Anthropic Study Shows LLMs Can Be Poisoned Easily

China is stirring the pot on the tech trade wars once again. Sora has grown faster than even ChatGPT did. Is OpenAI now better at vibe coding than Anthropic is? How just a handful of malicious documents and poison an LLM. And in the longreads, the one game that has the fate of EA on its shoulders.

• China blacklists major chip research firm TechInsights following report on Huawei (CNBC)
• OpenAI’s Sora hit 1 million downloads in less than five days (CNBC)
• OpenAI Is Catching Up To Anthropic in AI Coding (The Information)
• Meta Tells Its Metaverse Workers to Use AI to ‘Go 5X Faster’ (Wired)
• It's trivially easy to poison LLMs into spitting out gibberish, says Anthropic (The Register)
• Kalshi, a Prediction Market, Raises Funds and Expands Overseas (NYTimes)

Weekend Longreads Suggestions:

• Battlefield 6 is a pivotal moment for the series — and EA (The Verge)

Learn more about your ad choices. Visit megaphone.fm/adchoices