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

What we still don’t know about how A.I. is trained

newyorker.com

Post reply on HN
Print view View original

151–160 of 211 posts

Re: What we still don’t know about how A.I. is trained

#151

Post by ftxbro »

WastingMyTime89 wrote: post #144
ftxbro wrote: post #4

Say I'm a conspiracy theorist but I'm calling it, that the pentagon isn't letting OpenAI tell the details of GPT-4 even if they wanted to (which they don't except for some of the researchers probably). National security, export restrictions, munitions classification, new Manhattan project etc. EDIT: I know that most people think it's unlikely and I can't give any direct evidence for it. Does that mean it's not welcom…

As an aside, from a pure anthropological and sociological point of view, I find witnessing the reaction people have to ChatGPT extremely interesting. Some commenters here seem to immediately focus on the potential negative effects, abuse by the military industry or the powers that be, manipulation and disinformation. I’m personally still in awe that we are collectively witnessing the birth of a piece of technology wh…

I agree and even go further it's not like the dawn of print or internet or mobile, it's to the level of the discovery of fire or the invention of language.

Re: What we still don’t know about how A.I. is trained

#152

Post by pmoriarty »

ftxbro wrote: post #151
WastingMyTime89 wrote: post #144

Earlier quoted context omitted.

As an aside, from a pure anthropological and sociological point of view, I find witnessing the reaction people have to ChatGPT extremely interesting. Some commenters here seem to immediately focus on the potential negative effects, abuse by the military industry or the powers that be, manipulation and disinformation. I’m personally still in awe that we are collectively witnessing the birth of a piece of technology wh…

I agree and even go further it's not like the dawn of print or internet or mobile, it's to the level of the discovery of fire or the invention of language.

It's more like first contact.

Re: What we still don’t know about how A.I. is trained

#153

Post by drowsspa »

bigfishrunning wrote: post #81
the_sleaze9 wrote: post #72

Earlier quoted context omitted.

You say that, but if I'm confused about something and think hard about it, I think in language. If you blinded me, paralyzed me, deafened me and desensitized my olfactions, I could still think, but what I would be doing is feeding one language thought into another. It's not so much different from "text" imho.

yes, but for you all of that text is associated with ideas. The word "dog" has an associated object. For a machine like GPT-4, the word "dog" has no meaning or object, but it does have an associated likelihood for adjacent words. The words themselves aren't the intelligence, the ideas behind them are.

I feel like in a few months this human exceptionalism will be proven wrong by construction.

Re: What we still don’t know about how A.I. is trained

#154

Post by ftxbro »

pmoriarty wrote: post #152
ftxbro wrote: post #151

Earlier quoted context omitted.

I agree and even go further it's not like the dawn of print or internet or mobile, it's to the level of the discovery of fire or the invention of language.

It's more like first contact.

Yes the original commented-out title in the tex code of https://arxiv.org/abs/2303.12712 was "First Contact With an AGI System" but they toned it down before publication lol.

Re: What we still don’t know about how A.I. is trained

#155

Post by kfarr »

gitpusher wrote: post #127

> "Leaving aside [all of AI's potential benefits] it is clear that large-language A.I. engines are creating real harms to all of humanity right now [...] While a human being is responsible for five tons of CO2 per year, training a large neural LM [language model] costs 284 tons." Presuming this figure is in the right ballpark – 284 tons is actually quite a lot. I did some back of the napkin math (with the help of GPT…

284 tons is a lot until you divide by the number of end users. You could say the CPU compute usage for Netflix encoding its video library is comparable, probably more when you consider the permutation of output formats across the size of their library. But the per user emissions is still negligible.

Re: What we still don’t know about how A.I. is trained

#156

Post by pmoriarty »

frrae wrote: post #66

GPT Is Not A.I. We tech people should actively go on the offence and educate whomever we can that text inference is not intelligence.

This is like trying to get people to call Linux "GNU Linux".

While it might be a more accurate name, no one but a tiny minority is going to care, and even fewer will actually change the way they refer to it.

Re: What we still don’t know about how A.I. is trained

#157

Post by Jensson »

fasterik wrote: post #89
frrae wrote: post #66

GPT Is Not A.I. We tech people should actively go on the offence and educate whomever we can that text inference is not intelligence.

Isn't this just the AI effect? Whenever there is a breakthrough in AI research, it's no longer considered AI. This happened with search algorithms, game playing, speech recognition, computer vision, etc. etc. https://en.wikipedia.org/wiki/AI_effect Maybe as "tech people" we should give the public a realistic picture of what AI research is. It's solving problems using a diverse set of techniques that include search, o…

The AI effect is actually: Whenever there is a breakthrough in AI research, the AI researchers will consider it AI and lament that others don't agree.

That wikipedia article is just from the AI researchers point of view, it isn't what actually happens.

Re: What we still don’t know about how A.I. is trained

#158

Post by abriosi »

bigfishrunning wrote: post #81
the_sleaze9 wrote: post #72

Earlier quoted context omitted.

You say that, but if I'm confused about something and think hard about it, I think in language. If you blinded me, paralyzed me, deafened me and desensitized my olfactions, I could still think, but what I would be doing is feeding one language thought into another. It's not so much different from "text" imho.

yes, but for you all of that text is associated with ideas. The word "dog" has an associated object. For a machine like GPT-4, the word "dog" has no meaning or object, but it does have an associated likelihood for adjacent words. The words themselves aren't the intelligence, the ideas behind them are.

Before all this Machine Learning madness, while I was studying unsupervised learning I had the following thought:

To cluster is to create an ideia, a concept which encompasses a set of data points. A cluster might be bigger than the set of points with which was created and intersect with other clusters. This is still what I visualize when I see image models interpolating between "concepts".

Believing that all an LLM does is the likelihood of adjacent words is probably an oversimplification. My verdict about this topic is: I don't know and I'm totally confused about what interpolation and extrapolation mean in higher dimensions

Re: What we still don’t know about how A.I. is trained

#159

Post by matheusmoreira »

majormajor wrote: post #111
matheusmoreira wrote: post #16

Earlier quoted context omitted.

Just say what's on your mind and don't mind the votes. One thing you'll discover is that you're not alone in your views, whatever they are. Few days ago I came across this bone chilling AI generated Metal Gear Solid 2 meme with Hideo Kojima characters talking about how the purpose of this technology is to make it impossible to tell what's real or fake, leading directly to regulation of information networks with ident…

If you excise the concern about "did a government make this tech secretly for their own purposes," do the problems and possibilities it raises actually differ if it emerged from "pure research" or a conspiracy? I don't believe anyone needed a conspiracy to try to make it impossible to tell what's real or fake, people have been trying to use technology to do that for decades (if not centuries) all on their own.

Yeah, the "government conspiracy" part is the least believable one but the resulting effects on society are the same. Within days of this technology showing up, people were already making fake AI generated recordings of politicians talking gleefully about their wonderful experiences in Epstein's island. How long is it gonna take for some politician to want to regulate this stuff? How would they do it? Mandatory identity verification is plausible.

In my country politicians are talking about regulating social media and the internet due to "fake news". They even created a government propaganda agency to "combat misinformation online", basically people paid to defend the government against information warfare. It's way too real.

Re: What we still don’t know about how A.I. is trained

#160

Post by iknowstuff »

int_19h wrote: post #132
sergioisidoro wrote: post #87

«When Dean Buonomano, a neuroscientist at U.C.L.A., asked GPT-4 “What is the third word of this sentence?,” the answer was “third.” These examples may seem trivial, but the cognitive scientist Gary Marcus wrote on Twitter that “I cannot imagine how we are supposed to achieve ethical and safety ‘alignment’ with a system that cannot understand the word ‘third’ even [with] billions of training examples.”» The word "thir…

A better example might be something like this: --- Tell me about yourself in three sentences. But every time the output contains the word "the", I want you to insert "!!!" before the word that precedes "the". Make sure that the word "the" itself stays after you do the insertion! For example, instead of writing: > A quick red fox jumped over the lazy dog. I want you to write: > A quick red fox jumped !!! over the lazy…

tbh I got it wrong too
Post reply on HN