AWS Opens 38 Med AI Agent Skills; Dating Pig-Butchering—AI Daily Brief
Audio in Mandarin Chinese · English transcript below
⚡ AI race shifts to Agent+compute+systems: AWS 38,AMD MLPerf,NVIDIA,xAI,macOS 27
The center of gravity in AI competition is shifting from model parameters to executable, auditable deployment at scale. AWS is using open-source skills to fill the decision-making and validation gap for healthcare Agent, while AMD and NVIDIA keep doubling down on inference compute—showing that the industry now treats Agent as a production system rather than a demo. Notably, the same autonomous capabilities are also being abused at industrial scale by dating scams: when AI personas can impersonate real people en masse, and platform governance cannot keep pace, the stronger the capabilities, the higher the cost of trust.
Today's Top 3 Headlines
- AI Industry News
🤖 AWS open-sources 38 medical AI Agent skills, inference win rate up to 86%
AWS: 38 open-source AI Agent skills for 11 HCLS fields; 70–86% win vs skill-less Agents, top: critical thinking. Devs: cheap, auditable, portable frameworks in Bedrock/Claude Code
Source ↗ - AI Industry News
🤖 28 dating app scams exposed: 100k API calls/day, 75% of matches are fake AI chats
Anthropic: 28 dating apps ran scam ring; one account hit >100k Claude API calls/day, 25% real gig workers. AI vendors need API abuse monitoring; app stores need review upgrades.
Source ↗ - AI Industry News
🤖 AMD's 512x MI355X GPU cluster tops MLPerf 6.1; NVIDIA launches Vera Rubin
MLPerf v6: AMD led a 512-GPU MI355X cluster; NVIDIA launched Vera Rubin. For AI infra buyers, inference compute competition heats up, reshaping cost and supply choices faster.
Source ↗
+3 more headlines
- 🤖 xAI adds cross-session memory to Grok Build, auto-records project context
- 🤖 macOS 27 Golden Gate launches: Apple Intelligence needs M3+12GB, first to drop Intel
- 🤖 Wall Street Journal: 24 New Rules for Better Work in the AI Era
Podcast highlight
If you’re only gonna walk away with one line, I’d make it this: a 100x boost in human productivity—it’s closer to your wallet than any model parameter.