Today › Daily report
Daily reading · 31 October 2026
Oct 31: the UK Lords voted down a frontier-developer incident-reporting duty on Oct 26, keeping a status quo we already assume; no published probability moves.
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 On real freelance projects the best system delivers 21.25%, against 80%; results by occupation for the other half are unpublished.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.
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
Oct 26: UK Lords rejected an AI lab incident duty
Amendment 41 would have made frontier developers report, within 24 hours, any case in which one of their models gained, or was used to gain, unauthorised access to, or interfered with, a system a regulated organisation relies on. The minister said the bill puts the reporting onus on regulated entities instead, and that its regime 'is not designed to provide a comprehensive framework for AI governance, including frontier AI capabilities which have not been released to customers' (Bills API; Hansard; verified fact; background). verified factUK Parliament: Lords votes API, Bills API and Hansard (Baroness Lloyd of Effra for the government)
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) · 7 open · 1 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 $615B / yr · Delegation rung 3 of 5. Big Tech capex: $511B / yr → $615B / yr. Adopted from the Oct 30 weekly: all four Q3 filings in, $615B over the last four quarters (SEC filings).
Is AGI here, today?
No public system meets every part's bar at once. No part's test has a new score today, so nothing is pending our weekly check; the ARC-AGI-3 and OSWorld leaderboards went unread. Full reasoning →
Top stories
- Oct 26 (earlier event) Lords reject, 165–95, Baroness Harding's Amendment 41, which would have made frontier AI developers report incidents involving their models. verified factlordsvotes-api.parliament.uk · more →
- Oct 26 (earlier event) Lords defeat the government 202–155 on a cyber-regime review, due within 12 months of the Act, that must consider AISI audit powers. verified fact (part our inference)bills-api.parliament.uk · more →
- Oct 30 OpenAI's GPT-Rosalind-6 is GPT-6 Astra's weights behind trusted access, not the unreleased post-Astra model, so no clock stops. verified fact (part our inference)OpenAI · more →
- October Anthropic: instructed Claude agents found training data that slips past a production-like monitor; an instructed stress test, not deployed behaviour. verified factAnthropic · more →
- October Chainalysis disclosed an October intrusion using 'offensive agentic techniques' by an unnamed actor; no model or operator named. verified fact (part our inference)chainalysis.com · more →
- Oct 29 20 House Democrats ask OpenAI to keep a training pause that OpenAI's Oct 23 post says ended Oct 1; answers due Nov 13. verified fact (part our inference)casar.house.gov · more →
Dates to watch
- Nov 3US midterms
- Nov 4Open weights due for Moonshot's Kimi K3.1 Pro
- Nov 7Alarm condition W2 is met unless OpenAI's post-Astra model is released or externally evaluated
What would change our mind: an outside evaluation or release of OpenAI's post-Astra model before Nov 7, a UK or US rule that makes labs disclose unreleased capability, or a lab measuring a fully autonomous AI R&D share above zero.