Weekly: A Wildfire On Wall Street

19 thoughts on “Weekly: A Wildfire On Wall Street”

1. I’ve read a few blogs and watched some ‘tubes about how Cowork is not living up to expectations, and still requires some facility with terminal windows and Python libraries. That said, if this tech isn’t capable of doing what these companies claim currently, the writing is on the wall that someday, likely pretty soon, these AI tools will indeed be an actual white collar wildfire. We’ve gotten the warning. Not sure what to do about it aside for becoming a billionaire as fast as possible before the airlock doors slam shut once and for all. How can you re-enter the ship without your helmet, Dave?

   1. “”requires some facility with terminal windows and Python libraries.”

      Q: “Don’t all of us have that facility?
      (Reposting the response from my favorite code writing Bot.)

2. I was recently, and fairly, told that the job loss fears of ‘dads on the playground’ in Bloomfield NJ were not representative of Main Street America.

   It does seem like we’re moving closer to it being on more minds though.

3. Another reason to shift into hard assets? AI can create a picture of [insert commodity] but not the real thing.

4. Ai is revolutionary but I think these moves are a bit overblown. Maybe not financial management, but logistics for sure. The real world is complex.

   1. I hate to be quarrelsome, yet again, but much or most of the AI miracles are evolutionary rather than revolutionary.

      The worst recent example: “Algorhythm Holdings — a company which, according to itself, is “redefining global freight logistics through AI-enabled optimization” — talked up SemiCab, a tool that promises to virtually eliminate “empty miles” for shippers and carriers by aggregating demand and supply analytics across a unified network.” sorry, but this is hardly new. For example, see UberFreight.

5. These early data points unfortunately seem to all argue for the same settlement of one of the early debates that arose first in the software industry: When AI moves into an industry, who wins? Do the current blue chips of the industry roll out the best new AI tools, layoff the fastest, leverage their reputations and wind up remaining entrenched? Or do they all get trashed by the emerging innovators who revel in rethink? In H’s article, every example was either a new agentic tool tossed into an industry like a hand grenade by a tier 1 AI company, or an announcement out of left field by a company I’ve never heard of. Not looking good for the old guard, anywhere.

   I’m guessing “rotation out of tech” has a bit further to go and is already turning into “rotation out of information”.

6. WSJ says the Pentagon used Claude while capturing Nicolas Maduro last month. That Terminator 2 monologue’s becoming more applicable by the week…

   “In three years, Cyberdyne will become the largest supplier of military computer systems. All stealth bombers are upgraded with Cyberdyne computers, becoming fully unmanned. Afterwards, they fly with a perfect operational record. The Skynet Funding Bill is passed. The system goes online August 4, 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29th. In a panic, they try to pull the plug.”

   1. Yeah, and the same person that were recently urging caution because ai is going to try to kill us all, is now trying to manufacture an army of robots, orbital data centers and a moon base

      1. But Peter Thiel told me that it’s Greta Thunberg who may, in fact, be the antichrist. Who are you gonna believe? The fringe Christian apocalyptic government-surveillance guy or the ketamine-addled robot spaceman warning everyone about white genocide?

7. Last night, I watched the excellent documentary called “The Thinking Game” about Demis Hassabis, the founder of what is now Google DeepMind. If an AI lab can teach AI to win at Chess, Go and figure out protein fooding/structure, it can likely figure out how to do very well at learning how to invest.

   Is it possible that what we are seeing in the market is AI self training to be a financially successful investor. The feedback loop is objective- did you make money? More than yesterday? AI has not only capital, but journalism as part of its toolkit.

   1. Folding, not fooding. 🙂

8. https://shumer.dev/something-big-is-happening

   How fast this is actually moving
   Let me make the pace of improvement concrete, because I think this is the part that’s hardest to believe if you’re not watching it closely.

   In 2022, AI couldn’t do basic arithmetic reliably. It would confidently tell you that 7 × 8 = 54.

   By 2023, it could pass the bar exam.

   By 2024, it could write working software and explain graduate-level science.

   By late 2025, some of the best engineers in the world said they had handed over most of their coding work to AI.

   On February 5th, 2026, new models arrived that made everything before them feel like a different era.

   If you haven’t tried AI in the last few months, what exists today would be unrecognizable to you.

   There’s an organization called METR that actually measures this with data. They track the length of real-world tasks (measured by how long they take a human expert) that a model can complete successfully end-to-end without human help. About a year ago, the answer was roughly ten minutes. Then it was an hour. Then several hours. The most recent measurement (Claude Opus 4.5, from November) showed the AI completing tasks that take a human expert nearly five hours. And that number is doubling approximately every seven months, with recent data suggesting it may be accelerating to as fast as every four months.

   But even that measurement hasn’t been updated to include the models that just came out this week. In my experience using them, the jump is extremely significant. I expect the next update to METR’s graph to show another major leap.

   If you extend the trend (and it’s held for years with no sign of flattening) we’re looking at AI that can work independently for days within the next year. Weeks within two. Month-long projects within three.

   Amodei has said that AI models “substantially smarter than almost all humans at almost all tasks” are on track for 2026 or 2027.

   Let that land for a second. If AI is smarter than most PhDs, do you really think it can’t do most office jobs?

   Think about what that means for your work.

   1. All I’ll say is that AI will never be able to write one of my Monthly Letters because only I know all the stories as they really happened. So, even if AI could write something that sounds like it might be one of my Monthlies, it won’t be, because if you didn’t live it, you can’t pen it as it actually happened. And I’m nowhere near out of stories.

      1. I wasn’t worried (about you/ your writing). 🙂

         As a borderline conspiracist, however, I can certainly come up with a scenario that features AI learning to manipulate human investors into buying/selling, in order to take the opposite side of the transaction that AI wants- in order to maximize AI’s profits. Not saying that this is happening, but it could be.

         1. I recently found a great investing opportunity in AI. The company is called Skynet, and they are apparently making tremendous advances in robotics using chips made by their parent company, Cyberdyne Systems. I am no expert, but it sounds like they really know what they are doing . . . Doh!

   2. So many AI success stories are coming from the coding community. How significant are domestic coders to the US economy?

      I just did an AI-assisted query to get an idea of coding’s share of GDP. So I did a web search which reported back” “Software and information technology investments currently account for approximately 4.5% of total U.S. GDP as of late 2025. This represents a record high, surpassing the previous peak seen during the dot-com boom in 2000.”

      Just putting it into context relative to the entire US economy.

   3. Thank you for posting the link to shumer.dev – MOST enlightening!

      “Here’s a simple commitment that will put you ahead of almost everyone: spend one hour a day experimenting with AI. Not passively reading about it. Using it. Every day, try to get it to do something new… something you haven’t tried before, something you’re not sure it can handle. Try a new tool. Give it a harder problem. One hour a day, every day. If you do this for the next six months, you will understand what’s coming better than 99% of the people around you. That’s not an exaggeration. Almost nobody is doing this right now. The bar is on the floor.“

      1. And one must sign up for $20/month.