EVIDENCE
Published signals, the current ASI forecast and the review archive are kept together so every headline can be traced back to the underlying record.
SIGNALS
Z.ai says its upcoming open-weight GLM-5.3 reached 84.5% on CyberGym and will be released publicly in about two weeks after additional security work; the benchmark claims are not yet independently verified.
SOURCE ↗Google released Gemini 3.7 Flash on 13 August 2026 for software coding, multi-step agentic workflows and automated business tasks. Independent reporting describes meaningful gains in planning, instruction following and code generation, with lower pricing than the preceding Flash release.
SOURCE ↗DeepSeek released the GA version of V4 Pro on 13 August 2026 across app, web and API. DeepSeek reports substantial agent gains, including Terminal Bench 2.1 at 87.9, HLE at 42.7 without tools and 60.0 with tools, Toolathlon-Verified at 74.1 and DSBench-FullStack at 71.1. The API also adds native Responses API support and low/high/max thinking effort. This is a material frontier-model release, but independent evaluation is still needed before treating it as a new global capability ceiling.
SOURCE ↗SpaceXAI released Grok 4.6 on 12 August 2026. Independent reporting citing Artificial Analysis says the model improves roughly five points over Grok 4.5 on the Artificial Analysis Intelligence Index, lands around GPT-5.6 Sol overall, and is especially strong on agentic work while remaining materially cheaper than the top Anthropic models. This strengthens SpaceXAI's frontier position, but it does not currently establish a new absolute capability ceiling.
SOURCE ↗A Claude-assisted research workflow reportedly improved the proportion of non-trivial simple zeros shown to lie on the critical line from about 41.6% to 67.2%. This was a substantial partial result, not a proof of the Riemann Hypothesis. The work used roughly 31M output tokens, about 60 sub-agents, around 650 discarded strategies, thousands of computational checks and Lean formalization.
An analyst estimate cited in the Pulse put AgiBot at roughly 8,400 humanoids shipped in H1 2026, around 44% of global shipments. This is meaningful evidence of manufacturing scale, but shipments are not evidence of general-purpose autonomy: public uptime, intervention rates, teleoperation share, generalized-task performance and continuous production telemetry were missing.
A candidate result was reported in which Claude Fable 5 proposed an algorithm and GPT-5.6 repaired and simplified the argument. Review was still pending, so it was treated as an AI-for-Science signal rather than a confirmed breakthrough.
SOURCE ↗VibeMathed/arXiv recorded an AI-assisted result in which GPT-5.6 Sol/Codex derived proofs and counterexamples and Claude simplified or checked parts of the argument. It remained unreviewed and did not change the ASI trajectory.
SOURCE ↗VibeMathed recorded a proposed counterexample developed with GPT-5.6 Sol Ultra. It was marked resolved by the tracker but remained unreviewed, with no Lean proof, peer review or specialist independent verification recorded in the Pulse.
SOURCE ↗GPT-5.6 Sol Ultra reportedly carried a Dittert-5 candidate through problem selection, discovery, certificate generation, Lean formalization, verification and manuscript preparation. The workflow was unusually autonomous, but independent human audit and review were still missing.
SOURCE ↗The 7 August Pulse recorded WeatherNext Cyclones as roughly one day of additional warning and about a decade of meteorological progress, without a documented ASI forecast revision.
The 6 August Pulse included a rank-3 K-U-M result recorded as Lean-verified. It was notable but did not justify an ASI forecast change.
ASI FORECAST
No change
No change
No change
No change
No change
REVIEW HISTORY