Submissions from simonwillison.net

		Prompt Injection as Role Confusion (simonwillison.net)
		3 points by Brajeshwar 1 day ago \| past \| discuss
		Porting Moebius 0.2B image inpainting model to run in browser with Claude Code (simonwillison.net)
		3 points by lumpa 1 day ago \| past \| discuss
		Datasette Apps: Host custom HTML applications inside Datasette (simonwillison.net)
		156 points by lumpa 5 days ago \| past \| 68 comments
		GLM-5.2 is probably the most powerful text-only open weights LLM (simonwillison.net)
		28 points by Brajeshwar 6 days ago \| past \| 4 comments
		GLM-5.2 is probably the most powerful text-only open weights LLM (simonwillison.net)
		9 points by lumpa 6 days ago \| past \| discuss
		The Fable 5 Export Controls Harm US Cyber Defense (simonwillison.net)
		5 points by lumpa 8 days ago \| past \| discuss
		"They screwed us": Personality clashes sent Anthropic's models offline (simonwillison.net)
		4 points by lumpa 9 days ago \| past \| discuss
		Why AI hasn't replaced software engineers, and won't (simonwillison.net)
		5 points by SVI 9 days ago \| past \| discuss
		Claude Fable is relentlessly proactive (simonwillison.net)
		774 points by lumpa 12 days ago \| past \| 656 comments
		Anthropic Walks Back Policy That Could Sabotage AI Researchers Using Claude (simonwillison.net)
		1 point by lumpa 13 days ago \| past \| discuss
		Initial impressions of Claude Fable 5 (simonwillison.net)
		9 points by gillytech 13 days ago \| past \| 2 comments
		Initial impressions of Claude Fable 5 (simonwillison.net)
		20 points by mfiguiere 14 days ago \| past \| 2 comments
		If Claude Fable stops helping you, you'll never know (simonwillison.net)
		8 points by behnamoh 14 days ago \| past \| 2 comments
		Siri AI at WWDC 2026 (simonwillison.net)
		2 points by lumpa 15 days ago \| past
		Running Python code in a sandbox with MicroPython and WASM (simonwillison.net)
		110 points by theanonymousone 18 days ago \| past \| 41 comments
		Running Python code in a sandbox with MicroPython and WASM (simonwillison.net)
		15 points by pretext 18 days ago \| past \| 2 comments
		Uber's $1,500/month AI limit is a useful signal for AI tool pricing (simonwillison.net)
		624 points by pdyc 21 days ago \| past \| 769 comments
		Running Python ASGI apps in the browser via Pyodide and a service worker (simonwillison.net)
		4 points by Brajeshwar 24 days ago \| past
		We contain Claude across products (simonwillison.net)
		3 points by Brajeshwar 24 days ago \| past
		SQLite Does Not Accept Agentic Code (simonwillison.net)
		3 points by birdculture 25 days ago \| past
		SQLite Does Not Accept Agentic Code (simonwillison.net)
		6 points by signa11 26 days ago \| past \| 1 comment
		Anthropic's self-reported run-rate revenue growth is wild (simonwillison.net)
		5 points by pretext 26 days ago \| past \| 1 comment
		Claude Opus 4.8: "a modest but tangible improvement" (simonwillison.net)
		2 points by GavinAnderegg 26 days ago \| past \| 2 comments
		I think Anthropic and OpenAI have found product-market fit (simonwillison.net)
		1094 points by simonw 28 days ago \| past \| 1245 comments
		Notes on Pope Leo XIV's Encyclical on AI (simonwillison.net)
		62 points by pretext 29 days ago \| past \| 12 comments
		Show HN: Datasette Agent (simonwillison.net)
		10 points by simonw 33 days ago \| past \| 1 comment
		Gemini 3.5 Flash: more expensive, but Google plan to use it for everything (simonwillison.net)
		2 points by flyaway123 35 days ago \| past
		Is Claude Code going to cost $100/month? Probably not–it's all confusing (simonwillison.net)
		4 points by birdculture 35 days ago \| past \| 1 comment
		The last six months in LLMs in five minutes (simonwillison.net)
		804 points by yakkomajuri 36 days ago \| past \| 587 comments
		Using LLM in the shebang line of a script (simonwillison.net)
		4 points by dnw 42 days ago \| past
		More