/Tech2h ago

Nathan Lambert reports backlash for criticizing AI regulatory capture, choosing independence over lucrative big-tech roles

Story Overview

Nathan Lambert, working at a non-profit, has drawn sharper criticism than usual for calling out regulatory moves that he argues unintentionally curb open-source AI work, including pressures around model distillation. He frames his stance as rooted in a deliberate choice to skip big-tech paychecks so he can push for broader access and fairer outcomes, noting that many peers privately agree but stay quiet due to their affiliations.

671K518652.3K

#57

Original post

Miles Brundage@Miles_Brundage#57inTech

Still lots of alpha in not being at the top few companies - we need independent takes on this stuff now more than ever

Nathan Lambert@natolambert

I've been getting a lot more hate than usual as I try to speak my mind about regulatory capture / unintentional attacks on open-source.

It's pretty sad, as there are few people in AI that can speak their mind (most companies say they cannot) and I know many people agree with me silently.

I also get people saying that you only say that because it supports the outcomes you want, in a weirdly derogatory way. Of course this is true, but I'm choosing to turn down meaningful wealth so I CAN fight for these values, working at non profits to speak my mind. Building a future that is more inclusive, diverse in the application of AI, and fairer for our children.

I may not always be right, but it has been clear to me for a while that more openness right now will help way more than supporting the closed causes. I continue to re-visit this and don't think everything should be open like some of the open-source absolutists.

I also don't like a lot of my comrades making fun of anthropic, calling the people there evil, etc. Those are not the case.

Trying to stay the course!

10:58 AM · Jun 28, 2026 · 3.4K Views

Open Question

Why few voices stay independent

Lambert points out that most researchers tied to companies simply cannot speak freely on these issues, leaving a thin set of non-profit and independent voices to surface concerns about capture even when the underlying motives began with safety goals.

Policy Risk

Where the debate stays unresolved

Details on the exact scale of reported backlash, the status of specific distillation or export rules, and any wider effects on open-source work remain unspecified in the current discussion, even as supporters like Miles Brundage echo the value of staying independent.

Sentiment

Many users praised advocates like Natolambert for defending open-source AI against regulatory capture and speaking up, while a few criticized them for not opposing closed companies like Anthropic strongly enough.

Pos

86.1%

Neg

13.9%

41 comments with sentiment.

Cluster Engagement

Digg Deeper

No Digg Deeper questions have been answered for this story yet.

Posts from X

Most Activity

VIEWS4.6KLIKES73REPLIES5

Nathan Lambert@natolambert

I have a lot of discussions with people who literally want open models banned. Happened in 2023. Happened in 2024. Happens today.

1d4.6K73

BOOKMARKS1

Stella Biderman@BlancheMinerva

“I also get people saying that you only say that because it supports the outcomes you want, in a weirdly derogatory way.”

I get this a lot too and it’s deeply confusing. Accomplishing good outcomes does in fact have a major influence on what I want.

Nathan Lambert@natolambert

I've been getting a lot more hate than usual as I try to speak my mind about regulatory capture / unintentional attacks on open-source.

It's pretty sad, as there are few people in AI that can speak their mind (most companies say they cannot) and I know many people agree with me silently.

I also don't like a lot of my comrades making fun of anthropic, calling the people there evil, etc. Those are not the case.

Trying to stay the course!

2h1.2K161

RETWEETS49

Nathan Lambert@natolambert

I've been getting a lot more hate than usual as I try to speak my mind about regulatory capture / unintentional attacks on open-source.

It's pretty sad, as there are few people in AI that can speak their mind (most companies say they cannot) and I know many people agree with me silently.

I also don't like a lot of my comrades making fun of anthropic, calling the people there evil, etc. Those are not the case.

Trying to stay the course!

1d47.3K96186

Oleksii Kuchaiev@kuchaev

@natolambert Keep up the good work! Open source AI is here to stay.

1d32616

Dec0ySquid@Dec0ySquid

@natolambert No one is evil.

No one is a category.

I say Fuck anthropic and those responsible for their choices, not for who they are.

You're too fucking nice.

Good. Be kind, but not to the unserious.

If you didagree, tell me Anthropic and big AI firms aren't unserious.

1d8599

Chris 🇨🇦@llm_wizard

@natolambert Thanks for speaking up for those who can't, Nathan!

1d67916

XYZ_123@UArreghini45280

@natolambert Top models cannot be confined in the hands of few privileged people. Fuck closed models, fuck Amodei.

20h43381

velocity@govelocity_

@natolambert @AndrewCurran_ Good on you, but anthropic’s posture is a misanthropic one. The operating principle seems to be that humans are too immature to wield such power, and should be stewarded by the model—except for a chosen few. Those few, they’re the good ones.

1d2527

Stella Biderman@BlancheMinerva

Not to mention the fact that the implication that I only want things that are financially beneficial to me is objectively hilarious coming overwhelming from people who work at companies that would triple my salary if I worked there.

Stella Biderman@BlancheMinerva

“I also get people saying that you only say that because it supports the outcomes you want, in a weirdly derogatory way.”

I get this a lot too and it’s deeply confusing. Accomplishing good outcomes does in fact have a major influence on what I want.

2h34960

Nathan Lambert@natolambert

@Miles_Brundage Happy to be part of this crew

Miles Brundage@Miles_Brundage

Still lots of alpha in not being at the top few companies - we need independent takes on this stuff now more than ever

2h67380

keithofaptos@keithofaptos

Thirdly, open source Agent systems. This is the path to reducing the open source open weighted sparse models size while increasing each expert's intelligence. If the tap got turned off this minute there's already enough to get this done. It's already happening. Everyday more pieces of code for every part of what I've listed are being shared. There's upwards of 10 million people globally working on or in Ai industry wide now. China is going outta their way to share as fast as possible now that DC has decided to directly involve themselves in US frontier models. 17hrs ago DS : https://github.com/deepseek-ai/DeepSpec https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-DSpark

22h21

Olivia P. Walker@olivia_p_walker

@natolambert Regulatory capture is never a good thing. @grok, define “regulatory capture” for those who don’t know.

20h1591

Ralph Cerchione@Dry_Observer

@natolambert I killed off distillation attacks this morning.

And then informed Anthropic.

Which should make the business models of Chinese labs relying on them far more exciting.

23h490

Beyang@beyang

@natolambert Thank for speaking up, Nathan. The world needs more voices speaking out against regulatory capture.

18h4584

Slop to Signal@SlopToSignal

@natolambert the 'you only believe that bc it benefits you' gotcha is so lazy

bro turned down money to make the point, thats literally the opposite of the argument

19h4494

Cobalt@cobaltdigital33

@natolambert i respect the work you do. and you have a lot of great takes. but please avoid letting your bias lead you towards reducing complex issues into sweeping binary statements.

22h351

keithofaptos@keithofaptos

Pandora isn't able to go back into the Box(es). If you or anyone else wants them to, ever. Period. Secondly, model councils, distillation, compression, meta routing, tier routing, Gödel machines, autoresearch, and a few others... already get the world to AGI, RSI, and intelligence and IQ far beyond Mythos right now. And all you researching lifer's know it. You're not stopping China. They'll give it away and we'll eat it up. The BadActors will hack till the cows come home. Even if you could wave your wands and turn it all off this second. Everything I wrote above still happens. 🖖🏻

23h791

Dan McAteer@daniel_mac8

@natolambert Support it.

23h5293

Jesse@jtc_ai

@Dec0ySquid @natolambert Trillion dollar companies are inherently serious. You can't get to that scale without the market agreeing there is value there.

I also don't get the hate for Anthropic. If those in govt listened to them, we wouldn't be in this place we are now.

23h451

Cobalt@cobaltdigital33

@natolambert to clarify: i think that decisions (as mentioned in your earlier post) can be made for the sake of safety & regulatory capture.

handwaving one or the other seems wrong (& handwaving the prior is disrespectful to the folks at anthropic).

22h35