Had to share how I stay as up to date as humanly possible in AI. I can’t just re
By Allie K. Miller · October 16, 2026 · Curated by George's Blog
Had to share how I stay as up to date as humanly possible in AI.
I can’t just read every day. I actually have to become a stronger forecaster than my clients.
My forecasting happens through 4 lenses: research, startups, zeitgeist, and regulation.
👀 If you track research: you can see what breakthroughs might spur new companies and business lines. You can also see how fast research is moving (and what is moving slower than expected, which should dial back some predictions!). Best source is arXiv.
👀 If you track startups: you can see what is getting funded or laughed out the door. You can see which startups are locking down enterprise clients. You can see which types of companies are getting backed by Tier 1 VCs. You can predict what capabilities enterprises might be able to purchase over the next 2–3 years. Best sources are Y Combinator/Hacker News, Axios Pro Rata, and The Information.
👀 If you track zeitgeist: you know what people give a shit about. None of the startup and research information really matters if there is no public trust in the product. Best sources are TikTok comments and my own inbox.
👀 If you track regulation: you know where friction might happen or where forces might push a company in a different direction. If policy makes chip production in America cheaper, how might that shift global supply chains? If data centers face new environmental standards, how might that impact compute scalability? Regulation gives me better clues about where the broader economy is headed. Best sources are Politico, Platform Economy Insights, and anyone in DC (they’ll talk your ear off).
Pick something in AI you’re trying to predict, like when you’ll have a robot that stretches your legs for you. What do these four lenses tell you about it, and what evidence would change your mind?