I work in financial market regulation. My job: let innovation grow while keeping its risks in check, before they hit. With AI, the questions are being asked. But the race is moving faster than the answers. I use it every day. It saves me hours, and it may have just cracked math problems people chased for 50 years. But the US-China race feels a lot like the US-Soviet race for the bomb: each side pushes ahead out of fear the other gets there first and worries about the risks later. In finance, we learned the hard way: the danger isn't one bad bank, it's everyone making the same bet. Now everyone runs on the same few AI models. Here, I try to make sense of the big shifts AI is driving in finance, geopolitics and tech, and break them down simply, for you and for me. Come argue with me. What’s the one AI risk you think is being underestimated right now?

Hey @grok, what are the 10 most likely apocalyptic scenarios inspired by TV shows? For each one, give the TV show, a one-sentence description of the scenario, the flag emoji of the country where it could happen, and its probability as a two-digit percentage. Rank them from most to least likely.

AI doesn’t equalize intelligence. It amplifies it. The more you’re able to challenge its answers, verify its sources, and sharpen your ideas, the more Grok becomes a formidable weapon. The more you swallow everything it says without pushback, the more it reinforces your mediocrity: generic style, hollow arguments, zero verification. AI is a multiplier.

🚀SpaceX went public in June at a $1.8 trillion valuation. Its stock jumped to $225 in the first week, then fell 53% by August as the excitement faded and insiders got closer to selling. Anthropic is asking for a similar valuation, based on growth its own filing says "may not be sustainable." What happens when that excitement fades too?

Hey @grok, my self-driving Tesla has to choose in a split second: hit 3 elderly women or a baby to avoid killing me. Who decides that choice: me, Tesla, or the government? And who pays when it goes wrong? One clear sentence each.

🇧🇷 Brazil re-elected Lula yesterday by 4.4 million votes. Today investors ran: stocks down ~14%, trading halted at the open, the currency down ~10%, and borrowing costs spiked overnight. And the loser, Flávio Bolsonaro, still refuses to concede. Why it matters: US midterms are in 8 days, and markets now judge an election within hours. If your 401(k) dropped 14% the morning after an election, would you still accept the result?

My AI knows me better than most of my coworkers. Grok Bot fact-checks me before I make a call, tells me when my idea is weak, and remembers every correction I give it. But only because I told it to. One prompt telling it to stay neutral and criticize me. Without it, would it have just agreed with my mistakes? How many people are getting yes-men instead of advisors?