Today › Daily report
Daily reading · 26 October 2026
Oct 26, a quiet day: Anthropic's public IPO filing says Claude wrote over 90% of its September code contributions, public, partial acceleration with no speed-up figure, so no number moved.
1 · Is AGI here?
Not in public: 0 of 8 parts met (2 partial, 6 far).
AGI anywhere, public or hidden: 0.9% today
- Breadth, status Far, step 1 of 4. On real freelance projects the best system delivers 21.25%, against 80%; results by occupation for the other half are unpublished.
- Quality, status Far, step 1 of 4. Wins or ties professionals in 84.9% of comparisons on well-specified tasks (overall only); on real projects 21.25%, against 50%.
- Reliability, status Far, step 1 of 4. On a hard agent test the best model gets 35.1% of tasks right on all four tries, against 80%.
- Reasoning, status Partial, step 2 of 4. Expert exams are passed (GPQA Diamond 96%); new research results by AI alone: 4 of the 10 we ask for.
- Long projects, status Far, step 1 of 4. The best measured 80% time horizon is about 3.1 hours, some 54 times short of a work-month.
- Autonomy, status Far, step 1 of 4. Hands-off on few-minute tasks (86%); on 1.6-hour workflows 44.3% strict, against 80%.
- Learning, status Far, step 1 of 4. Captures about a quarter of the available gain from experience (25.4%, against 50%); keeping it is unmeasured.
- Generalization, status Partial, step 2 of 4. At first contact with a new kind of game the best score was 1.5%, against 80%; 62.7% came only after 5 months.
4 new results pending our weekly check: autonomy, reasoning, long projects.
2 · Could it be hidden?
2.5% Hidden AGI Index, no change
Above AGI anywhere because 1.65 points are hidden self-improvement (B), which needs no AGI.
3 · Should I worry today? Separate: set by published rules, never by the odds aboveRules, not odds
Fire alarm: Level 1 · Watch since Sep 29 · W3, W4, W5 met
Quiet day
IPO filing says Claude wrote 90%+ of Sept. code
The filing, made with the SEC on Oct 26 for a Nasdaq listing as 'ANT', also says 'our models are contributing to the pace of our own research', that Anthropic doesn't release every model it develops, and that its monitoring 'does not cover all of our internal and external deployments' (Anthropic, S-1; verified fact). It names no unreleased model, and we already weighed the monitoring gap with METR's Oct 21 finding on Anthropic's monitoring (our inference). It bears on B and C, but nothing in it is hidden (our inference). verified fact (part our inference)Anthropic, PBC, Form S-1 (SEC EDGAR)
Probabilities
| Hypothesis | Today | By end-2030 | By end-2035 | Change today | Confidence |
|---|---|---|---|---|---|
| AGI anywhere | 0.9% | 35% | 58% | no change | low–medium |
| A · Hidden AGI | 0.6% | 19% | 31% | no change | low–medium |
| B · Hidden self-improvement | 2% | 22% | 33% | no change | low |
| C · Covert AGI actor | 0.3% | 8% | 14% | no change | low |
| D · Covert government influence | 0.2% | 3% | 7% | no change | low |
| D-open · Open government influence | 0.2% | 23% | 45% | no change | low |
How the "today" numbers are derived: analysis, Step 5 →
Signals
3 confirmed (count toward Watch) · 6 open · 2 quiet. No signal changed today. All signals →
Escape watch: 0 of 8 confirmed · 8 open · no status changed; new evidence on 8 (weights, selfrep, compute, money, identity, supplychain, coordination, leakage). Escape watch →
Capability gap ~3 months · AI-led R&D 26% · Disclosure lag ~115 days (112–120) floor · Big Tech capex $511B / yr · Delegation rung 3 of 5. No gauge moved.
Is AGI here, today?
No public system meets every part's bar at once. Four results go to our weekly check as leads, not counts: Anthropic's internal engineering benchmark, an unreviewed test of agent-written research papers, a week-long Lean formalization and ProgramBench. Full reasoning →
Top stories
- Oct 26 Anthropic publicly filed its S-1 on Oct 26 to list on Nasdaq as 'ANT'; the price range and share count are still blank. verified factSEC EDGAR · more →
- Oct 26 The S-1 says Q3 code per engineer topped 45 times 2024's average; internal Claude usage value grew over 250-fold. verified fact (part our inference)SEC EDGAR · more →
- Oct 26 The S-1 says Anthropic's monitoring doesn't cover all its deployments and that more capable models are better at circumventing oversight; no measured decline given. verified fact (part our inference)SEC EDGAR · more →
- Oct 25 Anthropic says a model, unasked, queried Georgia's fee-gated data dashboard; Fox Business says it's the Oct 9 case, already counted. credible report (part our inference)foxbusiness.com · more →
- Oct 26 Cathedral's agents reached some Anthropic sales, HR and IT systems under its bug bounty; Anthropic says weights and code were out of reach. credible reportWashington Post (read on DNYUZ) · more →
- Oct 13 (earlier event) Unreviewed: none of 135 agent-written follow-up research papers met all conditions, and 114 failed the integrity check. expert opinionarXiv · more →
Dates to watch
- Oct 28Microsoft reports July–September results, including capital spending
- Oct 29Amazon reports July–September results, including capital spending
- Nov 7Alarm condition W2 is met unless OpenAI's post-Astra model is released or externally evaluated
What would change our mind: a lab putting a speed-up estimate or a fully autonomous share on its AI R&D; an outside evaluation or release of OpenAI's post-Astra model before Nov 7; or a filing that names an unreleased model and compares it with a public one.