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

Where does next-token prediction leave us?

pop.rdi.sh

Post reply on HN
Print view View original

81–90 of 221 posts

Re: Where does next-token prediction leave us?

#81

Post by slopinthebag »

brookst wrote: post #23
overgard wrote: post #19

Earlier quoted context omitted.

I hate the term "democratization". It's putting a respectable face on something that shouldn't be respectable at all. In the age of the internet, coding and other creative skills have always been largely democratized to people that care enough to learn. Nothing is being "democratized" by AI, there's simply (an attempt) at driving the value of actual skill to zero so the skill-less and stupid can purchase their way to…

It’s not coding that’s being democratized, it’s the ability to create games, tools, etc, that require coding. Suppose it became possible to buy a Ferrari (not that new hideous one) for $5k. That would be democratizing Ferrari ownership: far more people could do it. I’m sure there would be investment bankers who would complain it devalues their hard work (I am not a fan of those people, but it is notoriously long hour…

If that happened, Farrari's would simply lose their value and something more rare would take it's place. The scarcity is the point. Same with art, if you "democratise" it somehow humans will seek out a new form of artistic expression that is not accessible to people with a $10 midjourney subscription. And that is OK, good actually.

Re: Where does next-token prediction leave us?

#82

Post by treebeard901 »

sigmar wrote: post #2

>But lately I’ve been thinking if it is just a class issue? This cohort of people likely have a cushion that softens the concussive blows they are doling out right now. They perhaps have the luxury of a somewhat functioning government and a social safety net that they are witness to in all walks of life. Over half the world does not. Science and technology, I feel, has always had a certain apathy towards the plight t…

One of the billionaires on the All In Podcast, (which I listen to sometimes to get that kind of perspective), mentioned that one thing the U.S. and China could work on as far as regulating AI, is Know Your Customer controls. At first after one of the hosts said this I didn't think much of it but then started thinking through the second order effects later on. Why would that be such a main concern? He gave the example of dangerous people using AI to do bad things, but I think the subtext is much more complicated. At least from the ownership class point of view. Which also ties into the original idea behind OpenAI. Which is that the public should have access to the full abilities of any model equally. But the ownership class can't have this. In another post someone made a good point about licenses and this can be extended to stuff like the legal system, where at the end of the day the personal human connections decide things far more than what a model would compute the result of a case to be...

So because so much of maintaining the forms of power and order within the wealthy class has been reliant on information, and again the legal example can be used here because if you hire a lawyer, you're largely paying for information and access. Now the powers that be to compete with open information have to make the access side of the equation much more important to maintain the status quo with such a disruptive technology. And that is another layer of the need for being able to say, degrade a Chinese model if a U.S. citizen the Govt doesn't like is trying to bypass the restrictions implemented on them using models made in the U.S.

In other words, KYC is about restoring the historical aspect of needing money to have information. And there is a class warfare aspect to looking at it that way.

Re: Where does next-token prediction leave us?

#83

Post by rzmmm »

Gigachad wrote: post #76
fny wrote: post #69

Earlier quoted context omitted.

I don't think its entirely class. I've met many people who have convinced themselves they have some special sauce--even in software! And even those who are insulated like doctors and plumbers don't realize they're still at risk from second order effects. What happens when half of their customers are broke? That possibility is not an exaggeration. Even if everyone re-skills over night, people will default on their mor…

Even if you are a plumber and we don’t automate plumbing, you’ll have half the population displaced and switching to plumbing. Everything that isn’t automated will be over saturated.

Or average joe becomes DIY plumber with help of chatGPT and there is a lot less demand for plumbers then.

Re: Where does next-token prediction leave us?

#84

Post by dclowd9901 »

rootusrootus wrote: post #42
overgard wrote: post #40

I have a more depressing theory. I think class is part of it, but I'm starting to suspect that a shockingly large number of people lack the critical thinking skills needed to think out the implications of this stuff. I say that because I've met so many people of the managerial class that seem to think it's going to replace the annoying people they have to pay, but somehow they don't think it's going to come for them…

Some get it, though. When I quipped that Claude may eventually replace me, my manager was visibly shaken as he mentioned it would probably come for him first. I feel better about it than I did a few months ago. Still seems like there’s something missing that is going to be hard to make happen. I forsee humans being necessary for a long while yet.

"Ingenuity" is what I think is missing. The sheer _want_ of solving a problem that is distinctly a living creature's concern.

The irony is if we ever taught machines how to have this, they'd probably not want to work for us anymore.

Re: Where does next-token prediction leave us?

#85

Post by Gigachad »

rzmmm wrote: post #83
Gigachad wrote: post #76

Earlier quoted context omitted.

Even if you are a plumber and we don’t automate plumbing, you’ll have half the population displaced and switching to plumbing. Everything that isn’t automated will be over saturated.

Or average joe becomes DIY plumber with help of chatGPT and there is a lot less demand for plumbers then.

ChatGPT is probably less useful than youtube for this kind of stuff.

Re: Where does next-token prediction leave us?

#86

Post by 0x5FC3 »

simianwords wrote: post #62

This author fundamentally doesn't understand the mental model of the people they are describing and makes huge sweeping claims in what they think is a in a savvy manner but they are completely wrong. Here's the RIGHT mental model of the people the author is talking about. 1. If AI is good enough that it can boost productivity by 20% then it is good for society in general because the gains will be redistributed (as it…

> good for society in general because the gains will be redistributed

How exactly? Is this a version of trickle down economics I am unfamiliar with?

> What better thing can happen to the curious ones amongst us to get an oracle that can answer every question

How are the curious going to eat or have a roof over their heads? Or how are the curious going to pay for the tokens?

> civilisation will be changed in ways that English doesn't even have the vocabulary to describe

Maybe, we ought to think real hard and slow about it?

Re: Where does next-token prediction leave us?

#87

Post by exfalso »

fny wrote: post #69
overgard wrote: post #40

I have a more depressing theory. I think class is part of it, but I'm starting to suspect that a shockingly large number of people lack the critical thinking skills needed to think out the implications of this stuff. I say that because I've met so many people of the managerial class that seem to think it's going to replace the annoying people they have to pay, but somehow they don't think it's going to come for them…

I don't think its entirely class. I've met many people who have convinced themselves they have some special sauce--even in software! And even those who are insulated like doctors and plumbers don't realize they're still at risk from second order effects. What happens when half of their customers are broke? That possibility is not an exaggeration. Even if everyone re-skills over night, people will default on their mor…

Agree on the knock-on effects. My prediction is deflation. Money will be worth more and more. As a consequence governments will have to step in to ensure inflation(with e.g. universal income), otherwise the economy stops.

But honestly I'm not sure this will be enough for people to spend on e.g. restaurants or activities or oh I don't know, children. I think this will imply a freezing or even stepping back on the Maslow pyramid, the majority of people consolidating in the middle.

What I'm mostly concerned about is not even economic, it's psychological. With nothing to do, people will not have purpose, and bored people are a gunpowder keg.

Re: Where does next-token prediction leave us?

#88

Post by ninalanyon »

mmilunic wrote: post #8

> Non-technical middle managers who have not written a line of code in their lives, now feel that the biggest obstacle between them and greatness has lifted. I find it interesting how this is almost the “democratization” you mentioned that AI provides. While AI “democratizes” certain technical ability, in some ways the democratization of things can actually be bad, in that this “democratization” pushes us towards a s…

> individual bargaining capability

Thinking that "individual bargaining capability" capability is an especially useful thing for society as a whole is one of the things that has got us into the current Hobbesian mess

Re: Where does next-token prediction leave us?

#89

Post by adamddev1 »

sigmar wrote: post #2

>But lately I’ve been thinking if it is just a class issue? This cohort of people likely have a cushion that softens the concussive blows they are doling out right now. They perhaps have the luxury of a somewhat functioning government and a social safety net that they are witness to in all walks of life. Over half the world does not. Science and technology, I feel, has always had a certain apathy towards the plight t…

Education systems and cultures that value "sounding" correct, repeating patterns and information, and immediate practical gain appear to like AI more that the cultures with a strong tradition of critical thinking. The cultures that love AI are the same cultures that will tend to build something cheap and quickly so that it can sell. The cultures that are less favorable are the ones that invest in building things correctly and solidly and value long term thinking. People with the immediate-gain-over-truth mindset love AI.

> https://www.visualcapitalist.com/survey-how-21-countries-vie...

A graph of corruption levels would also almost mirror the values in the first link. The more corrupt a country is, the more the public loves AI. AI enthusiasm is almost directly correlated to a general erosion of truth.

In the same way, those with a high value of correctness and stability, like people who work w Haskell, Agda, or Type Theory are not as enthusiastic about AI as people who copy and paste boilerplate and snippets for quick, practical react apps.

Re: Where does next-token prediction leave us?

#90

Post by agnosticmantis »

simianwords wrote: post #62

This author fundamentally doesn't understand the mental model of the people they are describing and makes huge sweeping claims in what they think is a in a savvy manner but they are completely wrong. Here's the RIGHT mental model of the people the author is talking about. 1. If AI is good enough that it can boost productivity by 20% then it is good for society in general because the gains will be redistributed (as it…

> If AI is good enough that it can boost productivity by 20% then it is good for society in general because the gains will be redistributed (as it has always been)

This redistribution has never been as automatic and inevitable as you seem to assume.

> What better thing can happen to the curious ones amongst us to get an oracle that can answer every question?

Getting paid to answer said questions would be nice. The alternative is you'll have to work 14-hour shifts in a warehouse to be able to pay for ChatGPT 10.0 subscription with ads, but sure, it can answer your math problems and satiate your mathematical curiosity.

Post reply on HN