All The Headlines, All The Model Drops...

DeepSeek dropped V4 Pro and V4 Flash, undercutting US labs on price by roughly 5x. OpenAI fired back with GPT-5.5, reclaiming benchmark crowns from Claude. Meta confirms 8,000 layoffs on May 20, and Google plans to invest up to $40B in Anthropic.

• DeepSeek releases its new flagship models V4 Pro and V4 Flash in preview, saying V4 Pro trails the performance of state-of-the-art models by about 3 to 6 months (Bloomberg)
• Simon Willison's comparison chart of DeepSeek V4 pricing vs. US frontier models (Simon Willison)
• OpenAI says "GPT-5.5 matches GPT-5.4 per-token latency in real-world serving", performs "at a much higher level of intelligence", and is more capable for Codex (OpenAI)
• Meta plans to cut 10% of its employees, or ~8,000 jobs, on May 20 and won't fill 6,000 open roles, trying to boost efficiency and offset its heavy AI spending (Bloomberg)
• Google plans to invest up to $40B in Anthropic: $10B now at a $350B valuation, with another $30B if Anthropic hits performance targets (Bloomberg)

Weekend Longreads Suggestions

• A look at the AI nonprofit METR, maker of maybe the most important AI benchmark, whose time-horizon metrics are used by researchers and Wall Street to track AI development (NYT)
• The Infinite Machine Olto e-bike review: a more elegant solution for trips too long to walk but too short to drive (The Verge)

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

Support Brian's work on the Techmeme Ride Home directly!

For just $5/month, you can subscribe to the ad-free Premium feed.

Full details at tech.supercast.com

Thank you for your support of Brian and the Techmeme Ride Home!