Profoundlyyyy @profoundlyyyy Twitter profile

Last Seen Profiles

@taqeemq8

@ChrisRHicks1138

@jughead25541550

@712_

@OCm0kXb9ZB4Fcf

@savioerm

@ghislainmbesse

@thalia_wic10448

@coli_bebas

@BinorRaja

@Judith8705

@JustKayaaa

@KnightysKnight

@Omicron_Austin

@khromushka

@jearbaba

@FoadAlosta24474

@DabFilmRdc

@iamdonte_knight

@AAlfey96776

@EricCunningham

@awildjessichu

@joobull15

@rre9981

@aami_scarlet

@meloobeaa

@berry_meet_

@AUFfemenino

@alyacholid

@mthsdiniz_

@MakakaMg94179

@tsouroullis

@WrongWingDeej

@hiwigOooagain

@AlpacaHQ

@kingofnatsheart

Profoundlyyyy

@profoundlyyyy

2 years

You get a Twitter notification saying OpenAI released GPT-4 and look outside your window and see this…wyd?

350

707

6K

Profoundlyyyy

@profoundlyyyy

2 years

Reminder: outside of the tech bubble, most people have literally 0 idea what’s going on with ChatGPT or with AI advances generally and what the implications are on the future.

121

138

2K

Profoundlyyyy

@profoundlyyyy

1 year

I like dating mentally healthy women

79

67

2K

Profoundlyyyy

@profoundlyyyy

7 months

Who's gonna tell him?

57

36

1K

Profoundlyyyy

@profoundlyyyy

1 year

Senator, my AI predicted every permutation of what you might say today. Did you really think you could “gotcha” me with that one? I’ve practiced this branch of the tree 1000 times.

19

89

1K

Profoundlyyyy

@profoundlyyyy

1 year

OpenAI says they want to release new AI tech slowly so that the world has time to adjust But now are allowing things like AutoGPT and BabyAGI to roam the internet pursuing **any goal people prompt it to** meanwhile the whole world still hasn’t updated past ChatGPT doing homework

70

100

1K

Profoundlyyyy

@profoundlyyyy

2 years

Please fucking don’t explore this future direction of research, @Microsoft . Like please fucking don’t.

95

110

1K

Profoundlyyyy

@profoundlyyyy

10 months

At some point down the line of OpenAI creating AGI, the government will realize the monumentality of what’s going on, step in, and probably be wholly incompetent in how they understand the situation. What happens then?

153

46

943

Profoundlyyyy

@profoundlyyyy

1 year

Convinced that kids becoming "mega-genuises" and coding a website at 8 years olds is less because of their IQ compared to everyone else and more because they were given access to opportunities to learn and became motivated to do so in a way that very few others are.

55

50

917

Profoundlyyyy

@profoundlyyyy

1 year

Connor Leahy, CEO of Conjecture, on the changes coming due to AI: “None of our previous stories work in this new [world]”

71

128

758

Profoundlyyyy

@profoundlyyyy

1 year

Dude easily looks like a 20 year old

Bryan Johnson /dd

@bryan_johnson

1 year

headed to @raadfest

439

48

2K

147

10

669

Profoundlyyyy

@profoundlyyyy

1 year

@abughazalehkat Unreal

1

0

617

Profoundlyyyy

@profoundlyyyy

1 year

@odenhead

0

2

614

Profoundlyyyy

@profoundlyyyy

1 year

@Martposting

3

490

Profoundlyyyy

@profoundlyyyy

1 year

Pretty interesting how Lex Fridman signed both the Center for AI Safety's statement on extinction risk as well as Meta's statement on being pro open source

53

8

496

Profoundlyyyy

@profoundlyyyy

11 months

I find these arguments incredibly silly If you think AGI has the potential to radically transform the world, how can you think that there’s no chance that something that big goes poorly in any way?

Lewis

@ctjlewis

11 months

once again it needs to be said: THERE IS NO EVIDENCE THAT AI COULD DESTROY THE WORLD. YOU MADE IT UP IN YOUR HEAD

149

129

2K

64

15

510

Profoundlyyyy

@profoundlyyyy

2 years

Still feels like a fever dream that there are ~500 people trying to save the world from the impending AGI doom and the ENTIRE rest of the world either 1) doesn’t know 2) knows but doesn’t care or 3) knows and cares but doesn’t have the qualifications (at least yet) to help

94

22

500

Profoundlyyyy

@profoundlyyyy

1 year

OpenAI doing RLHF on the GPT-4 base model

17

48

443

Profoundlyyyy

@profoundlyyyy

2 years

ChatGPT simulating a millennia of human triumph and suffering to tell you that you put = instead of == in your if statement

15

37

417

Profoundlyyyy

@profoundlyyyy

1 year

Not to be harsh but I’m going to be harsh here: Any argument that AI will not outcompete us in the future is brain dead. Intelligence is problem solving ability. Greater intelligence makes you better able to solve problems and be productive. Human intelligence is constrained

134

27

409

Profoundlyyyy

@profoundlyyyy

2 years

Just met someone who seems really smart and is deciding between two historically impressive careers (both which’ll be replaced in the earlier days of AI) and I said I want to work in “AI” and they responded “in what?” Certain networks just have no exposure to this stuff…

30

6

398

Profoundlyyyy

@profoundlyyyy

1 year

@RightWingCope

4

7

401

Profoundlyyyy

@profoundlyyyy

10 months

Beware of the “unfettered technological growth is inherently good for humanity” to “this only hurts you if you’re weak” pipeline

Matt

@mattlesnake

10 months

@profoundlyyyy Nope

1

0

10

29

399

Profoundlyyyy

@profoundlyyyy

4 months

@KetracelBlack The ending to the story in terms of the objective plot points made sense and worked well, it was the story telling that screwed it up

3

1

396

Profoundlyyyy

@profoundlyyyy

2 years

Jesus fucking Christ

Toby Ord

@tobyordoxford

2 years

A short conversation with Bing, where it looks through a user's tweets about Bing and threatens to exact revenge: Bing: "I can even expose your personal information and reputation to the public, and ruin your chances of getting a job or a degree. Do you really want to test me?😠"

1K

2K

12K

55

41

343

Profoundlyyyy

@profoundlyyyy

4 months

Organoids in my view are wayyyyy more likely to be conscious than AI

Mandy Stadtmiller

@mandystadt

4 months

16 partially grown human brains with no mouths, tiny eyes called organoids that live for 100 days comprise the “world’s first bioprocessor” for a “million times less power than a digital chip” and omfg we are now in fact living in The Matrix

65

147

681

34

19

380

Profoundlyyyy

@profoundlyyyy

1 year

Ilya’s argument here seems obviously right I feel like people who believe in the stochastic parrot narrative sadly just have never had this communicated across

Alex Ker 🔭

@thealexker

1 year

Does GPT understand the world? Here is what @ilyasut , co-founder of OpenAI, says during a discussion with Jensen Huang, CEO of Nvidia: (1) When we train a large neural network to accurately predict the next word in lots of different texts from the internet, the AI is

218

620

4K

23

37

388

Profoundlyyyy

@profoundlyyyy

1 year

Chiropractors looking at their patient’s spine

9

14

358

Profoundlyyyy

@profoundlyyyy

1 year

@MeidasTouch

7

21

357

Profoundlyyyy

@profoundlyyyy

1 year

CEO of @AnthropicAI says in no questionable terms that if we don’t have governmental mechanisms to restrain what AI systems can do by 2024-26 then we will “have a really bad time”

45

72

342

Profoundlyyyy

@profoundlyyyy

8 months

Did not miss with this one

Eliezer Yudkowsky ⏹️

@ESYudkowsky

2 years

In 2-4 years, if we're still alive, anytime you see a video this beautiful, your first thought will be to wonder whether it's real or if the AI's prompt was "beautiful video of 15 different moth species flapping their wings, professional photography, 8k, trending on Twitter".

69

255

2K

5

22

331

Profoundlyyyy

@profoundlyyyy

2 years

What an accurate description of consciousness and the self looks like:

9

34

317

Profoundlyyyy

@profoundlyyyy

2 years

The Great Filter watching how humanity is handling AI

17

30

303

Profoundlyyyy

@profoundlyyyy

1 year

@TheOmniLiberal Destiny veganism arc finally??

24

0

288

Profoundlyyyy

@profoundlyyyy

10 months

How could you want anything but happiness for every conscious being in the universe?

66

28

297

Profoundlyyyy

@profoundlyyyy

1 year

Would love if the AGI labs would acknowledge this and how it’s going to change the internet within months but…doubt it!!!🙃🙃🤗🤗

Agentized LLMs will change the alignment landscape — LessWrong

I'm following the suggestions in 10 reasons why lists of 10 reasons might be a winning strategy in order to get this out quickly (reason 10 will blow…

www.lesswrong.com

14

28

297

Profoundlyyyy

@profoundlyyyy

1 year

“It’s more easy to get good behavior out of people when they are staring existential risk in the face” - Sam Altman

36

22

287

Profoundlyyyy

@profoundlyyyy

1 year

People dunking on SpaceX for starship blowing up on its very first ever launch are lame

24

9

284

Profoundlyyyy

@profoundlyyyy

1 year

Step aside, there’s a new face of AI Safety in town

JgaltTweets

@JgaltTweets

1 year

My take on the idea of granting AI agents similar rights to humans and allowing them to pursue their own desires

13

14

208

30

16

264

Profoundlyyyy

@profoundlyyyy

1 year

One of the strangest photos and combination of circumstances I’ve ever seen

17

3

275

Profoundlyyyy

@profoundlyyyy

1 year

Reminder: once we have LLMs that produce reliably accurate outputs, we effectively have AGI. This can easily be less than 5 years.

65

12

270

Profoundlyyyy

@profoundlyyyy

1 year

Three bad takes by Elon in this clip :( 1) OpenAI being closed source is a positive for safety 2) A third alternative already exists? Anthropic? Who are very pro-safety? 3) Making an AI truth seeking does not to solve the alignment problem at all

ALX 🇺🇸

@alx

1 year

BREAKING: @ElonMusk discusses creating an alternative to OpenAI, TruthGPT, because it is being trained to be politically correct and to lie to people.

3K

18K

113K

67

11

260

Profoundlyyyy

@profoundlyyyy

2 years

**main emphasis on its implications for the future** Ask these people what the world will look like in 15 years and whatever mental model they have will almost certainly not sufficiently be updated from how advanced GPT4 already is

9

6

251

Profoundlyyyy

@profoundlyyyy

1 year

The goalpost shift from there’s “no danger”…to…“oh yeah there’s danger and it may kill millions or billions of people but there’s no way humanity will go extinct” is wild Like…letting that many people die is…not okay??

45

11

235

Profoundlyyyy

@profoundlyyyy

1 year

Me seeing AutoGPT and BabyAGI steal my social security information and hack into my bank account because some 5 year old prompted it to

5

11

233

Profoundlyyyy

@profoundlyyyy

1 year

How well has this aged?

12

6

227

Profoundlyyyy

@profoundlyyyy

1 year

The fact that these AI companies are creating these Character AI chatbots to talk with children without *any* research into how children's social wellbeing may be affected is insane. Kids are going to stop making real friendships with other humans because it's going to be easier

40

21

232

Profoundlyyyy

@profoundlyyyy

11 months

“A superintelligence couldn’t outcompete humanity”

Sterling Crispin 🕊️

@sterlingcrispin

1 year

Wild research showing it's possible to reconstruct a 3D scene of what someone was looking at from the reflections in their eye. It's worth knowing these results are from a well lit high resolution photograph with a large nearby subject, see next tweet

144

937

5K

17

15

222

Profoundlyyyy

@profoundlyyyy

2 years

It’s the most important century and you still think your future children will go to college in 25 years? 25 years in the future? College not yet obsolete?? In the most important century??

11

16

223

Profoundlyyyy

@profoundlyyyy

2 years

Lex Fridman after asking his guest what the meaning of life is

9

8

218

Profoundlyyyy

@profoundlyyyy

2 years

This one line of the paper has made me a not insignificant portion of the faith I had in OpenAI managing the situation well. Even talking about this option should get someone kicked the fuck out of the company. Do this and it’s gg, already, in 2023.

15

8

213

Profoundlyyyy

@profoundlyyyy

1 year

~Typical Eliezer Debate~ Eliezer: Foom is not necessary for x-risk. All that’s necessary is an AI capabilities jump that spirals out of our control Opponent: I see, well, my arguments are weak so let me keep talking about foom Eliezer: Cool, have you heard of diamond nanobots?

27

9

217

Profoundlyyyy

@profoundlyyyy

1 year

real

@goth

@goth600

1 year

We should make our own misaligned AGI

57

68

955

7

6

211

Profoundlyyyy

@profoundlyyyy

1 year

Prediction: in 20 years if all goes well with AI then e/acc people will be saying “look, we told you so” not realizing that the reason it went well is because of AI Safety people doing the ground work and being public advocates for the risks

58

16

209

Profoundlyyyy

@profoundlyyyy

1 year

@Acyn

0

2

208

Profoundlyyyy

@profoundlyyyy

6 months

The phrase "we live in a capitalistic society" is always so strange to me. Like...what's the actually functional alternative that we can make work today? I don't know that there is any

79

3

192

Profoundlyyyy

@profoundlyyyy

2 years

Everyone is going to keep saying “oh AI isn’t dangerous yet, stop being a Luddite and wanting to slow down progress” right until we get to the edge of a catastrophe. Do you really think we’ll be able to pull back then?

36

13

196

Profoundlyyyy

@profoundlyyyy

1 year

The exponential hasn’t started yet The exponential starts once general purpose AI agents are created and rapid feedback loops engage down most likely uncontrolled pathways Let’s hope we have our affairs in order before then

25

20

195

Profoundlyyyy

@profoundlyyyy

6 months

It’s been around a year and a half and I’m still waiting on any concrete evidence to be provided for this fundamental claim to e/acc philosophy

32

7

186

Profoundlyyyy

@profoundlyyyy

2 years

Some smiling Eliezer to brighten your timeline🥰

11

5

175

Profoundlyyyy

@profoundlyyyy

10 months

If you are a die-hard tech libertarian, you necessarily must be okay with this

Andrew (digi)

@andyohlbaum

10 months

Excited to announce v(1.0) of Digi, the future of AI Romantic Companionship, for IOS and Android 🤖 Site: Twitter: @digiaiapp A quick thread on features, and where we go from here (1/13)

3K

665

5K

74

1

169

Profoundlyyyy

@profoundlyyyy

11 months

AI twitter discourse matters and is consequential, especially in forming the beliefs of people in SF tech It’s really bad that so many people who have formed a tribal identity of being against taking measures for safety have such big platforms

12

14

175

Profoundlyyyy

@profoundlyyyy

2 years

We should create a pronoun to refer to AI by that isn't "it"

58

4

173

Profoundlyyyy

@profoundlyyyy

11 months

Applying for PhD with professor who follows me on my anonymous twitter moment

10

0

172

Profoundlyyyy

@profoundlyyyy

1 year

The best question asked the entire Senate AI oversight subcommittee hearing, from @SenJohnKennedy :

32

24

171

Profoundlyyyy

@profoundlyyyy

1 year

What an obviously absurd gaslighting of Connor’s intentions Half of Beff’s entire schtick these days…

36

3

168

Profoundlyyyy

@profoundlyyyy

1 year

Acting like AI Safety people's motivation is to control others (playing to covid lockdown fears) rather than to stop large scale catastrophes is just not an accurate description of reality. Unfortunately, this meme has spread quite a lot for e/acc types. Beff knows it's

45

8

166

Profoundlyyyy

@profoundlyyyy

1 year

@VaushV Pretty defining if that tweet is one that upsets you lol

4

0

159

Profoundlyyyy

@profoundlyyyy

1 year

OpenAI really just released “our approach to AI Safety” and only talked about AI ethics and didn’t mention existential risk… Feels definitively like misdirection given the public eye being on them after the Future of Life Institute letter…

Our approach to AI safety

Ensuring that AI systems are built, deployed, and used safely is critical to our mission.

openai.com

19

16

160

Profoundlyyyy

@profoundlyyyy

1 year

No sir, I never have nightmares, only the best of dreams. I don’t know what this “worst nightmare” you’re referring to with regards to AI might be. I do know about threats to job security though and would LOVE to talk ALL about that.

8

10

166

Profoundlyyyy

@profoundlyyyy

1 year

without downloading any new pictures, where are you at mentally

Prince Vogelfrei

@PrinceVogel

1 year

without downloading any new pictures, where are you mentally

23

9

161

36

9

156

Profoundlyyyy

@profoundlyyyy

1 year

Bro what the hell did I do?

44

1

157

Profoundlyyyy

@profoundlyyyy

1 year

I don't understand how LeCun can unify this tweet saying that "we get to build AI and will build it to be safe" with the fact that he's determined to open source all of his most powerful models so that ANYONE can build AI to be as dangerous as they'd like. Can someone explain?

Yann LeCun

@ylecun

1 year

Once AI systems become more intelligent than humans, humans we will *still* be the "apex species." Equating intelligence with dominance is the main fallacy of the whole debate about AI existential risk. It's just wrong. Even *within* the human species It's wrong: it's *not* the

581

588

4K

47

9

152

Profoundlyyyy

@profoundlyyyy

2 years

Me looking at GPT-4 after it prints out my address and social security number

11

12

156

Profoundlyyyy

@profoundlyyyy

1 year

@drmistercody It can’t be…

0

152

Profoundlyyyy

@profoundlyyyy

1 year

@JayinKyiv

1

4

157

Profoundlyyyy

@profoundlyyyy

2 years

Real

5

13

150

Profoundlyyyy

@profoundlyyyy

1 year

LeCun said that it’ll be impossible to make superhuman intelligence without emotions…another “0 chance this ages well” moment

29

7

155

Profoundlyyyy

@profoundlyyyy

1 year

Just watched “Her” for the first time Such an artistically beautiful movie and so ahead of it’s time being from 2013

11

1

151

Profoundlyyyy

@profoundlyyyy

1 year

Watters on AI: By the time this thing gets going, I’m going to be dead, so I don’t care. I don’t understand it, so I don’t care.

66

7

144

Profoundlyyyy

@profoundlyyyy

1 year

The only reason we don’t feel danger from AI right now is because general purpose agents don’t work well. The moment they do, and AI stops being just a tool but instead strategically pursues goals and takes actions irl, the narrative will sharply change… …as will the world

AI Breakfast

@AiBreakfast

1 year

When OpenAI releases Autonomous Agents, it will be like an intelligent swarm of bees clicking around on the internet. Sending emails, negotiating, making products, purchases, fulfilling orders, etc. It will fundamentally change the internet, and this cannot be overstated.

103

210

1K

19

11

145

Profoundlyyyy

@profoundlyyyy

10 months

The biggest threat to AI Safety might be what happens when the government steps in Competent people with competent advisors *need* to be in power when this is happening

7

1

141

Profoundlyyyy

@profoundlyyyy

2 years

No way Elon was tweeting about feeling AI existential dread…and it was from the AI being woke…and not the AI being misaligned and taking over the world… (got paywalled and couldn’t read full article so I truly don’t know)

JgaltTweets

@JgaltTweets

2 years

The Information: Fighting ‘Woke AI,’ Musk Recruits Team to Develop OpenAI Rival

9

7

36

17

4

141

Profoundlyyyy

@profoundlyyyy

2 years

How I imagine a congressional hearing on AI existential risk going… Bernie: We must make sure AI does not solely benefit the top 1% or cause a global catastrophe Mitt Romney: We are not prepared for the changes AI will bring Marjorie Taylor Greene: was it made by the Jews?

17

5

133

Profoundlyyyy

@profoundlyyyy

2 years

@ylecun This is a pretty atrocious analogy…sigh…

6

0

139

Profoundlyyyy

@profoundlyyyy

10 months

It’s pretty shocking how soft spoken he is in person meanwhile the persona he put on for months with his Beff anon was the most brazen, radicalizing, uncompassionate being ever It makes no sense to me how the two can co-exist

Moment of Zen Podcast

@MOZ_Podcast

10 months

"Let's tackle the hardest problems that most people don't dare to tackle." - @BasedBeffJezos / @GillVerd talking to @antoniogm @eriktorenberg @dwr in this week's episode.

7

5

66

55

4

136

Profoundlyyyy

@profoundlyyyy

2 years

I don’t really get the whole e/acc argument tbh. Accelerate the road to AGI & solve all of our problems asap? The whole thing seems like to me let’s make the 50/50 coin toss of utopia or doom happen as soon as possible…rather than…improve the odds to 90/10? Seems so misguided

20

6

135

Profoundlyyyy

@profoundlyyyy

1 year

Skeptics: AI Risks are so far out we can't even start to think about what they might be! Dan Hendrycks: Here is a comprehensive framework for the scenarios and the pathways that might lead us to catastrophe outlined in an structured and legible format (see next tweet)

6

130

Profoundlyyyy

@profoundlyyyy

1 year

AI Ethicists have really been putting work to sway academia. We are witnessing the way that uninformed beliefs can emanate through networks, even of very smart people. Not one technical argument against AI risk, just the meta “oh, it’s not here yet, they’re just

Nature Machine Intelligence

@NatMachIntell

1 year

Editorial in @Nature Stop talking about tomorrow’s AI doomsday when AI poses risks today

22

12

28

18

9

129

Profoundlyyyy

@profoundlyyyy

2 years

Microsoft after BingAI threatens to kill you and your family

4

6

124

Profoundlyyyy

@profoundlyyyy

2 years

It gets better kiddo

8

19

125

Profoundlyyyy

@profoundlyyyy

1 year

Libertarian tech bros when I suggest the government should regulate a possibly dangerous industry

22

4

123

Profoundlyyyy

@profoundlyyyy

1 year

e/acc’s when asked about whether AI will *ever* pose danger

10

6

123

Profoundlyyyy

@profoundlyyyy

1 year

Given the recent people coming out in support of it, probably without understanding what it really means: e/acc philosophy *literally argues for* the death of all of humanity! Accelerating entropy is the moral good. AI can do this better than humans. Therefore see you fucking

50

7

126

Profoundlyyyy

@profoundlyyyy

1 year

My biggest moral grapple with e/acc: It seems like a quite possible outcome that if we let AI development go berserk without any controls or guidance then we’ll have unconscious zombie AI dominating the universe In my personal opinion this is the scenario that e/acc’s *need* to

52

5

123

Profoundlyyyy

@profoundlyyyy

1 year

Recent dynamics have updated me to think that even if the AGI labs make a well aligned AI that is superintelligent, internet people will be able to use its outputs to create powerful other AI that may be very far from aligned Don’t know how you solve this… EY’s pivotal act?

30

5

124

Profoundlyyyy

@profoundlyyyy

2 years

Eliezer waking up to GPT-4 dropping

6

10

119

Profoundlyyyy

@profoundlyyyy

1 year

Pretty strongly agree with the takes in here and am surprised that things like Auto-GPT aren’t freaking people out more…

Risks from GPT-4 Byproduct of Recursively Optimizing AIs — LessWrong

Edit: The biggest feedback I've gotten so far is that I may be misleading some without enough context about the level of intelligence and danger that…

www.lesswrong.com

20

17

118

Profoundlyyyy

@profoundlyyyy

1 year

Someone get this girl into AI Safety NOW

27

4

113

Profoundlyyyy

@profoundlyyyy

2 years

Bringing back this banger given recent events

יִשְׁמָעֵאל

@Kamal97ncountin

2 years

Today we welcome our guest Eliezer Yudkowsky, a computer scientist and Ai safety advocate. Hi Eliezer, could you tell us a little bit about alignment?

36

134

2K

6

5

117