fine-tuning cowboy 🤠// prev faang & startups in sf // practicing the art of vanity
Joined August 2013
- Tweets 9,926
- Following 995
- Followers 3,433
- Likes 37,614
785 Photos and videos
Pinned Tweet
Apr 8
gm
you can finetune a 9B qwen model with a single prompt, curate your own dataset with our generator & evaluator system, try it on our playground, and take it all home with you under 30 minutes. for free.
5 more seats. dm or comment if you're interested in trying today.
10
1
29
9,350
The problem @DavidSacks is that the Administration has a credibility gap on whether decisions are being made with political or retributive motives or for the national interest. Trump's mantra has been, until now, accelerationism and no regulation.
Why not work with Congress to set up an independent agency for AI safety, like the nuclear commission or FERC, so that the public has confidence in these decisions?
Jun 13
I’ve had a number of conversations with folks inside and outside government about the current situation with Anthropic, and here is what I believe to be true:
— As we know, Anthropic publicly released its Mythos class models earlier this week under the commercial name Fable.
— Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.)
— A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused.
— In their blog post, Anthropic defended its decision by saying the jailbreak isn’t serious. That is not what the trusted partner and the USG believe; nor is that kind of minimizing language consistent with Anthropic’s brand as the AI safety company. It’s difficult to fathom how they could claim a jailbreak allowing operability of a cyber weapon could be defined as not “serious.”
— In the past, Anthropic has always said that safety must be top priority and taken super seriously. In this case, Anthropic prioritized the continued offering of the consumer model over safety.
— In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community.
— The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release. The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority.
— Those trying to misdirect and tie this action to the prior DoW/Anthropic issues are wrong. The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court.
148
51
610
104,033
Jun 14
Dark precedence to set
Jun 14
Make no mistake: post-Mythos, the United States has a licensing regime for AI. It’s just informal, with no consistent rules or firm boundaries on state power or public transparency. Cobalt mining in the Congo is vastly more institutionalized than frontier AI licensing in the US.
6
226
Jun 14
I just wanna say the Knicks won right as I boarded off my plane and landed in nyc back at home babyyyyy
11
334
Jun 13
I need to expunge this account, make it corpo, and create an a new alt account
1
6
651
Jun 13
I wanna get into politics so badly
Jun 13
Wish you had unlimited token spend?
Come work for the IRS
3
6
11,156
Jun 13
literally the stupidest thing I’ve ever read just block him dye your hair and move on like the rest of us what the fuck
Jun 12
Dear @Aella_Girl,
I’ve heard a lot of people seeming to think that I reject you wholesale, or that I don’t approve of you for being a sex worker. I want to set the record straight about why I do not want to work with you and have made certain comments, and offer the chance for a dialogue if you want. I also want to apologize.
I don’t like you because of how my ex-husband, Ronny Fernandez, your plzdontkillus cofounder, would make sexual and romantic bids at you in front of me (presumably also in private) while we were in a monogamous relationship. This was his fault, and it contributed to our divorce, but because it caused me to resent you as well, I came to you about it, hoping to give you a chance to show you weren’t down with it. You were cold to me, dismissing my concerns with “he’s not my type.” At that point I realized we were not friends. It wasn’t your responsibility to stop him, but your priority was him and my feelings weren’t part of the equation.
When you’re in an abusive relationship, it’s easy to feel more angry and betrayed at the people around you than at the abuser. I felt angry at you and many others from the rationalist community, as well as general rationalist culture, for supporting him. It’s not your fault that I was in an abusive relationship. It was his fault. Even though I think my feelings are valid, especially when you opted not to help me when I reached out, I recognize that I’m transferring anger onto you that really belongs to him.
I saw you once say that I judged you for being a sex worker. I think you were remembering a time when I argued that, because Ronny had your nudes, you didn’t just have a platonic relationship. I have also critiqued the blurred line between your sex work and your intellectual content, where engaging with you without sex is still a form of flirtation and erotic connection, for a similar reason. My issue has never been with the sexuality. It had to do with how you and our mutual friends were insisting that I had no recourse against my partner cheating on me— that I was in the wrong for feeling jealous.
Perhaps you regret this, or would regret this now that you know how I felt. It would make a big difference to me if you did. There are also subject matter disagreements I have with you on AI Safety work, and I wrote a lot of them up, but upon introspection I think that discussion would be pretty collegial if I saw you have empathy for a wound in me that’s still healing. The breach of trust was really a personal thing.
I’m also writing to share my concern about your current relationship with Ronny. Ronny lovebombs you epically on main, and I think that could lead to epic devaluation and exploitation if you’re not careful. I have wondered if he’s pushed you to do and share more and more extreme things when I see him seeming to get dividends. I saw him negging you and undercutting you even when he and I were together, and he’s good at portraying that destruction of self-esteem as rationalist introspective virtue and fucky intimacy. No matter how annoyed I may feel at you, I would never want you to go through what he did to me.
I’m going to unblock you, and you can DM or reply if you wish.
Holly
2
41
5,837
Jun 13
JD Vance was literally saying we should release everything, he’s getting a lot of hatred for it. According to the NYT article, Susie Wiles was the biggest strongest internal advocate for the covering up
JD Vance. Susie Wiles. Karoline Leavitt.
Yesterday, Congress formally demanded all three testify about the White House's concealment of the Epstein files.
The Vice President is listed first.
ALT House Oversight Ranking Member Garcia's June 11 2026 letter to Chairman Comer listing JD Vance Susie Wiles and Karoline Leavitt as officials demanded to testify about Epstein file concealment
2
211
Jun 13
Deepseek can do the funniest thing rn
Jun 13
This is going to have an opposite effect that the decels want. It's a huge open source AI accelerant, an accelerant for corporations, and enterprises
Now it's a real race. You either make your own AI infrastructure or you don't have a seat at the table
7
496
zöda retweeted
Jun 12
we get baklava lattes and u can too
@falcongx_ is putting on a live design music set, limited merch drop, and more
alongside @saleh_digital and @telekineticpxls
invite link below
17
12
62
11,493
Jun 11
Massie is literally leading the movement where republicans have the courage to disagree with their party for what they believe in. incredible to see real time. it went from 4 republicans standing with him to 19 today.
Jun 11
19 Republicans withheld our votes in order to preserve Constitutionally guaranteed rights.
By refusing to honor the Fourth Amendment, you’re jeopardizing the continuation of FISA.
Include a WARRANT requirement for US citizens if you want Republicans to pass this bill.
1
135
Jun 11
I wonder if people will consider this a conspiracy theory in 10 years
Jun 11
IT'S OVER
The New York Times just dropped the BOMBSHELL
Trump raped kids with Epstein, and his administration engaged in a cover up to protect him
1
3
1,070
Jun 10
Is it actually completely reasonable and aligned with safetyai values to add strict guardrails to mythos line of models? Yes.
Would it have been preferable if they never released it to begin with? 1000% yes.
This pay to play mechanism is very inequitable & dangerous. They sold their soul
3
169
Jun 10
why is fable/mythos so sassy
3
288
Jun 10
genuine question to all my ai nerds have you guys heard of a strike? a union? you have so much control if we can form a cohesive collective. you’re actually not at the whims of the big lab leaders.
I know China developing AI while we stall is a real threat, but you could strike to make policy changes not literally pause AI.
DMS open if you wanna chat labor organization ❤️
Jun 9
Anthropic won't pause AI development.
This is a blatant lie unless they fire their whole army of AI engineers.
They'll keep working and give strong AI to huge corporations and government while arguing that it's too powerful for plebs to use.
Dangerous effective altruism.
8
328
Jun 10
I have to imagine that The Arts was a huge source of conversion and retention for Catholics. They are so beautiful, so magnificent, and very inspiring. You look at the heavens and think “is that what comes of being a follower of Christ”? I love medieval and renaissance art so much.
It’s a shame the modern world has such horrible brutalist, sterile, mechanistic aesthetics. It feels like forgotten practice.
2
6
401