Anthropic or xAI IPO in 2026? AGIX ETF Offers Exposure
Anthropic and xAI may be edging closer to the public markets. Click to read how IPO speculation is building and how investors can access both these private AI leaders today through KraneShares AGIX ETF.
When the cost of intelligence goes to zero
A thought exercise…
Think of a subject you know nothing about.
I’ll go first.
New England whaling in the 1800’s. I wouldn’t even know where to begin. Something about lamp oil and Herman Melville? I could probably bullshit you for five minutes. I don’t think I could do six.
But now I can type “You are a professor of New England history at a prestigious college. Prepare me for an appearance on Jeopardy where “Whaling” will be a category in one of the rounds. I’ll need enough superficial knowledge on the topic as I can possibly retain without too much so that I forget the important stuff. Make me appear to be knowledgeable on the subject within five minutes.”
ChatGPT will absolutely get this done. It will even offer a quiz after you’ve read the material to lock it all in. And if you go through this exercise, you can appear credible in a 20 minute conversation on the subject with someone who knows less than you about whaling. Which is everyone. Behold! Intelligence on demand.
But that’s not intel. Intelligence is now free. Intel is different. It’s the stuff you have to hunt down and assemble yourself. It’s the stuff insiders know that will not be readily apparent or accessible to the LLMs. In some cases, it’s the secret layer of knowledge that might actually be worth something in the context of trading or investing. Because you don’t get paid in the stock market for the stuff everyone already knows or can calculate. By definition, it is valuable because you can take action with it before others can figure it out if you work hard enough. LLMs are a shortcut to intelligence. They’re not going to provide you with the thing we think of as “intel.”
Did I ever tell you the story of Max Planck’s chauffeur?
Charlie Munger used to tell this story to distinguish the difference between learning and mimicry. It’s pertinent now because LLMs are pattern-matching word calculators that have perfected the art of mimicry.
Anyway, Max Planck was a Nobel Prize-winning physicist. He did a lecture tour of Germany and, at every stop, his driver would sit in the room as he expounded on his scientific ideas.
One day, after weeks of monotony on the road, the driver said “I have an idea. How about, today, I do the talk and you sit in the audience.” The physicist loved the novelty of the idea and agreed to switch places. His chauffeur stood in front of the lecture hall and delivered the Max Planck talk he had heard so many times. The audience dutifully applauded.
And then someone stood up to ask a technical follow-up question. The driver didn’t miss a beat. With a haughty look, he waved his hand and said "I'm surprised that in an advanced city like Munich I get such an elementary question. I'm going to ask my chauffeur to reply.”
What’s not going to change despite the widespread intelligence running like hot and cold water through ChatGPT and Claude is that there are still going to be experts. What will change is that we will also have legions of people with “chauffeur knowledge” - which is fine but is not a replacement for expertise. It’ll get you through a round of Jeopardy, but will anyone be paid for it? Why would you pay for what’s already free?
Sam Koppelman and Nathaniel Horwitz on The Compound and Friends
Michael and I with the Hunterbrook Media boys in NYC this week
I was reminded of this on Thursday when my friend Sam Koppelman and his business partner Nathaniel Horwitz came by The Compound to tape the podcast. Sam and Nathaniel founded Hunterbrook Media which is a hedge fund attached to a working, well-funded newsroom. They’ve hired some of the best investigative journalists they could find - including the legendary Bethany Frankel, whose landmark work unraveling the Enron fraud became one of the most consequential series of articles in Wall Street history. Hunterbrook is going to hunt for intel, manufacture the story based on the facts their reporters uncover and then trade on it. By definition, they will have something more valuable than the sort of chauffeur knowledge anyone else can order up from a chatbot. They’ll be using their own intel.
Intelligence is a commodity but intel is a valuable good.
Last year the hedge fund outperformed the S&P 500 with a long-biased book of stock trades and a few notable investigations that became national news. It’s an incredible story that Michael Batnick and I were excited to bring to you. If you haven’t heard it yet, you can listen or watch at the links below!
Inside the Baddest Hedge Fund in America
THE COMPOUND & FRIENDS
Inside the Baddest Hedge Fund in America
Michael Batnick and Downtown Josh Brown are joined by Nathaniel Horwitz and Sam Koppelman of Hunterbrook to discuss: how investigative journalism can be an edge in investing, the Sable Offshore story, how Sphere is changing entertainment, and much more!
