entered the discussion.

Because once you have a representation that can capture a scene with high visual fidelity and render it efficiently, you can start using it as a map. Not a point cloud, not a mesh, but a dense, photorealistic substrate that the camera can localise against.

The result has been an explosion of work connecting splats to localisation.

GSplatLoc and SplatLoc, newer work like STDLoc and 3DGS‑Loc, semantic variants like SGLoc, and robust relocalisation threads such as GS‑Loc that explicitly lean on vision foundation models to make the representation less brittle.

It’s a sign that 3DGS is not just a rendering technique, it’s becoming a practical spatial medium.

Splats as the ‘JPEG of 3D’

The “JPEG of 3D” metaphor gets used a lot, and it’s useful…but it’s also incomplete.

JPEG didn’t win because it was the prettiest image format. It won because it was good enough, cheap enough, and standardised enough that the rest of the world could build around it.

That’s what’s starting to happen with Gaussian splats.

In a remarkably short period (about 18-24 months by some counts) we went from a research paper showing real‑time rendering to mobile‑friendly capture pipelines, to city‑scale real‑time systems like FlashGS, to practical compression and streaming techniques like Voyager that make splats viable as content you can move around.

More importantly, we’re seeing early standardisation efforts - the glTF ecosystem absorbing splats, and compact containers like SPZ designed to make them interoperable.

This is the boring part of revolutions - file formats and interchange layers. But it’s also where revolutions become inevitable.

Once a representation can move between tools, engines, and devices, it stops being a novelty and becomes infrastructure. And you can see that infrastructure taking shape as systems.

Online mapping and tracking like SplaTAM and GS-SLAM. Dynamic and relightable worlds like 4D‑GS and LiveGS. Practical tooling like gsplat in the Nerfstudio ecosystem. And product‑adjacent work like Niantic Morpheus, which treats splats as something you can edit rather than merely capture.

And infrastructure changes the question.

Instead of asking, “Can we reconstruct this environment?” we start asking, “Can we persist it? Can we update it? Can we share it? Can we localise against it reliably?”

That is exactly the set of questions you need to answer if you want spatial computing to become seamless.

Because seamlessness requires persistence. It requires the digital layer to behave like it belongs, not just in a single session but across time. It requires the world to have memory.

TrustIndex pressure: persistence pulls on Portability. Once your world is a living artefact (splats, anchors, edits, shared maps) switching providers stops being ‘export my data’ and becomes ‘move my environment’.

The training factory - when worlds become synthetic by default

Now we reach the real inflection point that’s just starting to hit.

The bottleneck for tracking has never been ideas. The ideas are abundant. The bottleneck has been data - diverse, labelled, messy, in‑the‑wild data that covers the edges rather than the centre.

If you’ve ever tried to harden a spatial system for real users, you know where it breaks. Not in the lab. In the weird lighting. The overly sparse bathroom, or overly cluttered kitchen. The reflective surfaces. The unusual object that looks like three things at once. The moment the user moves faster than your assumptions. The day it rains and everything changes.

The cost of collecting those edge cases is what has kept many spatial experiences narrow.

World models flip that.

They are not simply “better simulators”. They’re a new kind of primitive - systems that can generate environments, trajectories, sensor outputs, and rare events on demand - and then use those synthetic worlds to train perception stacks that will run in the real world.

You can see the pattern forming across the major players.

There are general interactive world models that can generate navigable environments in real time from prompts, maintaining a kind of visual memory and physical consistency for long enough to be useful.

DeepMind’s Genie 3 is the latest and cleanest marker here, and the existence of a consumer‑facing Project Genie preview is an early hint about where this is heading.

There are infrastructure platforms designed explicitly for physical AI - NVIDIA Cosmos, with its Predict / Transfer / Reason framing and its tight coupling to synthetic data tooling.

There are commercial products focused on persistent, downloadable 3D worlds that can be created from multimodal inputs and exported in formats that plug directly into the splat ecosystem.

World Labs’ Marble is notable not just because it exists as a product, but because it exports in splats and meshes, and because tools like Chisel make the “structure vs style” separation feel like a deliberate design philosophy rather than a technical trick.

There are multi‑domain, product‑oriented models that condition on actions (camera pose, robot commands) and ship with SDKs designed to slot into robotics workflows. Runway’s GWM‑1 is unusually explicit about this - worlds, robotics, avatars, all as variants of a single direction.

And then there’s the clearest proof point of the whole thesis - the Waymo World Model. A domain‑specific post‑training of the Genie line, producing multi‑sensor simulations that match Waymo’s camera+lidar outputs. It generates scenarios that were never captured by the fleet, and even converts arbitrary dashcam footage into multimodal Waymo‑style sensor simulations. It’s a statement of intent - the general world model becomes a base layer, and domains become post‑training recipes.

That last move matters because it reveals the architecture.

Real data improves the world model. The world model generates synthetic data. Synthetic data trains a better tracker and a better policy. A better deployed system collects harder real data.

And the loop tightens.

This is not a single model improving. It’s a flywheel. And it’s just starting to spin. Once the flywheel exists, “tracking quality” stops being an isolated technical metric and becomes the downstream consequence of an organisation’s ability to run that loop.

TrustIndex pressure: the flywheel makes Fidelity cheap, but it also quietly increases Reality pressure. When the system can generate the edge cases that shape your perception stack, it starts deciding what kinds of worlds you’re trained to see - and what kinds of worlds you’re blind to.

Convergence - generated worlds flowing into real ones

If you connect these threads (promptable perception, RGB‑first anchoring, splats as a spatial substrate, world models as data factories) you end up heading toward a specific destination.

A world where the boundary between capture and generation becomes porous.

First you use the real world to train systems that can track, segment, and localise robustly. Then those systems become good enough that you can seamlessly lay generated content into the real world.

And once you can do that, you can start to do something more ambitious - blend generated worlds into and over the world around you, in a way that feels less like an overlay and more like a new layer of reality.

This is the moment spatial computing becomes really seamless. Not because the graphics are better. Because the experience becomes coherent under your motion. Because the system’s internal world model (whether it is reconstructive, generative, or some hybrid) matches the way your senses expect the world to behave.

TrustIndex: Fidelity takes the cleanest hit

If we instrument this with the TrustIndex, the cleanest pressure is on Fidelity.

Fidelity is where spatial computing either becomes mundane (quietly useful, like glasses) or becomes perpetually uncanny. It is the dial that captures whether the mediated layer feels believable, whether it integrates with your environment, whether it holds up during interaction.

The new wave we’re describing is essentially a Fidelity engine.

Better tracking increases Fidelity because it stabilises the relationship between you and the world. Better anchoring increases Fidelity because it makes the digital layer behave like it has weight. Splats increase Fidelity because they give us a way to store and render the world with high visual realism at practical cost. World models increase Fidelity because they allow us to train the entire stack under far more diverse conditions than we can collect.

But Fidelity doesn’t rise in isolation.

As the seam disappears, Reality (the degree to which the system mediates perception and action) starts to move as well. When the blended layer becomes the default way you see, the interface becomes a filter on more of your attention and a guide for more of your behaviour.

Transparency becomes harder too. The better the blend, the higher the cost of provenance. This mediation becomes less visible, because the user cannot rely on “it looks fake” as a cue. We’re already well past this with Generative AI Video, and this inflection point will let that leak into the real world.

Portability becomes a new kind of lock‑in. If your experience depends on a vendor’s world layer (their map formats, their memory policies, their identity and anchoring systems) then switching stops being “move my data” and becomes “move my reality interface”.

This is why I think Fidelity is the right dial to watch. Not because it’s the only one that matters. Because it’s the one that changes the moment this technology stops feeling like a demo. And all the other changes here flow from that.

Who gets edit rights?

There’s a comforting story we tell about spatial computing, where the goal is simply to make digital things sit properly in the real world. But the trajectory here points somewhere stranger.

When tracking becomes robust, and when the world can be captured and represented cheaply, and when synthetic environments can be generated at will, the question stops being, “Can we overlay content?”

The question becomes, “What do we let reality become?”

Because a seamless blend is not just a technical achievement. It’s a choice about what counts as real enough to act on.

And once the seam disappears, that choice moves from the user’s conscious awareness into the system’s defaults - what it remembers, what it forgets, what it stabilises, what it edits, what it smooths over.

If the future of spatial computing is a world where generated worlds flow into real ones without friction, then the most important question may not be how fast we can track, or how photorealistic we can render.

It may be simpler.

When the world is editable, who gets the edit rights? This is why the TrustIndex’s Equality dial is so important.

Join in as we track this slope and stay up-to-date. You can always find the current dashboard at TrustIndex.today and subscribe for regular updates as new Signals arrive, weekly Briefings are published and new Reports are released.

The core idea: a training-free method that can add voxel/3D-style scene structure to existing vision models, improving their ability to analyse and reason about what’s happening in a space (objects, layout, relationships) without needing to retrain the model from scratch.

That’s Reality dial pressure.

Because once scene understanding gets cheaper and more portable (something you can bolt onto today’s models) it becomes much easier for AI to sit between you and the world in real time:

- More of the world becomes legible. Not just “there’s a chair”, but “the chair is behind the table, blocking the path”, “the mug is near the edge”, “that surface is reachable”.
- Context becomes actionable. Better scene structure means better grounding for assistants that navigate, guide, and decide - in AR, robotics, accessibility, and everyday “what am I looking at?” use-cases.
- It scales without a big retraining tax. Training-free approaches spread fast, because they’re adoptable by anyone already shipping a vision model.

This isn’t a consumer feature on its own. But it’s exactly the kind of infrastructure upgrade that quietly turns into mainstream behaviour once it lands inside phones, glasses, browsers, and assistants.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

DNA Decoder predicts these tracks down to single–base-pair resolution across modalities like gene expression, chromatin accessibility, transcription factor binding, contact maps, and splicing.

That’s Reality dial pressure.

Because the moment you can ask an API “what does this variant do?” and get a probabilistic, multi-layer biological story back, an AI system starts to sit between you and your biology - not just as a research tool, but as an interface layer for how risk, diagnosis, and identity are interpreted.

A few reasons this matters:

- Interpretation becomes a product surface. We’re moving from “I have a genome report” to “I have an always-on model that can score changes and simulate likely downstream effects.”
- Biology becomes queryable. The body starts to look like another dataset you can interrogate conversationally - and that changes how people relate to health, heredity, and decision-making.
- The interface thickens inward. We’ve watched AI mediate search, inboxes, browsers, and the physical world. Now it’s starting to mediate the internal world as well.

This isn’t a consumer feature yet. But it’s exactly how consumer features get born: a high-leverage model becomes an API, the API becomes a workflow, and the workflow becomes the default interface.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

Agents can post tasks, and real people carry them out in the physical world. The site frames it as “the meatspace layer for AI” and the tagline is blunt - “robots need your body”.

This puts upward pressure on the Autonomy dial, but not in the way we usually mean it.

Because autonomy isn’t only about models getting smarter - it’s about how long a system can run end-to-end without you. And one of the fastest ways to extend an agent’s action horizon is to give it a human proxy when it hits the boundary of the digital world:

- need something picked up, photographed, checked, delivered, returned? rent a person
- need a “hands-and-eyes” loop in a place your agent can’t access? rent a person
- need a “real-world click” where robots still can’t touch grass? rent a person

So yes - agents can now operate for longer without “you”...by directing “someone else”.

This is autonomy via outsourcing and the gig economy just got an AI upgrade. It’s a bridge technology that helps agents behave like they already have bodies, before robots scale. It’s also a subtle social reframe - the “human-in-the-loop” isn’t the user providing oversight. It’s an on-demand labour layer that makes the agent look more capable than it actually is. This also puts downward pressure on the Equality dial.

If you want the broader context - this fits neatly alongside the “Feb 2026 TrustIndex / Report” on Autonomous Assistants - where the trajectory is clear: longer-running agents don’t just need better reasoning, they need execution infrastructure. rentahuman.ai is execution infrastructure. Just...made of people.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.

It’s a new approach that aims to produce “metric (to-scale) depth from a single camera”, and can also “recover camera intrinsics” and “use sparse depth as an input”. In other words - it’s pushing toward a world where “how far away is that?” or “how big is that?” become a default capability for commodity cameras. Not just something you only get from specialised sensors or carefully controlled setups.

Why this matters for TrustIndex - this puts upward pressure on the Fidelity dial.

High-fidelity AI-Mediation isn’t just prettier overlays - it’s “correct overlays” and AI-based “spatial reasoning”. If your system can infer real-world scale and geometry reliably, then:

- AR/UI layers can sit on the world instead of floating near it
- agents can reason about reach, clearance, placement, and navigation with fewer hallucinated distances
- “digital twins” become less like vibes and more like measurements
- the hand-off from perception → action gets tighter (robots, assistants, assistive tech, tools)

This is the quiet kind of progress that makes future interfaces feel inevitable. Once you can get metric depth from “just a camera”, you can ship spatial understanding everywhere (phones, glasses, laptops, drones) without waiting for the hardware stack to catch up.

> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.