Operation Spiderweb (Ukraine’s attack inside Russia, using mass drones) has world-changing implications for AI in defense and deep tech investing "There are decades where nothing happens, and there are weeks where decades happen." Decades just happened. weightythoughts.com/p/ai-is-overtu…

Coming Up at Tech Weekend – June 2025 Fireside Chat: Inside the VC Playbook — Investor Insights for 2025 Join us for an unfiltered conversation with two top investors as they share what they’re looking for in 2025, the trends they're betting on, and how founders can stand out

Not all data is equal in AI—and especially not in self-driving cars at this stage. What is needed to push them to the next level? Read, watch, or listen to my interview with Brandon Basso, Senior Director of Simulation at Cruise. weightythoughts.com/p/from-simulat…

Making lemons into lemonade…

My quick take on GPT-5 as well as some macro thoughts on how AI might change society (or not). weightythoughts.com/p/three-macro-…

Trade-offs exist, and it's good to talk about them. weightythoughts.com/p/tradeoffs-ex…

I guess I'm an author now.

Incredible. "What You Need to Know About AI" isn't even out yet, and we're already #6 on Amazon... ... in Web Development. I'm not sure how that works, but I'll take it. 😅

Can Korea compete in the global AI race — without the hype? 🧠 James Wang What You Need to Know About AI is the book every founder, investor & policymaker needs. ✅ Grounded. Clear. Globally endorsed. 📘 Out Oct 15 #KStartup #AI #DeepTech #AIStartups koreatechdesk.com/what-you-need-…

Yes, the rare earth/critical minerals export licenses from China are going to be a big deal. That being said, the harder one squeezes, the more market control slips through your fingers—let's look at this one and historical examples. weightythoughts.com/p/shortages-cr…

People keep saying wild and crazy things about how much power/water/resources AI will take. Part of that is simply because people are wrong or misinformed about what's actually happening in AI models. weightythoughts.com/p/models-are-n…

Conventional wisdom seems to be that we’re going to be overwhelmed by AI power demands. That wisdom is wrong—or at least hard to defend when not blindly accepted. Guest post from Josh Blanchfield at Avos weightythoughts.com/p/no-ai-power-…

OUT NOW opmpod with James Wang on all things happening in VC right now. -The RIAification of VC firms -The dry powder bubble -VC performance metrics you shouldn't trust -and more... YouTube bit.ly/3M8otns Apple bit.ly/3XPpJhV Spotify bit.ly/4adIhzJ

We’re starting to reach the “boring” but more important phase of AI—quieter but more productive uses in applications. It’s unlikely we’ll see AGI or Superintelligence anytime soon. It’s more likely we’ll see AI optimized to be more directly useful. weightythoughts.com/p/the-boring-p…

As we hit 2026 (Happy New Year!), we have reached 10 years into this cycle of AI. 2015 was when computer vision hit human levels of performance. Since then, we can see the rapid shifts in the landscape when we step back. weightythoughts.com/p/ten-years-of…

Economic stats may finally be starting to reflect the impact of AI—especially as it gets more “agentic.” Still early, but fascinating and worth watching. weightythoughts.com/p/ai-gains-sta…

Great discussion on AI’s impact on jobs and careers on Business Books & Co. podcast. Thanks David Kopec @hudaksbasement David Short for having me! pnc.st/s/business-boo…

TSMC’s supply constraints may break Nvidia’s moat because of inference. Curious what that means? Read more here! weightythoughts.com/p/nvidias-moat…