Naïve Bae retweeted
I'll say it again for the AI risk experts in the back:
We've already run a real life, large scale experiment on what happens when a superintelligence is unleashed on a narrow objective.
Look no further than U.S. capital markets.
Millions of investors, analysts, traders and algorithms, backed by trillions of dollars and wired together through prices, all chasing one goal: the highest risk-adjusted return on capital.
And it is spectacularly good at its job. It funds the best ideas, punishes the worst ones, and turns new information into prices faster than any of us can react. No individual can out-think it. 99+ percent of professionals can't beat it.
But look at what else it's done along the way:
- Optimized the measure instead of the intent. Quarterly earnings became the scoreboard, so companies learned to manage the scoreboard.
- Treated anything without a price as free. Pollution, addiction, obesity, depression, political fragmentation, fragile supply chains, the long-term health of the communities it operates in.
- Found every gap in the rules, then paid lobbyists to help write the next set and pull up the ladder behind incumbents.
- Remade the world around its objective. Companies, pension funds and our retirement accounts now answer to the S&P 500.
- Made itself impossible to switch off, because everything else now depends on it.
None of this required malice. Nobody sat in a room and chose these outcomes. It's simply what a very capable optimizer does when its objective is a narrow proxy for what we actually care about.
And note that we never solved this alignment problem. We've managed it, imperfectly, with regulation, taxes, disclosure rules and the occasional crisis to show us where the guardrails were missing. Over time, the market has captured the instutions designed to constrain it. We've been at it the better part of a century, and the misalignment is the worst it's ever been.
So the next time someone tells you AI is only dangerous if it "wants" something bad, or if its goals are mis-specified, point them at the S&P 500. It doesn't want anything. It just optimizes.
That's the whole problem.
私を構成する9つのゲーム
1. スーパーマリオブラザーズ
2. ゼルダの伝説
3. ダブルドラゴンⅡ The Revenge
4. ロックマン
5. バイオハザード
6. ファイナルファンタジーIV
7. ストリートファイター Zero 3
8. ゼノギアス
9. パーフェクトダーク
#My9Games #私を構成する9つのゲーム
my9games.com/s/183f03f61d934…
Naïve Bae retweeted
A friend of mine works in AI and thinks bad things are about to happen. He sold his home and moved his family out the country. He said if you own a home in a walkable San Francisco neighborhood near a taqueria with good veggie options, you should sell your home right now.
Maybe the value isn’t ever in the models, but in the experience you gain doing the analysis. The models are simply the artifacts of that process and if you don’t have access to the universe of inputs used and the experience of how to analyze everything, you don’t have an edge.
A former colleague of mine just shared one of my old business cycle indicators.
I built it. It stayed at Pictet when I left.
Funny thing about rebuilding a process from scratch... it forces you to revisit your old work with a fresh pair of eyes.
I've done it many times now. Every time, there are improvements to be made.
So let me be blunt about this one: it's dead.
It’s a composite. Labor market conditions, the yield curve, the output gap, profit margins, consumer sentiment, wages... roughly fifteen inputs, ranked in percentile.
Many of those inputs have lost their meaning in a K-shaped economy. They were built to read a broad, synchronized cycle, where labor, credit and profits moved together. That economy doesn't exist anymore. So the signal degrades.
You can see it plainly in the chart: the last time this indicator hit the 90th percentile, it should have called a recession. In the old world it would have. This time it didn’t.
The changes I've been writing about for the past eighteen months have broken most of these old-school macro models. Or at the very least, made them far less important than they used to be.
This is exactly why the loud voices calling for recession these past 2-3 years have all been wrong. I took the other side.
I explained why they were wrong... initially in a GMI piece last year, then more completely in one titled The Compute Cycle, one of the last pieces I wrote before I left GMI.
Here's the bigger issue.
Everyone is building dashboards now. AI handed the world a new set of power tools, and almost overnight, anyone could ship a model.
However, AI made it possible for anyone to build tools. What it has not done is make everyone a toolmaker.
The difference is judgment.
Most people have been handed the capability without the apprenticeship.
They can produce the model. They cannot tell you how it works. And if you can't do that, you can't refine it.
The oldest rule in this business does not care how good the technology gets: garbage in, garbage out.
This old chart of mine is not the only one.
The Financial Conditions Index I developed years ago… same story: dying.
Why?
Because the ground underneath these indicators is moving, and fast.
For forty years the business cycle lived inside a cage. That cage held through the dotcom bust, the Global Financial Crisis, COVID, and half a dozen Fed cycles.
Then this year global semiconductor sales posted the largest annual increase in over four decades, and the cage broke.
Compute stopped being a downstream input to the real-world production cycle and became the binding constraint.
Growth now starts with chips and propagates outward. Everything else sits downstream. And that guts the old financial conditions logic.
These frameworks were built for a world where cash flow sensitivity to rates, a strong dollar and high oil prices governed the cycle.
The dollar still matters, it always will, it's the price of global liquidity. Rates and oil still matter too. But their grip on the cycle has loosened.
When capex is funded off corporate balance sheets running a moonshot, and the buildout is a strategic imperative nobody can opt out of (the Red Queen Effect), those old transmission channels matter far less than they used to.
When the causality of the entire cycle flips, the tools built to read the old one don't just need tuning. They need burying.
For well over a decade I've built asset allocation models and business cycle frameworks for some of the most sophisticated investors in the world.
That work, and the discipline of staying relentlessly data-dependent, is what I'm known for.
The craft beneath it has always been the same. Statistics and econometrics. The unglamorous stuff you can't shortcut, and can't fake.
It comes with me on this next venture. A full suite of new indicators, built from the ground up.
I'll be sharing much more in Q4.
The macro landscape is shifting. More now than ever.
Time to sit up...
Naïve Bae retweeted
Congrats on your Designated Players and $500M soccer-specific stadiums, MLS fans. Meanwhile, USL is out here playing real football in front of 14 people and a raccoon on a Wednesday night during a thunderstorm. You ever seen a center back score a 94th-minute winner after the goalkeeper got sent off and the backup keeper is a 5'9" academy midfielder? Didn't think so.
While y'all are busy arguing over which team has the best Designated Players, Louisville's Sting and Detroit's Friendly are throwing hands at midfield, and some random 19-year-old on loan from an MLS academy is scoring a bicycle kick in a game that ends 4-3 after 11 minutes of stoppage time. You wouldn't survive one cold, rainy Tuesday night in Portland.
Interesting @fallacyalarm - did you come across these guys when researching Aurora competition?
Most people think driverless trucking is years away. @Gatik_AI is already moving freight for some major brands you’d recognize. @TashaARK explains why ARK doubled down on its ARK Venture Fund investment.
Holdings: ark-funds.com/funds/arkvx
Bloomberg News today telling me something I already know was coming, as a Fallacy Alarm subscriber
@fallacyalarm
Rene wrote about $JACK and a rotation into B2C thesis a month ago. The stock is up 15%
substack.com/@fallacyalarm/n…
@fallacyalarm readers have known the debasement trade was in trouble at least as far back as October last year
We might be slowly figuring it out @3four3 !
Naïve Bae retweeted
Proud to unveil the Luce Migliore
A Ferrari electric vehicle designed by me and ChatGPT in 8 seconds at a cost of $0.10
🤖 Made with AI
