Hi Reader,

Here are the most important AI stories I’ve found this month:

• Microsoft’s AI solves medical mysteries
• Cloudflare charges AI crawlers
• Books are legal for AI training
• AI model teaches itself
• AI researchers are paid like superstars

Details below! 👇

🩺 Microsoft’s AI solves medical mysteries

Microsoft’s new “AI Diagnostic Orchestrator” solved 85% of 304 real medical mysteries, whereas experienced doctors (who did not have access to colleagues or medical databases) only solved an average of 20%. Microsoft claims they are on “the path to medical superintelligence.”

My take: Although this is incredibly impressive, the more important test is how it performs on cases from the ordinary population (in which the vast majority don't have a complex or rare diagnosis) and in real-world settings (in which doctors can consult with human experts and use Google). See here for a longer analysis.

🕸️ Cloudflare charges AI crawlers

AI companies have built billion dollar businesses by training their models on free Internet content. This threatens content publishers, who have historically depended on website visits in order to survive. In response, Cloudflare (which manages 20% of all web traffic) launched pay per crawl, which allows publishers to charge AI crawlers for access.

My take: I’m hopeful this will prove to be a huge win for both large and small publishers!

⚖️ Books are legal for AI training

In a major legal ruling, a US judge declared that training AI models on books without author permission is “fair use”, but the books must be legally purchased. Anthropic (the makers of Claude) had been purchasing and scanning some physical books, but pirating others.

My take: I’m glad that LLMs have been legally vindicated as a “transformative” (rather than “reproductive”) technology, but it’s hard for me to imagine that the big AI companies will stop training on illegally obtained content.

🎓 AI model teaches itself

MIT researchers have built an AI model that can continually teach itself by generating new insights and updating its weights automatically.

My take: This is perhaps a step in the direction of “Artificial General Intelligence” (AGI) that some experts predict we will reach within 5 to 10 years!

💰 AI researchers are paid like superstars

Meta has been on an AI hiring spree, reportedly paying $100M+ for top people to switch from other leading AI companies.

My take: With AI researchers being paid like sports superstars, will they be the next idols of our youth?

I hope you enjoyed reading this as much as I enjoyed writing it!

Have a great weekend!

- Kevin