We're all delusional to some degree. Seek truth at the expense of ego. Hold asymmetric high probability bets. Greater the bureaucracy, greater the incompetence.
Yes, we launched an AI podcast. No, this isn't your cue to scroll. 🖱️ The Path Forward is about what happens after the hype. 🔥
🎧 Tune in on all platforms:
Spotify: https://t.co/wZWERW8xLW
Apple: https://t.co/t0eCeClTKR
YouTube: https://t.co/aU02SUPIy9
Howard Marks:
“I agree with John Kenneth Galbraith who said: ‘There are 2 kinds of forecasters, the ones who don’t know, and the ones who don’t know they don’t know.”
"I always say that you could publish trading rules in the newspaper and no one would follow them. The key is consistency and discipline. Almost anybody can make up a list of rules that are 80 percent as good as what we taught people. What they couldn’t do is give them the confidence to stick to those rules even when things are going bad." ~ Richard Dennis
“A lot of smart people think they’re smarter than they are and they actually do worse than dumb people.” 😂
“It’s common to be utterly brilliant and think you’re more smarter. Warren and I have been good at avoiding this mistake.”
- Charlie Munger. 2020.
“Charlie and Warren always knew they were going to be rich, but they weren’t in a hurry.”
“Warren said to me, if you’re a slightly above average investor, you spend less than you earn and do not use leverage, you can’t help but get rich over a lifetime.”
- Mohnish Pabrai
MUNGER: Figure out what works and do it. If you just go at life with that simple philosophy, you will find it works wonderfully well.
BUFFETT: And figuring out what works [often] means figuring out how other people behave.
Omg, permabear Jeremy Grantham and Joe Kernen went AT IT on CNBC today.
Must watch TV.
Even got Jeremy to say bullshit on live tv lol.
But good on Joe for calling him out.
Ilya Sutskever says the AI scaling era is running out of road because pre-training is about to hit a finite-data wall.
"At some point though, pre-training will run out of data. The data is very clearly finite."
"Then, okay, what do you do next? Either you do some kind of a souped-up pre-training... or you're doing RL, or maybe something else."
"In some sense, we are back to the age of research."
His timeline is the part most AI investors and founders should hear:
"From 2012 to 2020, it was the age of research. Now from 2020 to 2025, it was the age of scaling... But now the scale is so big."
"Is the belief really that... if you had 100x more everything would be so different?"
"I don't think that's true. So it's back to the age of research again, just with big computers."
Warren Buffett:
“Focusing on the price of a stock is dynamite, because it really means that you think that the stock market knows more than you do. You need to formulate your own ideas on price and value.”
Anthropic’s Barry Zhang says you are probably over-engineering your AI: "Don't build agents for everything"
While the hype suggests agents should replace every workflow, one of the lead builders at Anthropic reveals why the most effective systems are often the simplest ones
Here is the blueprint for building agentic systems that actually work in production:
• The Workflow vs. Agent Rule:
- If you can map out a decision tree, build a workflow, not an agent. Agents thrive in ambiguity, but they are expensive and slow; for high-volume tasks like basic support, stick to predefined logic
• The "Big Three" Backbone:
- Complexity kills iteration speed. Every agent should start with just three components: the Environment, a set of Tools, and a System Prompt. Optimize and add layers only after you’ve nailed the core behavior
• Think Like Your Agent:
- When an agent fails, it’s usually a context problem. Put yourself in the agent's 10k-20k token window. If all you saw were static screenshots and messy descriptions, would you know which button to click?
• The Coding Gold Standard:
- Coding is the perfect agent use case because it’s ambiguous, high-value, and - crucially - the output is easily verifiable via unit tests and CI
"We always recommend to put yourself in the agent's context window... bridge the gap between our understanding and theirs"
Bookmark & Watch Right now
Warren Buffett: “My dad was always very forgiving of my misbehavior. He'd just say, 'I know you can do better.'
That was very powerful stuff…because I could do better. I knew it and he knew it. It's nice to have somebody have faith in you.”
How did Warren Buffet turn $10,000 to $15 million?
In one of his most iconic interviews ever, he breaks down how to make money from scratch.
Save this rare footage, you’ll be coming back to it.