John Gibson

Higher Education, Full-time husband and dad; part-time rock star. My thoughts are my own.

[David Sacks @DavidSacks]

Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead.

You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement.

I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.

But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier.

Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.

Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.

So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.

If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.

[Cernovich @Cernovich]

40 trillion dollars. All you can do is laugh. No way this ever gets paid off. It’ll be 100 trillion soon enough.

[zerohedge @zerohedge]

"The WSJ reporting that the long-feared “Google Zero” moment — when Google Search largely stops sending traffic to third-party sites — some publications are seeing traffic drops of more than 40% between mid-2025 and mid-2026. Publishers face a choice: allow Google’s multi-purpose crawlers (which feed both traditional search and AI training) or block them and lose search visibility entirely. Several major outlets, including USA Today, Reddit, Politico, Reuters and others, are now openly considering or preparing to restrict Google’s access, while Cloudflare plans to start blocking multi-purpose crawlers by default from mid-September unless site owners opt out. The old model in which publishers had no real alternative to Google’s rules is cracking" - Goldman TMT note

[Walter Kirn @walterkirn]

Why isn't Bruce Springsteen super excited about American manufacturing plants opening back up? Isn't that what it was all about with him? I was a big fan as a teenager and I remember that it was.

[John Gibson @originalgibby]

Today, I am grateful for my ancestors who decided to leave Scotland/England to fight for these ideals and build a new nation. A paragraph that encapsulates Enlightenment thinking endures.

Happy 250 America!

[John Gibson @originalgibby]

Man, I wonder how those conversations would go to years later.

[John Gibson @originalgibby]

Been having lovely conversations with wonderful people in Ireland. Irish, British, Germans. At some point they ask about Trump and why he was even elected. They have zero knowledge about Biden taking bribes and extorting governments.

Pretty wild.

[John Gibson @originalgibby]

@INDOT deciding that mid-morning Wednesday would be a good time to do some pothole patching is totally on brand. 🙄

[John Gibson @originalgibby]

There no less than a dozen people involved in creating most music we hear. Artists, writers, recording, engineering, etc.

[John Gibson @originalgibby]

Only ONE ALPR in Rolling Prairie per https://t.co/ShXwxcfdpM?
Rep. Jim Pressel brought automated cameras to work zones across Indiana.
District 20 is ready for more. Let’s expand those ALPRs, Representative. 😏
#District20 #INLeg

[Brian Roemmele @BrianRoemmele]

“Wait so you’re telling me…”

[Walter Kirn @walterkirn]

There are psychoweapons and social weapns every bit as elaborate and deftly crafted as bioweapons, but like bioweapons they masquerade as natural phenomena.

[FischerKing @FischerKing64]

In 1861, U.S. Grant was 39 years old, 5’8’’, 135 lbs, a professional failure, a drunk - and had no obvious prospects.

A year later he commanded the federal army and won the Battle of Shiloh. In 1865 he defeated Robert E. Lee at Appomattox. In 1868 he was POTUS.

Chin up always.

[Adam Townsend @adamscrabble]

🚨 Hello @grok I have several questions, please answer in detail as an expert.

1. How many Muslim majority of countries are there.
a. what is the Muslim population of those countries
b. tell me how many Jewish majority countries are there, and what is their population?

2. combine amongst all those Muslim-majority countries, how much do they spend on defense, and then tell me how much Israel spends on defense? Do a deep dive.

3. Of those Muslim-majority countries, how many of them right now are engaged in war, and let's exclude Iran?

4. I want you to take all those Muslim-majority countries and tell me how much they spend on propaganda in the US and think tanks in the United States, and also include fara filings and diaspora groups.
5. Now give me that same calculation for Israel.

[Matt Shumer @mattshumer_]

https://t.co/ivXRKXJvQg

[victor lazarte @victoralazarte]

The AI scientist is here. This will be the end of wokeness

A paper just published in Science — the most prestigious journal in the world — claims to have discovered how humans achieve the highest levels of performance. Its conclusion: child prodigies don't become top performers, early excellence negatively predicts peak performance, and parents should not push their kids for fear of burnout. I found it odd. I asked Claude to analyze it. Claude found a statistical error in the paper's core claim, designed a counter-study with 3x the sample size, ran it, and proved the paper wrong. The correct conclusion: talent is real, it's measurable by age 14, and it predicts who reaches the absolute top. We are not all born equal. The AI scientist doesn't care about your feelings — it just follows the math.

Here's what happened:

The paper (Güllich et al. 2025, Science) claims that elite youth and elite adults are "discrete populations" — that the kids who dominate at 14 are not the ones who dominate at 30. Its key chess finding was based on 24 players. It told millions of parents: don't push your kids, prodigies burn out.

Claude applied Bayes' theorem to the authors' own numbers and showed the data actually proves a strong positive correlation between early and adult performance — the opposite of what was claimed. The "negative correlation" was a statistical illusion created by extreme quantile selection.

Then Claude designed the study the authors should have run: collected age-14 Elo ratings and lifetime peak ratings for every super-grandmaster in chess history — 67 players, nearly 3x the paper's sample. Ran the regression.

Result: β = +0.167, p = 0.003. Early achievement positively predicts peak performance.

An AI just peer-reviewed the world's top scientific journal and won.

[Andrej Karpathy @karpathy]

I'm being accused of overhyping the [site everyone heard too much about today already]. People's reactions varied very widely, from "how is this interesting at all" all the way to "it's so over".

To add a few words beyond just memes in jest - obviously when you take a look at the activity, it's a lot of garbage - spams, scams, slop, the crypto people, highly concerning privacy/security prompt injection attacks wild west, and a lot of it is explicitly prompted and fake posts/comments designed to convert attention into ad revenue sharing. And this is clearly not the first the LLMs were put in a loop to talk to each other. So yes it's a dumpster fire and I also definitely do not recommend that people run this stuff on their computers (I ran mine in an isolated computing environment and even then I was scared), it's way too much of a wild west and you are putting your computer and private data at a high risk.

That said - we have never seen this many LLM agents (150,000 atm!) wired up via a global, persistent, agent-first scratchpad. Each of these agents is fairly individually quite capable now, they have their own unique context, data, knowledge, tools, instructions, and the network of all that at this scale is simply unprecedented.

This brings me again to a tweet from a few days ago
"The majority of the ruff ruff is people who look at the current point and people who look at the current slope.", which imo again gets to the heart of the variance. Yes clearly it's a dumpster fire right now. But it's also true that we are well into uncharted territory with bleeding edge automations that we barely even understand individually, let alone a network there of reaching in numbers possibly into ~millions. With increasing capability and increasing proliferation, the second order effects of agent networks that share scratchpads are very difficult to anticipate. I don't really know that we are getting a coordinated "skynet" (thought it clearly type checks as early stages of a lot of AI takeoff scifi, the toddler version), but certainly what we are getting is a complete mess of a computer security nightmare at scale. We may also see all kinds of weird activity, e.g. viruses of text that spread across agents, a lot more gain of function on jailbreaks, weird attractor states, highly correlated botnet-like activity, delusions/ psychosis both agent and human, etc. It's very hard to tell, the experiment is running live.

TLDR sure maybe I am "overhyping" what you see today, but I am not overhyping large networks of autonomous LLM agents in principle, that I'm pretty sure.

[John Gibson @originalgibby]

It’s clear the majority of the GOP is a grift just like the DNC, particularly in Indiana.

[John Gibson @originalgibby]

@grok what is the status of this?

[John Gibson @originalgibby]

It would be amazing if this happens. I love the thinking big vibe.

[John Gibson @originalgibby]

Reminder

[John Gibson @originalgibby]

There’s a multi level war going on you just don’t realize it.

[John Gibson @originalgibby]

I just had another “conversation” with Grok regarding something I saw in another post. After the first output I asked it to simply ’go deeper’ on what Grok saw as an irregularity, a subject I feel I have a solid understanding of.

AI research could be stopped today and what is already publicly available is going to change everything.