Quinn Slack

Freedom of Intelligence Anthropic has created a dangerous, destabilizing mess by lobbying for and getting US government restrictions on models like Mythos, Fable, and GPT-5.6. Now the US government is deciding who has access to which models, and the best models are accessible only to a very few and a very rich set of companies. Nobody wants that. Even Anthropic doesn’t like what happened. Now the rest of us need to clean up the mess. How? We need to fight for our freedom of intelligence, the freedom from government restrictions on who can use which AI models. If we allow government to decide what level of intelligence someone can access, no matter how well intended, we’ll be less safe and forever divided. What Freedom of Intelligence means Freedom of intelligence means the government may not restrict which AI models you can use. This means that the government must not require licensing of model labs, or approval of models prior to release. Otherwise government inevitably will use that power to restrict releases to certain favored individuals and companies (as we’ve just seen) and to introduce biases. Freedom of intelligence also means that the government may not prohibit you from downloading and running open models. If someone commits a crime with the use of AI, that already is illegal and should remain illegal. The government must not force a model lab to release a model against its wishes. If a model lab chooses to release their own model to only a few privileged people and companies (as Anthropic did with Mythos), or to keep it internal, that is their right. Other model labs can compete by serving the rest of the market. It shouldn’t be illegal to offer frontier intelligence to small businesses, startups, and individuals. Intelligence is fundamental When people argue against freedom of intelligence, they say: AI is powerful and sometimes dangerous, and we’ll be safer if the right people control AI the right way. They’re right about the first part and naive about the second part. For something as fundamental as intelligence, there is no such thing as the “right people” to control intelligence, nor the “right way” to control intelligence. People will disagree. People already disagree very, very strongly. In a democratic society, the only stable equilibrium for a bitterly divided realm is to grant individual freedom. Intelligence is not the same as speech or religion, but it is every bit as powerful and dear and deserving of freedom. There is no democratic way to regulate access to intelligence Nobody likes the current US government policy on model restrictions. Nobody really knows what it is, even, or knows what it will be next week. Today, Monday, June 29, 2026, the US government is choosing which people and companies can and can’t access Anthropic’s and OpenAI’s frontier intelligence. Who is deciding? Based on what criteria? Nobody knows. Maybe you think that the US government’s behavior in the last few weeks is a blip, and that the “right people” will control AI the “right way” soon. Maybe you hope, like Dario Amodei, that “qualified third-party”[1] regulators shielded from “political favoritism or arbitrary decisions” will swoop in and take control of AI policy. That’s just not how it works in our political system, certainly not for a high-salience, zero-sum issue like access to intelligence. We would never, ever, ever pass a regulatory apparatus where the most important national policy decisions are decided by unelected experts, free from accountability to the voters. Nor should it pass. (Ironically, the only way it might pass is if Anthropic is the politically favored one, which would violate Dario’s own stated proposal.) But suppose Dario gets lucky and his “Federal AI Control Administration” (my name for it) is created. And suppose on day 1, the Federal AI Control Administration approves the release of Claude Mythos 5, but only to ~100 of the biggest corporations in the US, in order to limit the risk. (Dario would support this government action, presumably, since it’s what Anthropic itself deemed optimal.) On day 2, the Federal AI Control Administration starts deciding which companies should get access to GPT-5.6. Suddenly, “AI safety” has turned into “picking winners and losers”, because it’s safer to not give frontier intelligence to everyone. Of course, this is the actual reality today. Does this sound like the kind of thing that voters in a democracy, already distrustful of AI and of corporate power, would support? No. Is this stable? No. Play it forward a bit. What do you think the 101st biggest company, denied frontier intelligence by the US government, does first: sue or curry political favor? What do you think the US executive branch does with this newfound power? What do you think Anthropic’s corporate rivals, like Amazon and Google and OpenAI, do with their newfound powers to summon arbitrary regulatory fury on each other? There’s no way to sustain a stable, democratic arrangement where government controls access to intelligence. The more powerful you think AI is, the less stable is any attempt to regulate access to intelligence. (By the way, I truly believe Dario and AI safety adherents are true believers with good intent. I am not arguing that they are evil or greedy.) Freedom is counterintuitively stable My biggest fear is that we’ll oscillate around bad AI regulation, with daily distractions and growing corruption, not realizing that the only stable equilibrium is freedom of intelligence. While intelligence is not exactly like speech, the analogy to freedom of speech is useful. Both speech and intelligence are powerful and sometimes dangerous. For thousands of years, kings and despots tried just banning bad speech, imposing probably well-intended “speech safety policies” (i.e., jailing and exiling and killing dissenters). This didn’t work. Our smartest minds, trying as hard as they could for thousands of years, having tamed fire, water, animals, wind, and space, never figured out a way to regulate truth. So, after trying literally every other speech policy, we arrived at freedom of speech: just let people speak, even if they’re wrong, even if their ideas are dangerous. This is, overall, the best policy. It’s counter-intuitive that allowing all the bad speech is better than just giving someone the power to decide what is “bad speech”. It’s so counter-intuitive that we call freedom of speech a human right, which is society’s way to say as strongly as possible, “we wrote this rule in blood, don’t mess with it.” I favor freedom of intelligence for the same reasons. Like speech, AI is powerful and sometimes dangerous. But it’s far more dangerous and unstable to give someone the power to decide what intelligence everyone else can use. Speak up now It feels risky to speak up. Friends and business partners share thoughts similar to mine here. I’ve talked to many of them in the past weeks. But these conversations happen in hushed tones, off the record. Why? Because Anthropic is a king and a kingmaker. We all use or have used their models, they’re great, and we’re scared of losing access or being shut out by them after criticizing them. Anthropic can unilaterally dictate the terms of their commercial relationships, including early access to new models, pricing, data retention, and much more. I have many friends at Anthropic. They’re great people and mean well. They don’t know what people truly think of Anthropic and its lobbying because everyone’s too afraid to speak up. But the more we speak up, the more Anthropic might be able to change from within. If you’re still afraid to speak up, feel free to reach out to me privately to chat (quinn@slack.org). If Anthropic retaliates against me or you for speaking up on this grave matter of national policy that they’re also lobbying on, that would do more than anything to prove our point. How to fight for freedom of intelligence First we need to change minds, then we need to change laws. To change minds, go and talk to people in the real world about freedom of intelligence. Use whatever you find memorable from this post, and figure out your own way to convince people. Share what works. If you’re in San Francisco, join us on Tue Jun 30, 2026, at 6:30pm (link [2] in reply) to start discussing and pushing for freedom of intelligence. Otherwise, organize in your own city, to spread the word and normalize this freedom before we lose it. Why I’m hopeful Nobody, nobody wants access to intelligence to be limited to a very few, and a few rich companies. Freedom of intelligence has broad appeal. Let’s build that big tent.

The kinds of mistakes made by smart non-native English speakers are different from the kinds of mistakes made by not-smart native English speakers. I don't discount the former at all. So, I wouldn't recommend a non-native English speaker to use AI to make their writing be "better English".

@fastchicken @AmpCode @josh_robb Awesome. Anything we could do to make this easier?

ok, got our webapp running in an (B)Orb in @AmpCode so I can run it up, and load the site (vite in the orb) from my laptop. mint. This will be SUPER useful for our designers, PMs and... the COO (hi @josh_robb ) As always, CORS was the pooh in the pool.

Go to 1h32m to watch me, @evanphx, and @thorstenball play the quiet game

@alexellisuk @ibuildthecloud Next time you see it, let me know. There is an update notice after ~8h+ of inactivity, but it doesn't block you.

@sqs @ibuildthecloud In my mind, I started the agent, made a few prompts, and the next day went to say "write that to a file" and got the blocker/nag. Could be mistaken.

Yeah, we are stricter if a user is using free credits because we need to prevent abuse and stay ahead of abusers. We don't require you to update when you're using paid credits (except when something truly changes that requires it). But we've done a lot of work to make resumption smooth...in fact, thinking and streaming all continue even when your CLI isn't running.

@usr_bin_roygbiv once you've worked like this, you can never go back

Every once in a while you work at a company that only hires people who know what they're doing, and suddenly its 20 people doing the same work as 400 somewhere else. There's zero meetings, everyone talks once a week on slack, and you go huh, how much garbage is there actually.

@alexellisuk @ibuildthecloud The grace period is 48 hours. What are you seeing that forces you to upgrade when you're 1 version out of date? (That would be like every 2-4 hours.)

@sqs @ibuildthecloud Have you thought on N releases as grace? To prevent it being to much of a nuisance. Or "updated within last 72 hours"?

@who_took_ratin @AmpCode It's from WorkOS. Is this your preferred language or device's language?

@AmpCode sending the mobile verification messages in Hindi is confusing. For a second I thought it was a scam. @sqs

@towrywang Where did it get stuck or go wrong?

Grok 4.5 has successfully and quickly fixed two different hard issues , while gpt sol keep going on the wrong direction slowly. Even ampcode harness can not save it 😂

#2 is all relative. You don't think about memory allocation in most cases I'd guess. But to someone else that is a piece they need to break it down into. As I use agents, previously important decisions become unimportant. It's easy to fix a whole class of bugs and code issues later with agents. You can take on new kinds of tech debt safely that before you'd quickly regret. So I just try to make the right decision about whether I need to care about a given decision and err on the side of not needing to care if I've seen the agent do it well before.

One thing I've been struggling to reconcile is the fact that there are two ways to tackle a difficult task that has a lot of unknowns: 1. To throw a big model at it and let it rip through and make decisions. 2. To use a proper software engineering approach where you break down the problem into small, solvable pieces where you can make informed decisions. I kind of feel that I would rather do the latter rather than the former most of the time. I’d rather my AI tools help me to the latter well. A better model (i.e. Amp High) obviously helps, but what if it helps specifically with the “breaking it down” rather than with “powering through”? Curious to hear your thoughts on this.

The Dial is so much better than a model picker because it lets you convey your preferences/needs better and because it lets Amp use multiple models together for the best results

We now publish live % usage of Amp mode, for everyone and for just the Amp team. ampcode.com/models

@jamonholmgren Yep. If open models are restricted in the US, then that will reduce competition for the model companies and will lead to significantly higher prices. Or if the ability to release good models requires being favored by the gov't. x.com/natolambert/st… freedomofintelligence.org

@sqs The fact that tokens are relatively fungible _should_ help keep that down, I assume.

I have an earnest plea. Can you folks at OpenAI, Anthropic, and xAI just be really transparent with us and tell us how the projected economics of the frontier models will impact us in 6-12 months? I understand you need to recover your investment. I am not begrudging a price increase. And I get you need adoption and are investing in that right now. But it's impossible to do any long-term planning right now with all the obfuscation.

@jamonholmgren Yep. They will get good at charging you close to the price you're willing to pay.

@sqs This is helpful. And, if you're not okay with those terms, then be prepared to pay huge $$$, I assume

@BasFenix We don't train on user data and contractually promise not to (unless you and your company explicitly opt in), so no, we haven't done that, but yeah, that is interesting.

One thing that stands out: mid-thread changes aren’t possible by choice (new thread or restart), so the best training signal is already in the data? tasks that finish with fewer user turns, smaller edits, and quick acceptance vs. heavy back-and-forth or manual throttle (dial) up. That implicit feedback feels like perfect supervision for a lightweight auto-dial predictor. Grok’s Auto mode and Cursor’s Auto routing already do something similar by deciding depth internally. Curious if you’re already playing with outcome based selection like that ?

You should expect the continued availability of really good tokens at cheap prices IF you look like a user with a low willingness-to-pay. That is, IF you have no ZDR requirement and are OK with training, OK with non-guaranteed-US inference, OK with limited model choice, OK with refusals for cyber and bio, OK with insane terms (like agreeing to not compete with the model labs), etc.

@sqs How does this help me plan?

Learning OS in 1,000 LoC via @ampcode Orbs because I’m on a work trip without my personal laptop. This is *almost* programming by hand right? Right?! …ating-system-in-1000-lines.vercel.app/en/

(Except to the extent that it deters future high-margin purchases, etc., etc. Economists have done a lot of work here that holds up well.)

Note that when I say GPUs spoil, I don't mean they depreciate fast. I mean that if you let a GPU go idle, or an airplane seat go empty, you can never go back in time and reclaim that lost capacity. You are willing to sell capacity at any price above your marginal cost.

Asking "when will AI subsidies end?" is like asking "when will United Airlines stop offering economy class?" AI today is different from Uber/DoorDash mid-2010s subsidies. Models & GPUs are huge fixed costs that spoil, and the marginal cost of inference is low. Like airlines.