Suchir's Essay on Fair Use | Hacker News Reader

Suchir's Essay on Fair Use

suchir.net

45 pointshgarg14 comments

I never knew him, but it’s incredibly saddening to see someone around my age take their own life, or at any age for that matter. He seemed like he was a bright soul who fought for what he believed was right. We need more people like that. They’re the ones who inspire others to enact change, even if we disagree with them. Rest easy, Suchir.

I think beyond technicalities, the spirit of "fair use" is to allow individuals to build upon the work of others, provide critiques, and clarify existing issues. And the reason why this is a "good thing" is that in society, we want everyone contributing their ideas to make an intellectually better world.

In short, the true spirit of fair use is about people sharing their responses to the ideas of others. And the ultimate reason for that is that we want a free interaction of people.

AI should not be considered under fair use at all because it doesn't propagate the responses and reactions of people -- technically AI, creates, but since the creations of AI are not expressions of the experience of a person or group of people. Instead, the only goal of AI is to take away the opportunities that people have to express themselves because those opportunities often came with a small payment.

Therefore, even though at times AI might comply with fair use (and that's even be argued against here), whether AI violates fair use technically is not even the most relevant question. The question is, does AI violate the spirit of fair use? And the answer is unquestionably, yes. Fair use was designed to increase the production of works that reasonably include samples from other works for the express purpose of encouraging human creativity. HUMAN.

AI is just a tool that big tech uses to tip the balance of society and take more than their fair share of wealth. It's a shiny toy that is amusing but not really of much benefit to society and it uses the technicalities of the law (such as fair use) to crush its true intent for selfish gain.

> It's a shiny toy that is amusing but not really of much benefit to society

Sorry, maybe you haven't figured out yet how to apply it, but it is a very useful tool already for many people.

I don't want to apply it. And of course it is useful to many, but that utility comes at a cost to greater society and the utility itself is questionable: it many help people in their microcosm in the short-run but in the long-term it is only isolating people and making them more machine like.

Maybe people just don't like to talk to other people, hence everything is behind websites and APIs now, just so that human interaction is minimized.

You may oppose this. But the best move from a game-theoretic perspective, is not to not play.

> AI is just a tool that big tech uses to tip the balance of society and take more than their fair share of wealth.

Additionally, like it or not there will also be permanent job destruction / replacement with many naïve AI supporters suggesting utopian nonsense such as "UBI" for everyone (it has never worked nor was it economically sustainable).

Let's call it for what it really is: tools such as ChatGPT (OpenAI), Devin (Cognition), Computer Use (Anthropic) and Fin (Intercom) are here to eliminate jobs. The worst part is that there is almost no alternative for those lost jobs and they are not hiding that fact.

If having less jobs is so great and "benefits humanity", then why are there more folks here and across the world worried about their future job prospects?

First ChatGPT came for journalists and we said "Learn to code".

Then Gen-AI image generators came for digital artists and we said "Learn to code".

Then music generators came for musicians and voice actors and we said "Learn to code".

Then Copilot and Devin came for senior software engineers and AI engineers are now telling everyone:

Learn to cope, this is for "humanity"

> Let's call it for what it really is: tools such as ChatGPT (OpenAI), Devin (Cognition), Computer Use (Anthropic) and Fin (Intercom) are here to eliminate jobs. The worst part is that there is almost no alternative for those lost jobs and they are not hiding that fact.

Absolutely true, but it's more than that. AI itself is ultimately the next step towards completely autonomous technology. Big tech companies and programmers are just taking advantage of the fact that technology still needs us. They are using that to get ahead in life while putting the entire planet at risk.

Copilot and Devin never came for senior software engineers, you can't give that to a non-coder and expect it to do the work.

Due to the way we use code nowadays, I'm convinced we would probably replace the CEO itself before the devs anyways.

> Copilot and Devin never came for senior software engineers, you can't give that to a non-coder and expect it to do the work.

True now, but not true in 10 years. Or 20 years. The entire point is the total end result, and that's what we should be concerned about.

Maybe they will get exponentially better, I can't predict the future. Code generating software with a UI isn't exactly a new idea though, we said the same about Frontpage.

Indeed, but as with all unchecked developments, they eventually improve and sometimes they can improve too much. Like survellance tech -- too much improvement in my opinion. So when it comes to something really important like the future of jobs, perhaps we should be more cautious.

I'm definitely okay with this although historically, none of the technological revolution did actually collapse the employment market, they freed up future jobs which were impossible economically previously.

Not saying that can't happen, there's no hard science about this but when we experienced as many technological revolutions as we have, betting the opposite of what happened previously might not be the most likely option.

Look, an LLM is a tool to transform a string in one language into a string in another. We have a name for that: a compiler.

LLMs are magical but all that they mean for SWEs is that there's one more compiler to use - and this one is unreliable and opaque to boot.

Until AGI is solved, software engineers are safe. Our job is the most fundamentally human of all: telling the machine what to do.

Does UK or EU have "fair use" copyright laws? Also, it's a question of Pro-Privacy vs Pro-AI-innovation; finding the right balance between AI development and privacy is crucial.

Reply on news.ycombinator.com