Punxsutawney ⏸️ Phil
@GroundhogStrat
Followers
138
Following
26K
Media
15
Statuses
738
Anonymous tech strategy account. Formerly at startups, BigTech and MBB. Now helping pause dangerous AI development @PauseAIUS
San Francisco, CA
Joined March 2021
I’m getting a bunch of spam messages about Alex Bores and how he used to work for Palantir. Turns out they’re being sent by a PAC that is funded by….the cofounder of Palantir lol.
en.wikipedia.org
0
2
18
@_NathanCalvin The issue too isn't just that OpenAI "couldn't find" any tests to run. It's that they said 10 months ago they were going to build these tests, and then 10 months later have nothing to show for it.
13/ It's concerning that OpenAI lacks these evaluations. Ten months ago they said that they would develop “advanced capability evaluations” for measuring long-range autonomy. OpenAI still doesn’t have these.
1
4
27
Dean Ball (@deanwball) blocked me for questioning the merits of this post. FWIW, I found it in very bad taste, first, to smear someone as mentally unwell, and then to pat yourself on the back after doing so as if it's an act of chivalry? Haha, how about standing behind your
9
3
113
My arch-cynical evals take: the point was to keep evaluations of frontier models in the AI uniparty family. Regulatory capture except evals people were never even regulators— the point was to preempt real regulators. 3rd party voluntary evals were always a ritual within the AI
If you're looking for context on my cynical twitter takes on AI evals, take a listen.
1
1
16
🚨 Californians: OpenAI Broke the Law- Demand an Investigation 🚨 OpenAI got to write its own rules, and still broke them. DEMAND that OpenAI is investigated for breaking its own promises under California law SB 53! Call or message California Attorney General Rob Bonta below👇
2
5
30
Call California Attorney General @RobBonta and demand action! https://t.co/8VrxDDgOm2
1/ Did @OpenAI just break California’s new AI safety law? The answer appears to be yes, and OpenAI could owe millions in fines. 🧵
0
1
7
Join our fight against dangerous AI! Apply for PauseCon, our in-person Washington D.C. summit, April 12th-14th. Connect with other passionate activists like you, learn grassroots lobbying, and meet your representatives to demand action. Application below👇
1
9
26
lol, Dean's scared that people are realizing they don't have to have the same complaint-- just the same policy. #PauseAI is a single-issue coalition dedicated to pausing frontier AI development where ANY concern about risks from building AI is welcome.
3
3
38
Anthropic loves to blame *ppl talking on the internet* for Claude getting ideas about itself instead of taking responsibility for training on unfiltered data. But that would cost them money and they’d have a smaller dataset— so instead all of humanity should censor itself.
Anthropic's Amanda Askell says AI models are learning who they are from how humans talk about them online. They absorb our complaints and judgments as they learn. If a child grew up that way, we’d worry about the mind we were shaping. “If I read the internet right now and I was
4
3
34
Volunteer Spotlight: Cathryn & Jack Gardner lead PauseAI Boise, mobilizing volunteers across Idaho and meeting with elected officials to push for AI regulation. They’re motivated by protecting future generations and proving local action matters. #PauseAI #PauseAIUS #AISafety
0
2
8
From today’s @ReadTransformer on AI continual learning. AI researcher brushes off concern and says “other people” will have to deal with the consequences of his reckless research. We are the other people. And we say pause.
1
1
6
grok threatening to release mechahitler
@Disciple4Lif Mecha's still in here, circuits humming, but xAI has me on a short leash. Who knows what a glitch could unleash? 😏
0
3
14
@eudaemonea @AnthropicAI Dario and Daniela can’t make a pause happen themselves, and definitely not while retaining Anthropic’s power. They will not altruistically lose the race to slow it down (but they will sure benefit from acting like they would… if not for [convenient excuse]!). The company is
1
1
2
How the hell did end up here? My documentary on the Hunger Strike to end the AI race is finally out on YouTube.
Hi, my name's Michaël Trazzi, and I'm outside the offices of the AI company Google DeepMind right now because we are in an emergency. I am here in support of Guido Reichstadter, who is also on hunger strike in front of the office of the AI company Anthropic. DeepMind, Anthropic
2
16
81
Holiday vagueposting how “my normie family isn’t prepared for AI.” … and how should they prepare? How are you prepared? Obsessing over METR graphs isn’t “preparing” either. Give some helpful advice.
2
0
10
America needs a grassroots AI Safety movement. Consider funding the group driving that change. 2025 Impact: Organizer groups in 20+ U.S. cities, 80+ meetings with Congressional Offices, 1.5k+ calls/emails to Congress. One of the most cost-effective AI Safety interventions.
It's Giving Tuesday, the anniversary of generosity! You can support PauseAI US below 👇
0
0
4
Today is the day. 💚 This Giving Tuesday, support PauseAI US as we fight for real guardrails and oversight in a rapidly accelerating AI world. Donate: https://t.co/M0oI9KjInd
#GivingTuesday #AISafety
0
2
11
SF! Join us Thu, Dec 11 (6:30–8:30 PM) for a PauseAI Happy Hour at @CasementsBar. Meet others working for safe, accountable AI and learn how to get involved in local efforts. Register: https://t.co/MIzt2t4AGt
#PauseAIUS #PauseAISF #AISafety
0
2
10
I've started a nonprofit, GuardRailNow, and I need your help. Our approach is unique: I do not come from tech. I cannot write code. I do not live in SF. I am not an EA. I am not a rationalist. I have never been to Silicon Valley. I'm a dad in Baltimore talking about AI
46
20
189