Sam Patt

Rational optimist |

Worked on OpenBazaar |

Wrote a book about Bitcoin |

I love lifting / AI / Geoguessr / programming

[Sam Patt @SamuelPatt]

It's highly likely Bitcoiners are delusional, but our delusions are of a fairer and freer world, unlike the delusions of people who think the perpetuation of the status quo systems will lead to anything other than economic ruin.

[Aella @Aella_Girl]

AI is clearly very intelligent. I don't understand how someone can work with the models in any sort of depth and think it's 'simply autocomplete.'

[Andy Masley @AndyMasley]

A lot of people have started saying that the "AI uses all the water" meme started with Empire of AI. I think this is definitely wrong, and what actually started it was the Washington Post article from 8 months earlier claiming that each ChatGPT prompt used a whole bottle of water.

This kicked the meme into the stratosphere. There are so many infographics made about AI mentioning that it uses a bottle per prompt, so many thousands of popular videos and images where someone holds a bottle begging people not to use AI. It got so much news coverage.

My claim in my new post (link below) is that I've found extremely strong evidence that this claim was based entirely on napkin math that ignored lots of simple things we knew about the hardware and software running GPT-4, and if you just account for those the cost drops by 50 to 200 times. This shouldn't have ever been allowed to influence the discourse at all, never mind get an infographic made in a major newspaper.

I think the Washington Post should either retract or correct the original article and make it clear that it was based on napkin math, or publish the methodology. The authors have never made it clear how they got the number, and if I'm right it shouldn't have been allowed to ever be the basis for a claim this strong.

[Sam Patt @SamuelPatt]

LLMs are absolutely 100% not a fuzzy search engine.

They are not just giant databases.

If Jim spent 5 minutes talking to an LLM it could easily explain all of this to him.

[Jim Stewartson, Decelerationist 🇨🇦🇺🇦🇺🇸 @jimstewartson]

I’m getting tired of “experts” like this misunderstanding what they’re looking at.

LLMs are giant databases of stuff HUMAN BEINGS have done.
They are the EXHAUST of humanity.

Prompts are database queries into EXISTING DATA.
It’s a fuzzy search engine, not intelligence.

[Perry E. Metzger @perrymetzger]

We live in two worlds. In one of them, it was announced today that an AI had saved the life of a little girl and that another had made an important mathematical breakthrough. In another, the supply of water is about to somehow vanish because of buildings with computers in them.

[Sam Patt @SamuelPatt]

The one congressman opposing his own party on principles, and losing, while people spout impossible physics and conspiracies about data centers is too much for me.

I have no faith in a political process where people are so easily manipulated.

[Sam Patt @SamuelPatt]

New AI project idea: create an "ignorance threshold" and automatically run all viral tweets against it; if they don't surpass the threshold, then block the account along with everyone who liked.

This wouldn't pass the threshold.

[S.🎧 @1ssve]

why does no one care about the fact that we’re projected to run out of clean drinking water by 2039 because of how much water ai data centers use please stop using chatgpt y’all, i beg. you do not need a shitty looking cartoon of yourself, i promise. the planet is dying.

[Sam Patt @SamuelPatt]

Remember when @Snowden warned us that the only thing the government needed to turn us into a surveillance state was a bunch of new data centers?

[Sam Patt @SamuelPatt]

"Data centers with no limits"

The point of the market is to allow people to voluntarily exchange with each other so that information about resource allocation flows freely and creates natural limits (through the price mechanism) without needing artificial limits (through threats of violence; the political mechanism).

Limits always exist. Are they created by the political process or a voluntary one?

[Mark 🇺🇲 @IdleAn94328]

No Data centers at all is an extremist position.
Data centers with no limits is also an extremist position.

Now that the two extremes have been defined you can identify and dismiss the extremists.

The rational debate from reasonable people occurs between those two extremes.

[Sam Patt @SamuelPatt]

My anecdotal evidence from living in rural areas for half my life is that most people outside of cities know that there's tons of unused (or under-used) land available.

When you drive through nothing for 30 minutes to get to the next town this concern for the land seems a bit ridiculous.

[Sam Patt @SamuelPatt]

Bizarre line of reasoning on display here.

Polling data shows ~70% of people oppose local data center construction and ~30% of people support it.

This is presented as proof that accounts which support data centers are all "bad-faith" and must be fake or doing it for the money.

If you lined up 100 people and asked them their opinion, why would you only listen to the 70 which agreed with you and disregard the other 30?

[Silicon Valley Fodder @Playerinthgame]

Bad faith pro-datacenter noise is out in force but the polling data doesn't lie. The vast majority of people don't want hyperscale datacenters built.
We know these pro-DC accounts are either:
- people who profit or hope to profit
- paid accounts or
- agentic

[Sam Patt @SamuelPatt]

1.5m views for a claim that only takes a rudimentary understanding of physics to debunk.

20 degrees across an entire state is an unfathomable amount of energy.

[🦢 @ZainabSana2622]

At this point it’s obvious the billionaires are trying to kill us. What do you mean the new AI data center in Utah will raise the state’s nightly temperature by 20+ degrees???

[Sam Patt @SamuelPatt]

This feels a bit too much like labor theory of value, but I agree in most cases.

I would say ai generated text is usually bad because writing is sharing a part of your mind, a part you feel compelled to share and / or expect others to value. You can't tell an AI how to do that.

[eigenrobot @eigenrobot]

asking people to read ai-generated text is offensive.

this is not because ai text is intrinsically bad. rather, the author has not paid a cost to write the text himself. this cost is a credible signal he finds its communication important.

so: not paying that cost is telling

[Sam Patt @SamuelPatt]

That's the old GPT 3 version, I got the new copy

[alex peysakhovich @alex_peys]

got a framed copy to hang by the ai team

[Sam Patt @SamuelPatt]

from the board game perspective it's because they have the most uncontested room to expand (west) and no enemies on one side (north)

Also none of their enemies can focus solely on them - if anyone seriously tried to take them over they would immediately expose themselves to all other enemies they border

Florida cannot expand and the Gulf Coast or Piedmont are forced to capture them in order to have any chance

[Sam Patt @SamuelPatt]

Board game players know that the Great Lakes megaregion will eventually dominate.

Florida first to die.

[John Potter 🌐🩸 e/acc @jjohnpotter]

What’s life like outside of the megaregions

I feel like I’ve lived a lot of places but turns out never outside of a megaregion

[Sam Patt @SamuelPatt]

12 is a unique number that deserves a unique word.

It's divisible by 2, 3, 4, and 6, which is quite useful in everyday life, much better than 10.

[Ethel Braithwaite @Ethelbrait1941]

Never stop saying "dozen" and "half dozen". Never stop using the word you read in an old novella. Never stop using your regional jargon. Don't succumb to an internationalized English stripped of its whimsy and romanticism in the name of streamlining global commerce.

[Sam Patt @SamuelPatt]

"the economy (stock market)"

oof

[Richard Hanania @RichardHanania]

Basically, the outcome of the war with Iran has been

1) Hormuz remains closed
2) The economy (stock market) is doing fine

I didn't see anyone predicting this. People critical of Trump said Hormuz would stay closed and it would be a disaster. His defenders also assumed he needed to open it.

Is the lesson just that markets are much more resilient than we think, and the AI boom is so powerful we can withstand anything?

[Sam Patt @SamuelPatt]

I feel this way too. If you've ever thrown a really hard problem at a model and gotten truly insightful responses, it's hard not to scoff at the "it's just auto-complete" mentality. Dawkins' reaction is understandable.

[Robert P. Murphy @BobMurphyEcon]

I understand why everyone is dunking on Richard Dawkins, but fairness compels me to speak up: If you took a complex project you were working on, uploaded it to Claude, and had a 3-day conversation about it, you wouldn't be making fun of Dawkins for saying it's conscious.

[Sam Patt @SamuelPatt]

"Meanwhile the numbers are already tracking the curve."

The numbers they cite are layoffs, but they don't attribute this to AI solely and they demonstrate a change in demand.

This "race to the bottom" type of claim only works on paper.

[Elias Al @iam_elias1]

Two economists just published a mathematical proof that AI will destroy the economy.

Not might. Not could. Will — if nothing changes.

The paper is called "The AI Layoff Trap." Published March 2, 2026. Wharton School, University of Pennsylvania. Boston University. Peer reviewed. Mathematically modeled.
The conclusion is one sentence.
"At the limit, firms automate their way to boundless productivity and zero demand."

An economy that produces everything. And sells it to nobody.
Here is how you get there.
A company fires 500 workers and replaces them with AI. A competitor fires 700 to keep up. Another fires 1,000. Every company is behaving rationally. Every company is following the incentives correctly. And every company is building a trap for itself.

Because the workers who were fired were also customers.
When they lose their jobs faster than the economy can absorb them, they stop spending. Consumer demand falls. Companies respond by cutting costs — which means automating more workers — which means less spending — which means more falling demand — which means more automation.

The loop has no natural exit.
The researchers tested every proposed solution. Universal basic income. Capital income taxes. Worker equity participation. Upskilling programs. Corporate coordination agreements.
Every single one failed in the model.
The only intervention that worked: a Pigouvian automation tax — a per-task levy charged every time a company replaces a human with AI, forcing them to price in the demand they are destroying before they pull the trigger.

No government has implemented this. No major economy is seriously discussing it.
Meanwhile the numbers are already tracking the curve. 100,000 tech workers laid off in 2025. 92,000 more in the first months of 2026. Jack Dorsey fired half of Block's workforce and said publicly: "Within the next year, the majority of companies will reach the same conclusion."
Nobody is doing anything wrong. Companies are following their incentives perfectly. That is exactly the problem.

Rational behavior. At scale. Simultaneously. With no mechanism to stop it.
Two economists built the math. The math leads to one place.

Source: Falk & Tsoukalas · Wharton School + Boston University ·
https://t.co/4m8E9jQNYm

[Will Rinehart @WillRinehart]

Today I'm launching AI Policy Hub, a project I've been working on and developing the last couple months.

While I have plans for other pages in the future, it currently features
- A state AI bill tracker that automatically updates every Monday
- A federal AI bill tracker that also updates every Monday
- A list of major government actions on AI
- A curated list of FRED charts that are important for understanding AI's economic impacts
- A economic trends page built on my testimony to the JEC, describing what's happening accorrding to the data; and
- A narrative description of my AI work

This is a working project, so please send me ideas for additions or changes!

The page is here: https://t.co/Z56upt2LhZ
My Substack on the project is here: https://t.co/ldTPfEYqff