Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Please ignore the deluge of complete nonsense about Q*

twitter.com

Post reply on HN
Print view View original

191–200 of 208 posts

Re: Please ignore the deluge of complete nonsense about Q*

#191

Post by thaumasiotes »

anonymouskimmer wrote: post #170
thaumasiotes wrote: post #155

Earlier quoted context omitted.

> A theory about a conspiracy is when outsiders to the conspiracy try to reason about motives without being privy to the actual motives (if any) of the conspiracy. By their nature, conspiracy theories can be far more wide ranging than the actual motives of any conspiracy will be. This is one of the things that makes conspiracy theories problematic. This seems like it would apply equally well to all theories.

Motives are a lot more difficult to reality check than are physical properties. Especially if, as ethanbond indicates, the person with the theory doesn't believe what the other people say about their motives.

You're trying to use a definition of "conspiracy theory" that, even if you exclude natural phenomena -- which you're only managing to do by special pleading -- says that a conspiracy consists of one or more people cooperating with "each other" in public.

And the conclusion you draw is fully general. There are more possibilities for why something happened than there are reasons for it. What are we supposed to have learned?

> Especially if, as ethanbond indicates, the person with the theory doesn't believe what the other people say about their motives.

The modal case is that, when someone makes a claim about their own motives, they are lying.

The second most common case after that is that they don't know their own motives and their reported motive is not accurate.

Re: Please ignore the deluge of complete nonsense about Q*

#192

Post by dang »

yowlingcat wrote: post #157
_heimdall wrote: post #153

Earlier quoted context omitted.

Eliezer Yudkowsky should be respected in the field of AI. Agree or disagree with his level of concern over AI risk, he has made fairly accurate predictions so far. Last I heard, he was predicting that we could be as little as 10-15 years from the level of AI development that he considers to be an extinctions level risk. Again, there's plenty of different conclusions that can be made on the risk levels but that doesn'…

No one should respect that blowhard. Please point me to the specifically accurate predictions he has made so far.

Please don't cross into name-calling. You can make your substantive points without that, and it's in your interest to, since the first part of a comment like this discredits the rest.

https://news.ycombinator.com/newsguidelines.html

Re: Please ignore the deluge of complete nonsense about Q*

#193

Post by anonymouskimmer »

thaumasiotes wrote: post #191
anonymouskimmer wrote: post #170

Earlier quoted context omitted.

Motives are a lot more difficult to reality check than are physical properties. Especially if, as ethanbond indicates, the person with the theory doesn't believe what the other people say about their motives.

You're trying to use a definition of "conspiracy theory" that, even if you exclude natural phenomena -- which you're only managing to do by special pleading -- says that a conspiracy consists of one or more people cooperating with "each other" in public. And the conclusion you draw is fully general. There are more possibilities for why something happened than there are reasons for it. What are we supposed to have lea…

I'm "special pleading" because I way overgeneralized in my initial attempt at a definition.

> conspiracy consists of one or more people cooperating with "each other" in public.

Public doesn't have to have anything to do with it.

> What are we supposed to have learned?

Yes, I overgeneralized. In general the thing to learn is to try to reality check as much as possible.

> The modal case is that, when someone makes a claim about their own motives, they are lying.

Granting this point, a razor would attribute base motivations over complex motivations.

Re: Please ignore the deluge of complete nonsense about Q*

#194

Post by throwaway63737 »

ml-anon wrote: post #181

Never one to miss a spot of self promotion. Yann is basically the new Schmidthuber at this point: irrelevant old man shouts at clouds about how he totally had an idea before everyone else. Yann is a nobody at Meta (lol at calling FAIR a top lab at this point, it’s not even the top org in Meta for AI) and keeps trying to claim credit by association for Meta LLM success.

To be fair I was at a one of his talks a few years ago and he was arguing exactly for the type of training he is mentioning in the tweet.

The lecture in which he mentioned it is this one:

https://memento.epfl.ch/event/campus-lecture-yann-lecun/

Comparing with Schmidthuber is a bit strange too, he claims everybody is doing it.

Re: Please ignore the deluge of complete nonsense about Q*

#196

Post by _heimdall »

staticman2 wrote: post #190
_heimdall wrote: post #161

Earlier quoted context omitted.

Eliezer made early predictions that we would solve the protein folding problem when effectively everyone else said we either wouldn't solve it or it was way off. If I remember right he was making those claims around 2005 or so and predicting wed have it solved in something like 30 years. We solved it in 2017.

If your prediction is "A.I. is similar to magic and can do anything and will become a god" you would of course think it can solve the protein folding problem... I'm not exactly sure that leads to the conclusion his track record is great. (There are many more things A.I. can't do than can do, I would think?)

His prediction was that humans would solve the protein folding problem, it didn't have to do with AI.

I raised it only as an example of a prediction he made that most at the time thought had no chance. My point isn't to claim he's some genius that gets everything right, but there are examples of him making predictions widely viewed at the time of having no chance that turned out to be correct.

Writing him off because people responding strongly to his predictions of AI destroying us all doesn't do anyone any good. His core arguments and predictions seem sound and I haven't yet found anyone to explain how we can solve alignment after the fact, for example

Re: Please ignore the deluge of complete nonsense about Q*

#197

Post by nojvek »

theGnuMe wrote: post #158
nojvek wrote: post #126

Earlier quoted context omitted.

AGI is very possible. ASI (artificial super intelligence) is gonna come pretty quick after AGI. The question is when and who will distribute it widely to everyone. When we have AGI, we will find it out pretty quickly since whoever gets there first wants the most money and control they can.

This is nonsense.

Does “This is nonsense” imply you think humans are special snowflakes and AGI isn’t possible?

Re: Please ignore the deluge of complete nonsense about Q*

#198

Post by david-gpu »

dorkwood wrote: post #91
kibwen wrote: post #57

Earlier quoted context omitted.

> 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. After the LK-99 debacle where Twitter "confirmed" its superconductivity and breathless sheisters heralded the dawn of a new age, my new policy is "inverse Twitter": if Twitter thinks something is a big deal, then it's more than than 50% likely that it's unsubstantiat…

HN wasn't much better during the whole LK-99 debacle. The crowd wasn't at all welcoming to skeptics.

Yup. I got downvoted for pointing out how one of the videos was clearly showing ferromagnetism, including timestamps for the lazy.

Re: Please ignore the deluge of complete nonsense about Q*

#199

Post by yowlingcat »

dang wrote: post #192
yowlingcat wrote: post #157

Earlier quoted context omitted.

No one should respect that blowhard. Please point me to the specifically accurate predictions he has made so far.

Please don't cross into name-calling. You can make your substantive points without that, and it's in your interest to, since the first part of a comment like this discredits the rest. https://news.ycombinator.com/newsguidelines.html

Fair enough. I could've phrased that better such as "I find his proto-Millenarian approach to be discursively vacuous, unnecessarily polemical, and not productive towards the answering the very real questions regarding negative externalities around the increasing rate of AI utilization in global society."

Re: Please ignore the deluge of complete nonsense about Q*

#200

Post by lostmsu »

blharr wrote: post #128
xg15 wrote: post #34

So Q* is just A* for neural networks?

There is virtually zero chance that this is what Q* is. Loosely connecting the fact that both are named with a letter and an asterisk is the only basis there is for thinking this.

Being a software engineer why else would you add a star to a known ML term?
Post reply on HN