#SimonWillison discusses the impact of #AI on #softwareengineering. He highlights November 2025 as a turning point when #AIcodingagents became reliable. Willison also emphasises the need for #security measures against #promptinjection and predicts the rise of “dark factories” where AI autonomously…

Merriam-Webster crowns “slop” word of the year as AI content floods Internet https://arstechni.ca #machinelearning #merriam-webster #SimonWillison #wordoftheyear #dictionaries #generativeai #AIcriticism #socialmedia #Biz&IT #aislop #Oxford #slop #spam #AI

Anthropic says its new AI model “maintained focus” for 30 hours on multistep tasks https://arstechni.ca #Computer-UsingAgent #largelanguagemodels #AIdevelopmenttools #computerusemodel #machinelearning #AIcomputeruse #SimonWillison #AIassistants #AIbenchmarks #generativeai #Programming…

This is a fantastic and horrifying demonstration of how widespread and dangerous these #AiVulnerabilities still are, almost three years after we first started talking about them.

#simonwillison

Under the heading #MonthofAI Bugs he has been publishing one report per day across an array of different tools, all of which are vulnerable to various classic #promptInjection problems.

#simonwillison

The Summer of Johann: prompt injections as far as the eye can see Independent AI researcher Johann Rehberger (previously) has had an absurdly busy August. Under the heading The Month of AI Bugs he has been publishing one report per day across an …

simonwillison.net/2025/Aug/15/...

The Summer of Johann: prompt injections as far as the eye can see
Independent AI researcher Johann Rehberger (previously) has had an absurdly busy August.

#simonwillison

New AI browser agents create risks if sites hijack them with hidden instructions https://arstechni.ca #computerusemodel #promptinjections #AIvulnerability #browsersecurity #Chromeextension #machinelearning #SimonWillison #webbrowsers #AIsecurity #agenticAI #Anthropic #AIagents #AIsafety…

New Grok AI model surprises experts by checking Elon Musk’s views before answering https://arstechni.ca #machinelearning #SimonWillison #AIassistants #JeremyHoward #AIalignment #AIbehavior #aisearch #ElonMusk #Twitter #Biz&IT #grok #xAI #AI #X

Anthropic summons the spirit of Flash games for the AI age https://arstechni.ca #largelanguagemodels #AIdevelopmenttools #AnthropicClaude #machinelearning #AIprogramming #SimonWillison #AIassistants #generativeai #Flashgames #newgrounds #vibecoding #vibecoding #agenticAI #Anthropic #AIagents…

Microsoft surprises MS-DOS fans with remake of ancient text editor that works on Linux https://arstechni.ca #softwaredevelopment #Vintagecomputing #retrocomputing #SimonWillison #commandline #Programming #texteditors #opensource #microsoft #retrotech #Windows #GitHub #MS-DOS #Ubuntu #Linux…

Hidden AI instructions reveal how Anthropic controls Claude 4 https://arstechni.ca #machinelearning #SimonWillison #Anthropic #AI

Video scraping: extracting JSON data from a 35 second screen capture for less than 1/10th of a cent The other day I found myself needing to add up some numeric values that were scattered across twelve different emails. I didn’t particularly feel like copying and pasting all of …

#VideoScraping: Extract JSON data from 35-sec screen capture for <$0.001 using #GoogleGemini. #SimonWillison demonstrates powerful technique for data extraction from protected sources. #AI #DataJournalism #Productivity
simonwillison.net/2024/Oct/17/...

#ai #Google #Gemini #llm