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

The AI Battle

aifuture.substack.com

Post reply on HN
Print view View original

21–30 of 137 posts

Re: The AI Battle

#21

Post by rendang »

psi75 wrote: post #3

If we don't implement universal basic income, we are utterly screwed. Markets where one side is desperate are inelastic, which means that a small change in supply or demand can cause huge price swings. Usually, this works against those who are out of power; for example, oil producers can cause price spikes if they can hold a cartel together. The job market is also extremely inelastic. We might only see 10 or 20 perce…

Is there any historical precedent for a short term, big dive in the wages of a certain sector based on a 20% decline in demand for that labor? If that were the expected outcome, wouldn't one have predicted that the economy-wide declines in the workforce in the Great Recession would have sunk wages?

Re: The AI Battle

#23

Post by ajafari1 »

qualudeheart wrote: post #20
slowmovintarget wrote: post #10

This very much requires acceptance of the worse-is-better model. That "beautiful" drawing of an an astronaut riding a horse is aesthetically... crap. It may take a person 5 hours to paint or draw that, but, arguably, a person wouldn't. GPT-3 might generate a novel, but, generally speaking the prose is jarring and awful to read, because there is no intent (danger: philosophically loaded word) behind it. It's all autom…

I agree that the horse picture sucks. Other people might say that they like Dall-E’s output on Twitter, for signalling purposes, but I still think they don’t really believe that. Still, it’s only a matter of time till nicer graphics are possible. Most graphic designers and artists are still worse than this thing, especially the type that bought a $200 drawing tablet on Amazon and sell their services on Fiverr. Those…

Exactly. Any content writer AI, developer AI, graphic designer AI just has to be good enough to cause an economic labor shift. There are plenty of non big tech companies that want affordable labor that can just get the job done. It doesn't need to be insanely good, although imo it will be in short order anyway.

Re: The AI Battle

#24

Post by slowmovintarget »

mrshadowgoose wrote: post #15
slowmovintarget wrote: post #10

This very much requires acceptance of the worse-is-better model. That "beautiful" drawing of an an astronaut riding a horse is aesthetically... crap. It may take a person 5 hours to paint or draw that, but, arguably, a person wouldn't. GPT-3 might generate a novel, but, generally speaking the prose is jarring and awful to read, because there is no intent (danger: philosophically loaded word) behind it. It's all autom…

This is the classic (and flawed) "it's gotta be perfect otherwise it's useless" argument that you see in other AI fields, such as self-driving. You don't have to be anywhere near perfect to cause tremendous disruption. "Polyester" grade artwork will be perfectly acceptable for a lot of use cases. Just like how self-driving will decimate the trucking workforce, all while people continue to screech about how "it can't…

Polyester was hugely disruptive. I'm not saying we can ignore this, or even that it is useless.

Re: The AI Battle

#25

Post by f38zf5vdt »

As someone working on some of this stuff:

(1) Open source models for image generation that are comparable to what DALLE2 or Imagen can spit out will be available by the end of next year. There will be nothing for OpenAI or Google to monetize.

(2) These models aren't magic. Often to get the content generation you want you need to finetune the models. 21st century graphic designers are well aware of AI image generation are looking forward to streamlining animation, etc. Half of the people I know doing ML pipelines right now are actually graphic designers. In a few years what formerly took whole studios will be able to be done by a single person. Want to make your own animated movie? With enough effort, fine tuning, and going through the generation of millions of images, it will soon be possible.

(3) By finetuning with your own unreleased images, you will always get an image model better at making your style of art than whatever is available to the public. You can use your proprietary fine tuned model to create art you enjoy or sell.

(4) Everyone should keep in mind that 50 years ago, to make a high definition film required a studio and millions of dollars to pay production crew. We're getting to the point where a single person can do that on their phone. I don't think the future is as dark as what everyone thinks it will be, and shaking out the tedium of jobs, even art, is going to help redirect our collective brain effort to more important tasks or more beautiful works.

Re: The AI Battle

#26

Post by marvin »

qualudeheart wrote: post #14

Dark times ahead. If you haven’t already made it big with crypto or startups, you are toast and will be trapped in the permanent UBI underclass. That assumes that there will be a UBI at all. Combined with inflation this will mean that only those with $5 million or more in the bank will live what still count as middle class lifestyles. Those with less money will be forced into poverty. Many will adopt petty crime for…

Things will be shaken up, the way they were when the steam engine was invented. But there's no reason to believe we're headed for a dystopic future.

The result of this development and the underlying trend is vastly increased wealth creation. That's a good thing. You'd have to be a luddite to think otherwise.

Read e.g. https://moores.samaltman.com/, where Sam Altman (OpenAI CEO) suggests a scheme for wealth taxation that gives everyone direct ownership of the AI-created wealth and the machinery to make it.

I'm pretty confident that there would be strong democratic support for a scheme along these lines, given how few are able to participate in the highest echelons of the wealth creation in the kind of AI-driven society that will happen before AI becomes truly autonomous.

Re: The AI Battle

#27

Post by Jack000 »

I think there's some confusion over the terms graphic design, art, and image generation, which are somewhat overlapping but distinct fields. Dalle-2 doesn't really do graphic design, but a future version of it might.

Re: The AI Battle

#28

Post by rendang »

ajafari1 wrote: post #8

Any bets on when the majority of software development is automated? My estimate based on the exponential improvement rate of AI on various tests like MMLU and MATH is within 3 years. I know that might sound crazy, but a year ago it would have been crazy to think that graphics design jobs could be automated by an AI in 2022. Feel free to disagree and discuss why.

Pick a baseline, like 1970, and call the level of coding productivity based on its contemporary languages, processes & tooling "0% Automated". Wouldn't we already in 2022 be well over "50% automated" perhaps even 80% or 90%? So we may indeed see some improvement in 3 years, but it would be a more accurate perspective to call that a move from 90% to 95% automation or something, not 0 to 51%.

Think about agriculture, for example - based on a preindustrial baseline we are at least 98% automated now. Likely also quite high for textiles and other industrial production. It often seems that in discussions of near-future automation, the vast gains in productivity seen in the past 2 centuries are ignored.

Re: The AI Battle

#29

Post by tialaramex »

slowmovintarget wrote: post #10

This very much requires acceptance of the worse-is-better model. That "beautiful" drawing of an an astronaut riding a horse is aesthetically... crap. It may take a person 5 hours to paint or draw that, but, arguably, a person wouldn't. GPT-3 might generate a novel, but, generally speaking the prose is jarring and awful to read, because there is no intent (danger: philosophically loaded word) behind it. It's all autom…

> those systems are not a path to AGI

If that's the actual meat of your claim, it's weird to just put it unsupported at the end. If you think it's an aside you've got a big problem, because without it the rest is pretty weak.

We saw this with chess. Anybody who was clinging on to the idea that well, the machines can't really play chess, because technically there do seem to be a few humans who are better, was screwed the moment the machines begin routinely beating grand masters. Either this categorically isn't something that the machines can do, or, it is and so it's important that they can do it at all. We shouldn't expect any half measures on this.

Do you consider Rob Liefeld an artist? What do you reckon the chances are that the machine can get human anatomy right more often than Rob Liefeld? Rob is a human, so it seems like that should give him an advantage. However unlike Rob DALL-E has learned by seeing lots of existing pictures of humans that they don't look the way Rob draws them...

Re: The AI Battle

#30

Post by bearjaws »

qualudeheart wrote: post #20
slowmovintarget wrote: post #10

This very much requires acceptance of the worse-is-better model. That "beautiful" drawing of an an astronaut riding a horse is aesthetically... crap. It may take a person 5 hours to paint or draw that, but, arguably, a person wouldn't. GPT-3 might generate a novel, but, generally speaking the prose is jarring and awful to read, because there is no intent (danger: philosophically loaded word) behind it. It's all autom…

I agree that the horse picture sucks. Other people might say that they like Dall-E’s output on Twitter, for signalling purposes, but I still think they don’t really believe that. Still, it’s only a matter of time till nicer graphics are possible. Most graphic designers and artists are still worse than this thing, especially the type that bought a $200 drawing tablet on Amazon and sell their services on Fiverr. Those…

The way I have put it to my artist friends is in the form of litmus test: If I sat at a local farmers market selling this art, people would likely buy some of it.

If the delineation can only be made because you are an art major, or practicing artist, that is not really compelling to your own market.

We are on the third generation of these type of "AI" and they are already past the point where people would exchange cash for a print of this work. It is only a matter of time until people are using these to generate the general picture and then drawing it using their preferred medium (e.g. oil, watercolor).

Post reply on HN