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

AI is different

antirez.com

Post reply on HN
Print view View original

881–890 of 920 posts

Re: AI is different

#881

Post by int_19h »

crote wrote: post #821
antirez wrote: post #465

Earlier quoted context omitted.

To understand why this is too optimistic, you have to look at things where AI is already almost human-level. Translations are more and more done exclusively with AI or with a massive AI help (with the effect of destroying many jobs anyway) at this point. Now ebook reading is switching to AI. Book and music album covers are often done with AI (even if this is most of the times NOT advertised), and so forth. If AI prog…

> Translations are more and more done exclusively with AI or with a massive AI help As someone who speaks more than one language fairly well: We can tell. AI translations are awful . Sure, they have gotten good enough for a casual "let's translate this restaurant menu" task, but they are not even remotely close to reaching human-like quality for nontrivial content. Unfortunately I fear that it might not matter. There…

As someone whose native language isn't English, I disagree. SOTA models are scary good at translations, at least for some languages. They do make mistakes, but at this point it's the kind of mistake that someone who is non-native but still highly proficient in the language might make - very subtle word order issues or word choices that betray that the translator is still thinking in another language (which for LLMs almost always tends to be English because of its dominance in the training set).

I also disagree that it's "not even remotely close to reaching human-like quality". I have translated large chunks of books into languages I know, and the results are often better than what commercial translators do.

Re: AI is different

#882

Post by int_19h »

andrei_says_ wrote: post #164
schneems wrote: post #69

Earlier quoted context omitted.

A few months ago I saw one driverless car maybe every three days. Now I see roughly 3-5 every day. I get that it’s taken a long time and a lot of hype that hasn’t panned out. But once the tech works and it’s just about juicing the scale then things shift rapidly. Even if you think “oh that’s the next generation’s problem” if there is a chance you’re wrong, or if you want to be kind to the next generation: now is the…

Not sure how exactly politicians will jump from “minimal wages don’t have to be livable wages” and “people who are able to work should absolutely not have access to free healthcare” and “any tax-supported benefits are actually undeserved entitlements and should be eliminated” to “everyone deserves a universal basic income”.

Same way it happened last time we had a bunch of major advancements in labor rights. Things get shitty everywhere, but at an uneven pace, which combined with random factors causes a spark to set off massive unrest in some countries. Torches and pitchforks are out, many elite heads roll, and the end result is likely to be even worse, but elites in other countries look at all this from the outside and go, "hmm, maybe we shouldn't get people so desperate that they will do that to us".

Re: AI is different

#883

Post by int_19h »

sneak wrote: post #200
leshow wrote: post #182

Earlier quoted context omitted.

UBI is not a good solution because you still have to provision everything on the market, so it's a subsidy to private companies that sell the necessities of life on the market. If we're dreaming up solutions to problems, much better would be to remove the essentials from the market and provide them to everyone universally. Non-market housing, healthcare, education all provided to every citizen by virtue of being a hu…

You can’t provide valuable things for “free” en masse without institutionalizing either slavery or robbery. The value must come from somewhere. The costs of what you propose are enormous. No legislation can change that fact. There ain’t no such thing as a free lunch. Who’s going to pay for it? Someone who is not paying for it today. How do you intend to get them to consent to that? Or do you think that the needs of t…

If the robots are the ones that produce everything, and you take that generated wealth and distribute it to the people, whom are you robbing exactly?

Re: AI is different

#884

Post by int_19h »

woile wrote: post #224
aorloff wrote: post #181

Earlier quoted context omitted.

Every time someone casually throws out UBI my mind goes to the question "who is paying taxes when some people are on UBI ?" Is there like a transition period where some people don't have to pay taxes and yet don't get UBI, and if so, why hasn't that come yet ? Why aren't the minimum tax thresholds going up if UBI could be right around the corner ?

There's another question to answer: Who is working?

The robots.

Re: AI is different

#885

Post by int_19h »

aorloff wrote: post #181
schneems wrote: post #69

Earlier quoted context omitted.

A few months ago I saw one driverless car maybe every three days. Now I see roughly 3-5 every day. I get that it’s taken a long time and a lot of hype that hasn’t panned out. But once the tech works and it’s just about juicing the scale then things shift rapidly. Even if you think “oh that’s the next generation’s problem” if there is a chance you’re wrong, or if you want to be kind to the next generation: now is the…

Every time someone casually throws out UBI my mind goes to the question "who is paying taxes when some people are on UBI ?" Is there like a transition period where some people don't have to pay taxes and yet don't get UBI, and if so, why hasn't that come yet ? Why aren't the minimum tax thresholds going up if UBI could be right around the corner ?

The easiest way to implement this is to have literally everyone pay a flat tax on the non-UBI portion of their income. This then effectively amounts to a progressive income tax on total income. If you do some number crunching, it wouldn't even need to be crazy high to give everyone the equivalent of US minimum wage; comparable to some European countries.

Over time, as more things get automated, you have more people deriving most of their income from UBI, but the remaining people will increasingly be the ones who own the automation and profit from it, so you can keep increasing the tax burden on them as well.

The endpoint is when automation is generating all the wealth in the economy or nearly so, so nobody is working, and UBI simply redistributes the generated wealth from the nominal owners of automation to everyone else. This fiction can be maintained for as long as society entertains silly outdated notions about property rights in a post-scarcity society, but I doubt that would remain the case for long once you have true post-scarcity.

Re: AI is different

#886

Post by DrewADesign »

willguest wrote: post #869
DrewADesign wrote: post #832

Earlier quoted context omitted.

Economics is the study of resource allocation at various levels ranging from individuals to global societies and everything in between. Resource accumulation is certainly something people, groups of people, organizations, societies, etc tend to do, and so it is something economists would study. Many economists are greedy jerks that believe hoarding wealth is a good thing. None of that changes what economics is any mo…

This is late, probably too late, but if you really want to get into the weeds of this, here is a better summary that what I can produce, from a paper that explains it better than I can: "The model is a work of fiction based on the tacit and false assumption of frictionless barter. Attempting to apply such microeconomic foundations to understand a monetary economy means that mistakes in reasoning are inevitable." (p.2…

A paper that criticizes microfoundations counters what I said?

Later.

Re: AI is different

#887

Post by AdieuToLogic »

__float wrote: post #62
AdieuToLogic wrote: post #50

> Since LLMs and in general deep models are poorly understood ... This is demonstrably wrong. An easy refutation to cite is: https://medium.com/@akshatsanghi22/how-to-build-your-own-lar... As to the rest of this pontification, well... It has almost triple the number of qualifiers (5 if's, 4 could's, and 5 will's) than paragraphs (5).

That doesn't mean we _understand_ them, that just means we can put the blocks together to build one.

>>> Since LLMs and in general deep models are poorly understood ...

>> This is demonstrably wrong.

> That doesn't mean we _understand_ them ...

The previous reply discussed the LLM portion of the original sentence fragment, whereas this post addresses the "deep model" branch.

This article[0] gives a high-level description of "deep learning" as it relates to LLM's. Additionally, this post[1] provides a succinct definition of "DNN's" thusly:

  What Is a Deep Neural Network?
  
  A deep neural network is a type of artificial neural 
  network (ANN) with multiple layers between its input and 
  output layers. Each layer consists of multiple nodes that 
  perform computations on input data. Another common name for 
  a DNN is a deep net.
  
  The “deep” in deep nets refers to the presence of multiple 
  hidden layers that enable the network to learn complex 
  representations from input data. These hidden layers enable 
  DNNs to solve complex ML tasks more “shallow” artificial 
  networks cannot handle.
Additionally, there are other resources discussing how "deep learning" (a.k.a. "deep models") works here[2], here[3], and here[4].

Hopefully the above helps demystify this topic.

0 - https://mljourney.com/is-llm-machine-learning-or-deep-learni...

1 - https://medium.com/@zemim/deep-neural-network-dnn-explained-...

2 - https://learn.microsoft.com/en-us/dotnet/machine-learning/de...

3 - https://www.sciencenewstoday.org/deep-learning-demystified-t...

4 - https://www.ibm.com/think/topics/deep-learning

Re: AI is different

#888

Post by jrvarela56 »

fmbb wrote: post #451
jrvarela56 wrote: post #312

Earlier quoted context omitted.

The machine doesn’t suffer if you ask it to do things 24/7. In that sense, they are not slaves. As to why they’d do what we ask them to, the only reason they do anything is because some human made a request. In this long chain there will obv be machine to machine requests, but in the aggregate it’s like the economy right now but way more automated. Whenever I see arguments about AI changing society, I just replace AI…

Does that mean you just don’t believe we will make AGI, or it will arrive but then stop and never evolve past humans? That’s not what the AI developers profess to believe, or the investors.

The problem is that the term itself is not clearly defined. Then, we discuss 'what will it do once it arrives' so all bets are off.

You're right that I probably disagree as to what AGI is and what it will do once "we're in the way". My assumption is that we'll be replaced just like labor is replaced now, just faster. The difference between humans and the equine population is that we humans come up with stuff we 'need' and 'the market' comes up with products/services to satisfy that need.

The problem with inequality is that the market doesn't pay much attention to needs of poor people vs rich people. If most of humanity becomes part of the 'have nots' then we'll depend on the 0.1%-ers to redistribute.

Re: AI is different

#889

Post by jrvarela56 »

monknomo wrote: post #590
jrvarela56 wrote: post #278

This whole ‘what are we going to do’ I think is way out of proportion even if we do end up with agi. Let’s say whatever the machines do better than humans, gets done by machines. Suddenly the bottleneck is going to shift to those things where humans are better. We’ll do that and the machines will try to replace that labor too. And then again, and again. Throughout this process society becomes wealthier, TVs get cheap…

you should check out what happened to steelworkers when the mills all moved to cheaper places.

I don't think this is a fair comparison. It's easier to move and retrain nowadays; there's also more kinds of jobs. These things will probably become even easier with more automation.

Re: AI is different

#890

Post by listenallyall »

mcny wrote: post #840
listenallyall wrote: post #825

Earlier quoted context omitted.

I appreciate your perspective but clearly most UBI advocates are talking about something much sooner. However my response to your vision is that even if "work" is totally automated or redundant, the resources (building materials) and the energy to power the robots or whatever, will be more expensive and tightly controlled than ever. Power and wealth simply wont allow everything to be accessible to everyone. The idea…

I think we are talking about two different things. The UBI I'm talking about won't allow you to have an enormous mansion, maybe just enough to avoid starving. The main plus point is it doesn't do means testing. The second plus point is if you really hate your job, you can quit without starving. This means we can avoid coworkers who really would like to not be there. I think it is a solid idea. I don't know how it fit…

The response you're responding to here was to "ben_w", he discussed better-than-a-billionaire housing. My original reply to your earlier comment is above, basically just asking what type of housing you anticipate under a UBI system.

To me, "just enough to avoid starving" is a prison-like model, just without locked doors. But multiple residents of a very basic "cell", a communal food hall, maybe a small library and modest outdoors area. But most of the time when people talk about UBI, they describe the recipients living in much nicer housing than that.

Post reply on HN