[HN Gopher] Please ignore the deluge of complete nonsense about Q*
___________________________________________________________________
Please ignore the deluge of complete nonsense about Q*
Author : mfiguiere
Score : 99 points
Date : 2023-11-24 20:14 UTC (2 hours ago)
(HTM) web link (twitter.com)
(TXT) w3m dump (twitter.com)
| 6gvONxR4sf7o wrote:
| "Please ignore all the speculation... and now for my
| speculation."
| michael_nielsen wrote:
| LeCun at least knows a lot about AI. Most of the Q* stuff is
| coming from people who know almost nothing.
| 6gvONxR4sf7o wrote:
| But his guesses match the consensus, so it doesn't add to the
| stuff he's criticizing.
| michael_nielsen wrote:
| Fair enough. I suppose the useful bit is that lots of
| ignorant people have gone nuts ("OMG this is AGI!") without
| any details. That's just hype. But, yes, to your point,
| there is also some substantive and interesting speculation
| from more knowledgeable people.
| Cacti wrote:
| Same could be said for you, who is also contributing
| nothing.
| samrus wrote:
| It sounds like you prefer something that's more sensational
| simply because it's more sensational
| wkat4242 wrote:
| Seems to be a common theme with stuff starting with Q...
| imjonse wrote:
| I've run ollama on my macbook, watched a couple videos on
| prompt engineering, tried out stable diffusion on my phone. I
| am even working on a startup that is basically a shiny
| website plus an OpenAI API wrapper on the backend. What do
| you mean I am not qualified to speculate on what Q* from
| OpenAI is and how it is transformational for society!!?? /s
| qualifiedai wrote:
| Yann has been a refreshing source of reason and common sense with
| regards to AI safety, regulation and open-source. I wish we had
| more people like him and less AI doomer cultists.
| hackinthebochs wrote:
| People like his takes because he gives an authoritative gloss
| to what they already believe. But his points are usually
| lacking in argument or rigor. Anyone that essentially expects
| the public to trust them when it comes to the outcome of AI/AGI
| should be view with suspicion.
| peyton wrote:
| I mean I don't think predicting the future is something that
| typically involves rigor. The outcome is pretty clear:
| whatever makes a ton of money. Probably a trusted friend in
| your pocket that sometimes helps you buy stuff. The most
| negative predictions are silly because they don't involve
| making a ton of money for anybody.
| makeitdouble wrote:
| The point about money is important, but we should also keep
| in mind most outcomes will make a ton of money for someone
| somewhere.
|
| Hell, there's wars killing tens of thousands of people
| going on right now, and a ton of money is changing hands
| making a juicy business for whole industries.
| qualifiedai wrote:
| On the contrary, he gives good arguments about why open is
| safer and closed is more dangerous whereas other side gives,
| imho, convoluted arguments and asks for them to be proven
| wrong (as opposed to trying to prove themselves right).
| hackinthebochs wrote:
| His arguments in defense of barreling forward with AI are
| terrible. They have zero chance to convince someone who
| doesn't share his intuitions/interests. For example:
| https://twitter.com/ylecun/status/1718764953534939162
|
| How easily smart people convince themselves of what they
| want to be true with zero self-awareness makes me much more
| fearful of what's to come.
|
| >other side gives, imho, convoluted arguments and asks for
| them to be proven wrong (as opposed to trying to prove
| themselves right).
|
| The question is what should our default stance be until
| proven otherwise? I submit it is not to continue building
| the potentially world-ending technology.
| qualifiedai wrote:
| The default in science is that the side arguing a point
| has the burden of proving it correct. Not asking the
| other side to prove them wrong.
| avsteele wrote:
| Please point me to an example of his good arguments.
|
| I only see his posts on Twitter but haven't been impressed.
| mycologos wrote:
| At the same time, it seems like _some_ antidote is needed to
| the breathless, quasi-mystical hype that cryptic OpenAI
| claims seem designed to stoke. Demanding precise and
| substantive criticisms of something about which almost no
| technical details have been provided seems like an unfair
| bar.
| UniverseHacker wrote:
| His arguments about safety are all just wishful thinking- he
| never addresses the substance of "AI doomer" concerns or
| arguments.
| uoaei wrote:
| It's hard to call anything that comes from LeCun "news": any time
| you hear of a phenomenon in the ML/AI space, you know pretty much
| exactly the sentiment he's going to express. His entire brand is
| "doomers are wrong, you can trust me, I am AI daddy".
| renewiltord wrote:
| Doomers thought GPT-2 was too dangerous to release. I guess one
| can be Dalio successful by calling doom at every instant like
| Dalio.
| uoaei wrote:
| And one can be unfalsifiably successful by using mass media
| to proclaim safety at every instant. The irony is that if
| things do fail these kinds of pronouncements can no longer be
| made using tools like Twitter, etc. on which influencers like
| LeCun build their brands, since their existence and utility
| depends on the stability of society.
| PopePompus wrote:
| Yes, the Anti-doomers will be right many times, and wrong at
| most once.
| wkat4242 wrote:
| The point is moot anyway.
|
| If someone can build it someone will. Laws and impending
| Doom or not. It's probably going to be better if it's us
| than Russia or China.
| benatkin wrote:
| That's a reductive argument, and won't work with me. What
| we have is tantamount to an arms race, and trying to
| suppress another country's development of tech that could
| be used against them is a thing. We have already
| restricted access to our microchips to China specifically
| and Russia through broad sanctions.
| https://www.nytimes.com/2023/10/17/business/economy/ai-
| chips... Russia isn't poised for the current crop of AI
| tech anyway.
| TeMPOraL wrote:
| Well, GPT-2 led directly to GPT-3. GPT-3 led to GPT-3.5 and
| then to GPT-4. GPT-4 might lead to all of us losing our
| source of income, so they may yet be proven right (economic
| turmoil _can_ be an x-risk if it 's large enough).
| s-xyz wrote:
| I should maybe dig a bit deeper into what he is saying, but
| every-time I get excited about some development I get discouraged
| by his views. Perhaps they are realistic, but I prefer to dream
| sometimes.
| andy_xor_andrew wrote:
| I took an AI course in college back around 2015. Just a bit
| before AlphaGo.
|
| One project was to implement a simple Q-learning action/value
| system to play simple games, like Pacman.
|
| The crypto-bros-turned-AGI-experts on twitter are spouting the
| most uninformed, misguided garbage about this whole thing, it's
| quite amazing to watch.
|
| And I'm not saying that I am smart or an expert about Q* because
| I took an introductory college course. I'm saying that even I,
| _someone who knows basically nothing beyond the introductory
| concept_ can identity that these people have _no_ clue what they
| are talking about, and yet the have this incredible talent of
| speaking in such an authoritative and faux-intelligent tone. It
| 's amazing.
|
| My favorites are the tweets that sound like this:
|
| "So, now we know that [insert something totally wrong]. Well,
| what if extend that further, by [another totally wrong
| conclusion]. Here's an explanation of how this all works. A
| thread, 1/N"
|
| followed by a full thread, images included, of drivel.
| nothrowaways wrote:
| Exactly, a VC hype.
| kypro wrote:
| I'm confused. I thought people were worried about the danger of
| some AI breakthrough? If researchers at OpenAI have developed an
| LLM more advanced than GPT4 which can also plan is that not
| potentially a worrying breakthrough?
| laserbeam wrote:
| It's just rumours. Everything I've read about that breakthrough
| sounds about as thoroughly backed in reality as a generic
| conspiracy theory.
|
| There may be a breakthrough, there may not, but nothing on the
| topic is convincing or worth reading.
| jordanpg wrote:
| Rumors that were reported on by a reputable news agency:
| https://www.reuters.com/technology/sam-altmans-ouster-
| openai...
| mianos wrote:
| They clarify their own article as unsubstantiated by
| seconds sources:
|
| > Reuters could not independently verify the capabilities
| of Q* claimed by the researchers.
|
| True, they are a reputable news agency, but the parent is
| also correct, it's not highly credible.
| martythemaniak wrote:
| Q* is the new "GPT-4 has a hundred trillion parameters".
|
| https://thealgorithmicbridge.substack.com/p/gpt-4-a-viral-ca...
| riazrizvi wrote:
| What are the more popular themes of the complete nonsense about
| Q*? Anyone know? I deleted my twitter account.
| wkat4242 wrote:
| Wow I work in AI (implementation) and I have zero idea what he's
| talking about lol
| layer8 wrote:
| I love that reply:
|
| LeCun: "[Note: I've been advocating for deep learning
| architecture capable of planning since 2016]."
|
| Reply: "My understanding is Schmidhuber already solved that 10
| years ago. Just no-one knows it yet."
| ren_engineer wrote:
| The leak about Q* feels like an olive branch to let the former
| OpenAI board and Ilya save a bit of face, probably part of the
| terms for Sam coming back plus it distracts from all the drama
| and puts a positive spin on things.
|
| >"The board didn't handle things well, but they were right to be
| concerned because OpenAI did have some sort of research
| breakthrough"
|
| Not coincidence that this leaks after Sam comes back, rather than
| before when it could have made the board look more justified in
| their decision. This changes the story from incompetence to "it's
| a problem only OpenAI has because they are so far ahead and close
| to AGI". Masterful PR move to leak this and shift the narrative
| iepathos wrote:
| Good point about timing of the leak. IMO the whole fiasco still
| reeks of incompetence and no PR move can wash that clean.
| eigenvalue wrote:
| To be fair, Altman separately alluded to a recent big
| discovery/breakthrough in a talk shortly before all the drama
| went down.
| xg15 wrote:
| So Q* is just A* for neural networks?
| trhway wrote:
| A conspiracy minded me may see all that saga of the last week as
| a marketing campaign to generate excitement for the GPT-5-now-
| with-30%-more-Q release.
|
| Ane even more conspiracy-minded take - from the Russian news
| naturally - is that it is the Great Battle for the future of
| humanity between "doomers" (Oh, no! the AI is going to kill us
| all, we need to stop all the work and control the GPUs like guns)
| and "Effective Altruists" (We can do all the evil today in order
| to achieve greater good tomorrow)
___________________________________________________________________
(page generated 2023-11-24 23:00 UTC)