Submissions from simonwillison.net

		ChatGPT is three years old today (simonwillison.net)
		6 points by ingve 36 days ago \| past \| 2 comments
		ChatGPT prompt consumes equivalent to 10s of Netflix (simonwillison.net)
		8 points by makeavish 38 days ago \| past \| 4 comments
		Claude Opus 4.5, and why evaluating new LLMs is increasingly difficult (simonwillison.net)
		6 points by jonesn11 40 days ago \| past \| 1 comment
		Building Effective Agents (simonwillison.net)
		1 point by gearnode 41 days ago \| past
		Claude Opus 4.5, and why evaluating new LLMs is increasingly difficult (simonwillison.net)
		1 point by gingersnap 42 days ago \| past \| 1 comment
		Claude Opus 4.5, and why evaluating new LLMs is increasingly difficult (simonwillison.net)
		7 points by janpio 42 days ago \| past \| 2 comments
		Olmo 3 is a fully open LLM (simonwillison.net)
		5 points by lumpa 44 days ago \| past \| 2 comments
		LLM cmd, an LLM plugin to prompt and edit a shell command (simonwillison.net)
		2 points by skeledrew 45 days ago \| past
		The fate of "small" open source (simonwillison.net)
		2 points by synergy20 45 days ago \| past
		Nano Banana Pro is the best available image generation model (simonwillison.net)
		3 points by xnx 46 days ago \| past
		Google Antigravity (simonwillison.net)
		2 points by bilsbie 47 days ago \| past
		Google Antigravity (simonwillison.net)
		4 points by vismit2000 48 days ago \| past
		What happens if AI labs train for pelicans riding bicycles? (simonwillison.net)
		5 points by Jap2-0 48 days ago \| past
		Trying out Gemini 3 Pro with audio transcription and a new pelican benchmark (simonwillison.net)
		181 points by nabla9 48 days ago \| past \| 58 comments
		Netflix issues GenAI rules:no generative AI in final deliverables (simonwillison.net)
		3 points by isaacfrond 56 days ago \| past
		Reverse engineering Codex CLI to get GPT-5-Codex-Mini to draw me a pelican (simonwillison.net)
		168 points by simonw 58 days ago \| past \| 76 comments
		Using Codex CLI with GPT-OSS:120B on an Nvidia DGX Spark via Tailscale (simonwillison.net)
		4 points by kgwgk 59 days ago \| past
		Could LLMs encourage new programming languages? (simonwillison.net)
		3 points by coloneltcb 59 days ago \| past
		Using Codex CLI with GPT-OSS:120B on an Nvidia DGX Spark via Tailscale (simonwillison.net)
		5 points by GavinAnderegg 60 days ago \| past
		Code research projects with async coding agents like Claude Code and Codex (simonwillison.net)
		3 points by simonw 60 days ago \| past
		I’m worried that they put co-pilot in Excel (simonwillison.net)
		494 points by isaacfrond 62 days ago \| past \| 334 comments
		New prompt injection papers: Agents rule of two and the attacker moves second (simonwillison.net)
		114 points by simonw 64 days ago \| past \| 44 comments
		Thoughts on Cursor 2.0 and Cursor Compose (simonwillison.net)
		7 points by azhenley 68 days ago \| past \| 1 comment
		Hacking the WiFi-enabled color screen GitHub Universe conference badge (simonwillison.net)
		31 points by meetpateltech 69 days ago \| past \| 4 comments
		The PSF has withdrawn a $1.5M proposal to US Government grant program (simonwillison.net)
		239 points by lumpa 70 days ago \| past \| 9 comments
		Setting up a codebase for working with coding agents (simonwillison.net)
		3 points by FromTheArchives 72 days ago \| past
		Dane Stuckey (OpenAI CISO) on Prompt Injection Risks for ChatGPT Atlas (simonwillison.net)
		5 points by coloneltcb 75 days ago \| past \| 1 comment
		Living Dangerously with Claude (simonwillison.net)
		250 points by FromTheArchives 76 days ago \| past \| 108 comments
		ChatGPT Atlas (simonwillison.net)
		4 points by gk1 76 days ago \| past \| 1 comment
		Claude Code for Web — a new asynchronous coding agent from Anthropic (simonwillison.net)
		2 points by codeclimber 76 days ago \| past \| 1 comment
		More