I have a theory that the holding of the most advanced US model is an A/B testing to see if the Chinese models' advancement come from the latest models from US. The downside is only the commercial opportunity window (the ban period).
This is at best self-delusional and quickly falsifiable in reality. Your brain will quickly question your idea of 'temporary' and will question your own judgement a while ago.
All MacBooks in local Bay Area stores are sold out, they’re all back-ordered online, and Apple is now raising prices by up to $1,000.
A sure sign everyone is running more agents?
@Jessicalessin But looking back, the researchers and engineers were the ones that could have predicted the rise of Nvidia back in 2020/2022, not the business people.
@fahdananta Yes. But for the ones that don't practice what they preach (case in point here), you would want to step back and think about their incentives.
@quxiaoyin Build more lock-in features such as shared memories, collaborative applications, faster and more RAM efficient applications, fear mongering about security of substitutes from other countries.
Blocks is such a prototypical low bar AWS move. You have a great customer/partner in Supabase, doing incredibly well. Clearly there's a niche. So you come in and you copy them almost directly. Same thing as Kiro vs. Cursor where Cursor was a run away hit. Cheap/easy to 2-pizza team out. Requires minimal PM brainwork. Immediately makes startups question working with you.
@cwarzel Looking back how small when these other companies were at their IPO, a small seed can grow into a big tree, and now we are seeing a giant seed. This economic book might help https://t.co/bwYKquqYbm