Hacker News Reader

Comment by vlunkr | Hacker News Reader

This is just my cynical opinion, but I think humanity would be better off if we all just agreed to throw our AI research in the garbage. What is the upside to all of this? Worker A does less work, Worker B loses their job, and Worker A continues to make the same salary and work the same hours they did before. And on the other end, customers and consumers get more SEO spam, more low-quality art, music, tv shows, and books. As if we need more of anything. I'm not against all technology, this just feels like a particularly dangerous development.

View on HN

We should throw out photo copiers and cameras in the garbage as well. After all, people have used them to reproduce the work of others since their inception.

AI isn't creating faithful reproductions of artwork. It's creating strange remixes of prior work without the artist's input or consent.

I thought it's blatant copying that is objectionable in the artistic industries where remixing is everywhere. Also, where does consent go into this? Do I need permission to produce a disrespectful satire (a remix) of the Nike logo?

> Do I need permission to produce a disrespectful satire (a remix) of the Nike logo?

We have developed Fair Use laws over many years and many court cases to answer these kinds of questions. AI tools presently don't understand or care about that.

I don't see why ML tools need to understand this from the law's perspective. The publisher & content producer already has responsibility to verify. There are already tools that allow an author or artist to sue if the work is too similar, or to request a publisher to do a takedown. If generative ML cannot sufficiently distinguish itself from a blatant copying machine, then we already have a system in place to deal with that problem.

But there seems to be the real possibility that ML tools can create art that's sufficiently different, and the degree to which they can generate sufficiently new art is the degree to which they warrant a new policy framework.

IMO the question of an artist's copyright ownership is the wrong framing here with regards to fear of ML. If self driving technology gets better, taxi and truck drivers do not have any informal or formal ownership over the art of driving, but they have the same fear.

Ironically, published remixes of original works (specifically, music remixes) pay a large portion of their revenue to the original artists.

> Do I need permission to produce a disrespectful satire (a remix) of the Nike logo?

You'd probably be fine, but Nike's lawyers could bleed your pocketbook dry to find out. Fair Use being a subjective thing and all.

Sounds like the everyday job of a club DJ!

Which is even better, and not at all a problem.

I don't have to consult with Walt Whitman to write a poem in his style, and nor should I.

I don't have to get consent from Slash to shred on my guitar in his style, and nor should I.

All of human endeavour has been the result of one ape looking at what another ape is doing, saying, "Me like! Me do too!" and then a third ape looking at them both and saying, "Me also like! Me do too, but do little different!"

We're entering the next era of our evolution, people... and its going to be guided and shaped and molded by our technology, just like every other major era of our evolution... from fire, to metallurgy, to the printing press, to the transistor... we're just in the next phase, and like every other phase, a lot of people stand to lose - but many more stand to gain.

There's a growing sentiment among the youth that I'm calling Luddism-Nouveau.

My brother is a public school teacher and he's telling me that the kids really like talking about Ted Kaczynski...

All of this to say; your opinion, while cynical, is not unique.

I don't agree with any of it, by the way. I like that we have modern medicine.

I'd call it Butlerian Jihadism or just Butlerism, but then again, it's kinda niche :p

It's not as niche as you'd think. The kids are talking about the unabomber's manifesto. I'm pretty concerned by that.

Here's[0] an article Bill Joy wrote at the turn of the millennium wherein he quoted the unabomber and spent a lot of time thinking about the power of then emerging technologies. Kaczynski was obviously intelligent (with academic creds too) and had thought through many of the implications of tech long before most others had. His actions were insane, but his malcontentment with technology perhaps was not.

[0]https://www.wired.com/2000/04/joy-2/

I'd be more concerned at any point if kids stopped taking things to extremes and missing out on the nuances of things. It's what kids do.

It's more concerning when in their mid twenties, early 30s, they don't grow up. Which seems to have more to do with the general infantilization arising from a variety of factors, perhaps most notably crippling debt for which they themselves are the pledged collateral, by which of course I mean student loans, which stops them from starting families, owning homes, and generally getting the experience of actually having responsibility for various systems that require having an intimate experience with nuance.

Like I say, I'm not against developing technology. Conflating AI and medicine seems nonsensical. As another example, what about nuclear weapons? Does anyone think that giving humans the ability to self-annihilate was a good development? We did it because we feared other humans would do it first. We would be better off if such weapons simply didn't exist, IMO.

So I understand progress can be good, but sometimes we ought to take a step back and reflect. Unfortunately our societies haven't really developed mechanisms to do that.

>As another example, what about nuclear weapons?

Mutually assured destruction have so far prevented the worst.

It's yet to be seen if nuclear electricity generation is a net positive for humanity, but it could be.

Except that AI is being used to develop drugs right now. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7577280/

None of this exists in a vacuum.

I almost can't believe what I'm reading. Every single leap forward in human history has been because people could literally just do more stuff. The entire industrial revolution happened because what took one person a week could now be done by that same person in a day. AI should hopefully do the same for many things. Sure, there will be people who will have trouble adapting. Just like when we get cars or when the industrial revolution happened.

I don't think it's worth it to leave the future society stunted. I'm very happy I don't live in a cave warmed by some burning dry wood and I don't have to go out tomorrow in freezing rain to hopefully catch the next thing to eat.

Under this framework, all labor saving devices should be thrown out. Guess we'd go back to the days of 90% or whatever of the population being employed in agriculture.

What labor is it saving though? Seems like the bulk of what we can do with AI is create inferior creative works, since we can't trust it to do anything important for the foreseeable future.

Not cynical at all. Nobody asked for this and it's being forced on us by bigcorps who don't give a shit about people, backed by asocial amoral people who don't give a shit about people.

The upside is that in the end, humanity only needs to consist of the 1%. There is no upside for the 99%. But they do not matter.

you say "loses their job" I say "doesn't need to work".

The world has survived many a technological advancement but, if AI were to make it so Worker B "doesn't need to work" going forward, our society is not structured to support Worker B in any meaningful way, and that's something to be concerned about.

Yeah, if somehow Basic Income works out, this will be less concerning. But I think that's the harder problem to tackle.

Not really, we'll do what we already do - leave them to die.

"But oHhHHh noOooOOo, that's sooo awful, cbozeman! We can't do that!"

Bullshit. We do it every single day. Every single day, someone, somewhere, starves to death because they have no marketable skills whatsoever, and because no one cares if they live or die. This is just going to accelerate the process. We'll either get Star Trek, or we'll get that Matt Damon movie.

Maybe not, but they might still need their health insurance.

Reply on news.ycombinator.com