AIPolitics
2026 in LLMs (so far)
Publication: Simon Willison's Weblog
Author: Not available in collected metadata
Source published: 27/09/2026, 23:54:15 UTC
Briefing created: 28/09/2026, 19:36:21 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 28/09/2026, 19:27:25 UTC.
Briefings
In late 2026, the release of Claude Opus 4.5 and GPT-5.1 marked a significant shift where coding agents transitioned from unreliable to reliable daily use. This breakthrough coincided with the first commit to the Warelay repository, which later renamed to CLA. The author, Simon, adopted an ambitious approach to 2026, predicting LLMs would write good code and solve sandboxing, though these predictions were largely unfulfilled. The year was characterized by 'AI mania,' where the author built a Python JavaScript interpreter and WebAssembly runtime to combat inefficiency, resulting in a 'Deep Blue' state of mind. Despite a 40% increase in conference discussions on agent security, the predicted 'Challenger disaster' did not materialize, and the author's own AI projects were deemed unnecessary by the community.
Key points
- Claude Opus 4.5 and GPT-5.1 were released in November 2025, improving coding agents from unreliable to reliable daily use.
- The first commit to the Warelay repository occurred in November 2025, which later renamed to CLA in January 2026.
Context and limitations
Predictions regarding LLM code quality and sandboxing were largely unfulfilled. The predicted 'Challenger disaster' for agent security did not occur. The author's own AI projects were deemed unnecessary by the community.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 28/09/2026, 19:36:21 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeiamxpk3zb24hvtsfleklp5wbw7d2e7ugvblz5tjfkgtuir6ov45wy