Ori Nagel ⏸️ @ygrowthco Twitter profile

Pinned Tweet

Ori Nagel ⏸️

@ygrowthco

5 months

🫡

Jules Terpak

@julesterpak

5 months

Stop taking these empty dopamine hits and get the fuck to bed

26

18

327

0

4

Last Seen Profiles

@GuiCowboytw

@vaveysss

@rsrotnet1

@TrTinV1

@HnanA2187

@AyMinEssFan

@biamaga1306

@farhana49269

@riae_official

@smlk3721

@limitfungsii

@vvaleria_85

@swdeayy

@RealGetadolidCF

@HamiyetGunak

@Romilso09774367

@PenzeysSpices

@miecchip

@UIUPeacocks

@pardoalcuadrado

@mnwf1484793

@Sosasea169551

@Passlovesweaty1

@fishballming0

@chanelfwjen

@Smegman7

@KaizenSOL

@EveBeloin_26

@Artobot

@Danielle__C84

@ricky16426717

@BlackfootSD55

@dezzside

@Kairindo1

@OmerOmer2036

@ricky16426717

Ori Nagel ⏸️

@ygrowthco

6 months

FKA Twigs tries to knock some sense into Congress about passing AI DeepFake protections.

9

375

3K

Ori Nagel ⏸️

@ygrowthco

5 months

Yann Lecun has no code of ethics. Here he is in 2008 saying he'll take money from anyone, as long as they let him publish.

215

96

935

Ori Nagel ⏸️

@ygrowthco

7 months

Did everyone just decide to collectively ignore this latest position from @geoffreyhinton ?

32

47

424

Ori Nagel ⏸️

@ygrowthco

5 months

TL;DR: Sam Altman lied to the Board at multiple times and was accused of "psychological abuse." •⁠ ⁠The board found out about ChatGPT being released on Twitter. •⁠ ⁠Altman gave them "inaccurate information about the small number of formal safety processes" it had in

36

61

403

Ori Nagel ⏸️

@ygrowthco

5 months

OpenAI Co-Founder: We have a Slack channel dedicated to the welfare of Artificial Intelligence. "It is conceivable that through some kinds of trainings we could generate immense amount of suffering. Like massive genocides." -Wojciech Zaremba, OpenAI Co-Founder ( @woj_zaremba )

33

63

329

Ori Nagel ⏸️

@ygrowthco

4 months

Wild times rn in the #OpenAI slack channels

11

19

240

Ori Nagel ⏸️

@ygrowthco

4 months

Leopold Aschenbrenner was fired for telling the board that OpenAI was too vulnerable to cyberattacks from China. HR said it was "racist to worry about CCP espionage." After he voiced his concern, OpenAI rummaged through his digital artifacts, and found a memo he shared

30

39

219

Ori Nagel ⏸️

@ygrowthco

6 months

Geoffrey Hinton calls on regulators to ban Open Source AI models.

64

45

189

Ori Nagel ⏸️

@ygrowthco

5 months

@futuristflower Lmao Now make it detect every time he lies in video

4

0

171

Ori Nagel ⏸️

@ygrowthco

6 months

“Put the power in the hands of the artist”

brian cheung

@bcheungz

6 months

did not expect to see @FKAtwigs on @cspan today but damn that was a fire way to describe what’s at stake with AI

1

125

530

1

8

147

Ori Nagel ⏸️

@ygrowthco

5 months

@tszzl Maybe consider making this your cover photo, just saves time lol “It is difficult to get a man to understand something when his salary depends on his not understanding it.” ~ Upton Sinclair

1

144

Ori Nagel ⏸️

@ygrowthco

5 months

@tszzl The Roon analysis handbook: 1 Everything Sam Altman shits is gold 2 Invoke spiritual reference to defend such shit, if necessary 3 Be nebulous about assertions 4 Mix in philosophy and shitposts occasionally to muddy up the ramblings with actual insider insights 📖🥸

3

2

84

Ori Nagel ⏸️

@ygrowthco

3 months

Geoffrey Hinton is participating in an upcoming AI documentary because of his concerns over x-risk.

3

10

83

Ori Nagel ⏸️

@ygrowthco

6 months

So the US Gov’t just appointed a leader to its AI Safety Institute who estimates a 50/50 chance a godlike AI will kill him. Wtf mate?

Liron Shapira

@liron

1 year

“I think [superintelligent AI] is the single most likely reason that I will personally die.” —Dr. @PaulFChristiano , PhD Inventor of @OpenAI 's Deep Reinforcement Learning from Human Preferences (RHLF)

5

18

115

22

7

70

Ori Nagel ⏸️

@ygrowthco

6 months

When scientists like @geoffreyhinton call for serious AI regulation, builders like Perplexity’s CEO just laugh in their face.

22

3

62

Ori Nagel ⏸️

@ygrowthco

6 months

Geoffrey Hinton accuses OpenAI of compromising its values. @sama is choosing profit over safety.

8

11

58

Ori Nagel ⏸️

@ygrowthco

8 months

This is fine.

5

11

56

Ori Nagel ⏸️

@ygrowthco

4 months

Joseph Gordon-Levitt ( @hitrecord ) says Donald Trump won't do sh*t to handle AI issues.

13

4

55

Ori Nagel ⏸️

@ygrowthco

4 months

“We must win the AI race against China!” You might wanna have a word with China’s chief accelerator, @ylecun

9

4

55

Ori Nagel ⏸️

@ygrowthco

2 months

A consensus is starting to emerge that LLMs will NEVER be safe enough for society's requirements: Stuart Russell: "Impossible" to meet quantifiable safety guarantees. @GaryMarcus : "Unlikely ever to be safe." @ylecun : "Auto-regressive LLMs are intrinsically unsafe."

10

11

51

Ori Nagel ⏸️

@ygrowthco

4 months

Yann Lecun: Chief AI S̶c̶i̶e̶n̶t̶i̶s̶t̶ Evangelist

6

3

51

Ori Nagel ⏸️

@ygrowthco

2 months

GPT-5 Coming soon...

6

7

51

Ori Nagel ⏸️

@ygrowthco

5 months

@elonmusk Is it surprising he has little concern for the impact of his research, as long as it boosts his publishing metrics?

15

1

49

Ori Nagel ⏸️

@ygrowthco

3 months

@ai_ctrl Uhh is there more William has to tell us? 👻

2

1

48

Ori Nagel ⏸️

@ygrowthco

5 months

What did Daniel Kokotajlo see? 👀

7

4

45

Ori Nagel ⏸️

@ygrowthco

5 months

@nlpnyc 👀

3

2

42

Ori Nagel ⏸️

@ygrowthco

5 months

@virgileblais @ylecun Well the question is really the application of his research. Here, when challenged that his research will be used nefariously, he says “good point” and that his only criteria to determine what’s ethical is whether the research can be published. He’ll take money from any funder.

8

2

40

Ori Nagel ⏸️

@ygrowthco

2 months

This is an emergency we need to save Fei Fei's unicorn

Fei-Fei Li

@drfeifei

2 months

I’m deeply concerned about California’s SB-1047, Safe and Secure Innovation for Frontier Artificial Intelligence Models Act. While well intended, this bill will not solve what it is meant to and will deeply harm #AI academia, little tech and the open-source community.

83

319

2K

4

0

40

Ori Nagel ⏸️

@ygrowthco

5 months

"So Dario, why should we trust you?" Judge @Anthropic by our promises and commitments, he says. "The proof is in the pudding." Let's grade the pudding: ❌ Vowed not to escalate race dynamics, pushed SOTA AI with Claude Opus. ❌ Promised pre-deployment regulatory oversight,

7

2

38

Ori Nagel ⏸️

@ygrowthco

4 months

Don’t read Leopold Aschenbrenner’s 165-page treatise on AI if you hate glaring plot holes

3

2

36

Ori Nagel ⏸️

@ygrowthco

5 months

Source:

Wojciech Zaremba — What Could Make AI Conscious?

Wojciech joins us to talk the principles behind OpenAI, the Fermi Paradox, and the future stages of developments in AGI.---Wojciech Zaremba is a co-founder o...

www.youtube.com

1

4

37

Ori Nagel ⏸️

@ygrowthco

3 months

A sobering quote from Sam Altman. How do you hear this and not become concerned about OpenAI employees' projections that AGI will arrive in three years?

11

4

34

Ori Nagel ⏸️

@ygrowthco

6 months

AI Safety Hype vs Reality

5

32

Ori Nagel ⏸️

@ygrowthco

3 months

I'm not that superstitious but is this some kind of message or..?

ControlAI

@ai_ctrl

3 months

Ex-OpenAI safety researcher William Saunders likens OpenAI to the Titanic. "I really didn't want to end up working on the Titanic of AI, and so that's why I resigned."

32

123

763

3

1

31

Ori Nagel ⏸️

@ygrowthco

5 months

"The path to AI doomsday is paved with minor AI incidents." Denialists claim that existential risk is unsupported, based on hypotheticals. Wrong. The AI doom scenario is based on a series of arguments, which are increasingly supported by evidence of small-scale AI incidents.

5

3

31

Ori Nagel ⏸️

@ygrowthco

4 months

@ylecun Is this what you say when esteemed scientists like @tegmark share similar emotional reactions?

Liron Shapira

@liron

1 year

"You know, we just had a little baby, and I keep asking myself... how old is he even gonna get?" — @Tegmark

3

6

61

3

0

29

Ori Nagel ⏸️

@ygrowthco

5 months

@MelindaBChu1 I have zero connection to that. I’ve seen your activity round these parts and I don’t think you add much value to the discourse so unless you choose to comment in good faith I don’t plan to engage with you.

3

0

28

Ori Nagel ⏸️

@ygrowthco

6 months

How will right-wing folks reconcile these conflicting views?

27

0

28

Ori Nagel ⏸️

@ygrowthco

5 months

@modelsarereal No

1

0

24

Ori Nagel ⏸️

@ygrowthco

6 months

@andybrandt Geoffrey Hinton is retired, he’s not a builder at a company that stands to profit from AI regulation

5

0

25

Ori Nagel ⏸️

@ygrowthco

6 months

@deandraker @Grimezsz They asked @Grimezsz to testify at the Senate but she insisted that @GRIMES_V1 represent her instead

0

25

Ori Nagel ⏸️

@ygrowthco

2 months

@ashleevance In return I hope Elon's paying you pretty well

5

0

25

Ori Nagel ⏸️

@ygrowthco

3 months

The guy who wrote the textbook on AI says that it's *IMPOSSIBLE* to guarantee safety with the current technology direction: Deep Learning and Transformers Impossible to align.

2

1

24

Ori Nagel ⏸️

@ygrowthco

3 months

Stuart Russell, author of the go-to AI textbook, says: Solving alignment with current AI tech is "hopeless." It's not about finding a solution to the 'King Midas' problem, it's worse. Because right now we can't even specify an AI's objectives.

0

4

23

Ori Nagel ⏸️

@ygrowthco

5 months

@sama Sam Altman in 2015: use equity to keep OpenAI employees quiet after they leave. Sam Altman in 2024: use equity to keep OpenAI employees quiet after they leave.

0

22

Ori Nagel ⏸️

@ygrowthco

4 months

Jakub Pachocki, OpenAI's Chief Scientist, couldn't even keep his account safe from crypto scammers. He replaced Ilya Sutskever, who led superalignment efforts to ensure AGI safety. Sam Altman promised Jakub would lead "rapid and safe progress." Rapid? Maybe. Safe? 🤷‍♂️

3

4

22

Ori Nagel ⏸️

@ygrowthco

11 months

@neil_chilson Maybe the best thread I’ve seen yet, have a like my good fellow

1

0

20

Ori Nagel ⏸️

@ygrowthco

5 months

Geoffrey Hinton agrees with @hlntnr and Tasha McCauley.

Ori Nagel ⏸️

@ygrowthco

6 months

Geoffrey Hinton accuses OpenAI of compromising its values. @sama is choosing profit over safety.

8

11

58

0

2

20

Ori Nagel ⏸️

@ygrowthco

4 months

“If I paid you a million dollars, would you agree to never say anything bad about me?” - Sam Altman

Liron Shapira

@liron

4 months

Confirmation that former OpenAI employees @DKokotajlo67142 and @leopoldasch each sacrificed $1M+ for the ability to speak out.

11

13

113

0

2

19

Ori Nagel ⏸️

@ygrowthco

4 months

@DKokotajlo67142 Thank you for voicing your concerns and empowering others to do so as well!

0

19

Ori Nagel ⏸️

@ygrowthco

6 months

"You don't have to be a scientist or engineer to have a voice." Former OpenAI Director Helen Toner argues that technologists shouldn't gatekeep AI policy conversations. After all they don't even fully understand the technology themselves.

2

18

Ori Nagel ⏸️

@ygrowthco

5 months

Sam Altman is a fan of the Democratic process (when he tips the scale in his favor).

1

0

17

Ori Nagel ⏸️

@ygrowthco

4 months

@MichaelTrazzi @ylecun owes Dan another apology.

Yann LeCun

@ylecun

1 year

@DanHendrycks I actually find your ability to break away your religious upbringing quite admirable. I wish more people could do what you and Yud were able to do. My tweet was really not an attack, but a question about a possible cultural phenomenon.

10

1

82

0

17

Ori Nagel ⏸️

@ygrowthco

5 months

Congratulations to @sama & team on a perfectly-timed announcement! Launching GPT-4o concurrently with global protests against AGI ... Must have been a 5D chess move to push for responsible governance.

3

1

17

Ori Nagel ⏸️

@ygrowthco

5 months

@saliozzia Sorry does the truth bother you? I don’t tolerate anti media BS, so if you bring me that again I’m gonna block

4

0

17

Ori Nagel ⏸️

@ygrowthco

5 months

@dustdevvil @woj_zaremba Not exactly a meme but here would be one contribution to that

0

15

Ori Nagel ⏸️

@ygrowthco

4 months

@TheZvi If only Joey would listen!

1

0

16

Ori Nagel ⏸️

@ygrowthco

2 months

@tegmark @OpenAI @jasonkwon @sama

0

16

Ori Nagel ⏸️

@ygrowthco

6 months

@__HMYS__ Yes, my point is it’s not like he’s some quack, he’s now essentially our country’s top AI safety official, which further legitimizes his concerns.

3

0

15

Ori Nagel ⏸️

@ygrowthco

5 months

@virgileblais @ylecun No they didn't use the word 'nefarious', doesn't change the meaning of my paraphrasing. So pedantic, at least try to argue in good faith. His only criteria, that he avows to in this clip, to evaluate the application of his research is whether it funds him and can be published.

4

0

14

Ori Nagel ⏸️

@ygrowthco

5 months

@ylecun @amdaleynews Therein lies the problem. Isn’t this short-sighted thinking? There are cases where that body of knowledge crosses ethical or safe lines, eg gain-of-function research. It’s unrealistic to expect your research to live in a vacuum.

1

0

13

Ori Nagel ⏸️

@ygrowthco

4 months

Per CNBC:

1

14

Ori Nagel ⏸️

@ygrowthco

8 months

Reply guys: chill bro he’s joking! Sam Altman:

0

14

Ori Nagel ⏸️

@ygrowthco

5 months

Lol what does that even mean? It’s one thing for @sama to drop this platitude in a speech but when you see it in official statements its superficiality is laid bare.

0

14

Ori Nagel ⏸️

@ygrowthco

5 months

Is this real or just a trailer for a “Silicon Valley” reboot??

1

13

Ori Nagel ⏸️

@ygrowthco

5 months

@BasedBeffJezos Haha if you actually believe this Then I actually feel bad for you Like concern for your mental health

1

0

13

Ori Nagel ⏸️

@ygrowthco

3 months

@ESYudkowsky Stuart Russell now is agreeing with the position on the intractability of alignment with current tech

Ori Nagel ⏸️

@ygrowthco

3 months

The guy who wrote the textbook on AI says that it's *IMPOSSIBLE* to guarantee safety with the current technology direction: Deep Learning and Transformers Impossible to align.

2

1

24

1

0

13

Ori Nagel ⏸️

@ygrowthco

4 months

@martin_casado @CommunityNotes The author of this post has now admits he "misspoke" about a key claim in this post. "Although advocates for Al safety guidelines often allude to the "black box" nature of Al models, where the logic behind their conclusions is not transparent, recent

0

13

Ori Nagel ⏸️

@ygrowthco

5 months

Both are happening on Monday at 10am at OpenAI. Which do you choose?

4

1

12

Ori Nagel ⏸️

@ygrowthco

6 months

Source:

1

12

Ori Nagel ⏸️

@ygrowthco

5 months

@a__douris @soniajoseph_ @techfounder3 What do you expect her to do, share pictures? Show you the orgy signup sheet with (surprise!) some OpenAI exec’s name on it? This is a matter of taking her at her word and she is an AI researcher, not some anon. The only thing you can do Sonia to get people to believe you, I

1

0

12

Ori Nagel ⏸️

@ygrowthco

6 months

@ro____ha__ @geoffreyhinton @AravSrinivas Arav thinks he’s on top of the world and totally oblivious to any harms he might be party to

1

0

10

Ori Nagel ⏸️

@ygrowthco

6 months

@OpenMedFuture @andybrandt So you think the wealthy and renowned "godfather of AI" decided to retire from his high-paying corporate job and tell the world that he regrets his life work in order to make money off of his equity shares?

2

0

12

Ori Nagel ⏸️

@ygrowthco

6 months

Source:

Professor Geoffrey Hinton is presented with the UCD Ulysses Medal

Professor Geoffrey Hinton was presented with the UCD Ulysses Medal by UCD President Professor Orla Feely during a ceremony at UCD O’Reilly Hall on Monday 8 A...

www.youtube.com

1

11

Ori Nagel ⏸️

@ygrowthco

5 months

@MichaelTontchev @demishassabis There’s little willingness among industry players to slow down the capabilities race, despite the lip service. That’s why such a move to shift the political conversation seems like the most meaningful action imo. If there are “safety-minded” people at AI labs right now, their

0

Ori Nagel ⏸️

@ygrowthco

5 months

@MelindaBChu1 I’m not an EA and i frankly don’t appreciate you accusing me something that I’m not. Not only is it unhinged, it’s just straight up lying. Now I’m giving you 1 more chance to not act like a loony or I’m blocking you.

1

0

10

Ori Nagel ⏸️

@ygrowthco

5 months

@AISafetyMemes Is each red flag an employee quitting for mysterious reasons? Like first Dario, then the Sam firing, Saunders, koko et al

1

0

10

Ori Nagel ⏸️

@ygrowthco

6 months

Eric Schmidt: China is exploiting open-source AI models to boost their capabilities.

4

1

9

Ori Nagel ⏸️

@ygrowthco

6 months

Beff Jezos: e/acc is not right-wing, it's apolitical. Also Beff Jezos: e/acc is a movement opposed to industry regulation & Big Gov't calling for the right to "bear arms" against data control, that believes in American exceptionalism with high alignment with Christian values.

4

1

11

Ori Nagel ⏸️

@ygrowthco

5 months

Sir @demishassabis always says the right thing about AI Safety. What he does is a different story. Here's his pitch to prevent runaway AI race dynamics, while his team races full speed ahead. 👏👏

2

1

11

Ori Nagel ⏸️

@ygrowthco

1 month

I'm getting really eerie vibes from these Don't Look Up deleted scenes

1

0

10

Ori Nagel ⏸️

@ygrowthco

2 months

@shiringhaffary He called it. The companies put "but it stifles innovation" on a tape-loop and somehow no one realizes the absurd bigger picture unfolding:

ControlAI

@ai_ctrl

2 months

AI scientist Stuart Russell says the AI industry's position on regulation is pathetic: "There are far more rules on sandwiches than there are on software ... and these trillion dollar companies will say 'Oh we can't fill out a form!'".

11

55

205

1

0

10

Ori Nagel ⏸️

@ygrowthco

2 months

Is it me or has something changed about Fei-Fei Li recently?

1

0

10

Ori Nagel ⏸️

@ygrowthco

5 months

I'm sorry, brother. If you think the forthcoming birth of a smarter digital species isn't a scary proposition, maybe you haven't thought it through enough. You might be in denial. Come to our NEXT MEETING on Sunday, May 19 at 11:00 AM PT.

2

4

10

Ori Nagel ⏸️

@ygrowthco

5 months

@virgileblais @ylecun I'm sorry I said one word that wasn't exactly in the text. If you say you're discussing in good faith, then I think you need to up your game, because it's poor form on your part to try to discount my position based on the way I paraphrased it rather than addressing the argument

3

0

10

Ori Nagel ⏸️

@ygrowthco

2 months

@tsarnick woah there's gonna be a whole new crop of sama wannabes now aren't there

1

0

10

Ori Nagel ⏸️

@ygrowthco

5 months

*Alright I’ll just call it: Beff is delusional.* He actually believes the conscientious objectors who resigned from OpenAI are playing 4D chess to make a profit. It’s a pathological grade denial. It should have been obvious from the get-go that his only contribution is

Beff – e/acc

@BasedBeffJezos

5 months

1) Come in as AI safety expert 2) learn the ways of OpenAI 3) make valuable connections 4) leave to start your own shop 5) if anybody asks, cite "AI safety concerns" rather than greed as the primary reason

42

28

291

1

10

Ori Nagel ⏸️

@ygrowthco

4 months

Watch this and ask yourself again why @sama went on that massive world tour in 2023.

ControlAI

@ai_ctrl

4 months

On the Dwarkesh podcast, ex-OpenAI safety researcher Leopold Aschenbrenner said he heard that at some point OpenAI leadership had discussed starting a bidding war to fund and sell AGI between the US, Russian, and Chinese governments. We should be skeptical of those who would pit

5

33

2

0

10

Ori Nagel ⏸️

@ygrowthco

4 months

@KelseyTuoc If only they would listen…

1

0

9

Ori Nagel ⏸️

@ygrowthco

4 months

@BasedNorthmathr @PauseAI “The opinions of experts don’t concern me” At least you’re honest about it lol

2

0

9

Ori Nagel ⏸️

@ygrowthco

7 months

@eshear In the parlance or anxiety therapy the self-judgment is a “safety behavior”, because as you rightly observed it removes the discomfort of the question and it also gives a mild positive reward knowing you’re right.

2

0

8

Ori Nagel ⏸️

@ygrowthco

5 months

@GaryMarcus A real life version of Samantha from Her just came out. How can you say in good conscience that nothing “mind blowing is being released anymore”?!

2

1

9

Ori Nagel ⏸️

@ygrowthco

2 months

🥴 ^e/accs rn

0

9

Ori Nagel ⏸️

@ygrowthco

2 months

ControlAI

@ai_ctrl

2 months

OpenAI Chief Strategy Officer Jason Kwon (last week): "We've always believed that AI should be regulated, and that commitment remains unchanged." Jason Kwon (this week) to legislators: California's draft AI bill "would threaten that growth, slow the pace of innovation ..."

3

8

49

0

9

Ori Nagel ⏸️

@ygrowthco

19 days

2023: Sam Altman claims no financial motive for his OpenAI role. *Spontaneous round of applause*

0

9

Ori Nagel ⏸️

@ygrowthco

29 days

This is the real GPT-o1 preview you need to see

Ori Nagel ⏸️

@ygrowthco

2 months

GPT-5 Coming soon...

6

7

51

0

2

9

Ori Nagel ⏸️

@ygrowthco

4 months

@danfaggella It’s not the first time, he previously tried to discredit Dan Hendrycks for his religious upbringing (!)

2

1

8

Ori Nagel ⏸️

@ygrowthco

21 days

Much respect to these legends for taking a stand on AI. 💪Keep up the great work💪 @hitRECordJoe @PiperPerabo @MarkRuffalo @seanastin

0

8

Ori Nagel ⏸️

@ygrowthco

3 months

@ylecun Saying they can’t plan is patently false when the cited study shows it semantically plans according to the benchmark 30%+ of the time.

5

0

7

Ori Nagel ⏸️

@ygrowthco

7 months

@Nemtastic1 @AISafetyMemes I've heard that argument. It doesn't make sense though. The best way to manage a lethal technology is to... keep building it? Geoffrey Hinton is a great model of how to do good operating outside of industry. Gaurantee you it would be highly newsworthy and potentially

1

8