Computer Science on Yik Yak | i honestly think ai has done more harm than good. I would say the government should regulate it but, they would abuse it at this rate

Yik Yak

Download

Join communities on Yik Yak Download

Computer Science 8w

i honestly think ai has done more harm than good. I would say the government should regulate it but, they would abuse it at this rate

Anonymous 8w

The only reason why a government would want to regulate it is make it harder for others to get into the industry because the big players are much just bankrolling them. Regulating would only make competition harder for the small guys, and barely affect the large players (and therefore benefit them)

Anonymous 8w

Yeah regulation would be a terrible idea. In general when govts get involved in things they don’t need to, those things get worse; more expensive, less efficient, lower quality, less innovation, less transparency, more fraud, etc.

Anonymous

#2 8w

I saw they were tryna limit how many floating point operations you can use to train a model, it was a high number iirc but still kinda bs

Anonymous

#3 8w

Yeah one of the reasons why it’s a terrible idea is bc they don’t have the technical understanding to even try to regulate it in a way that makes any sense.

Anonymous

#2 8w

Usually (and historically) they get around the expertise issue by finding experts in the field to participate in what's essentially a court case

Anonymous

#1 8w

This is true in part but it doesn’t mean regulations aren’t needed. That why OpenAI and other big companies were pushing certain regulations that favor big companies. However, big or small, they shouldn’t be able to plagiarize books or develop tools that cause people and especially teens to engage in self harm.

Anonymous

#5 8w

Bruh why shouldn't they be able to pirate books, you need big data

Anonymous

#3 8w

Piracy puts all companies at the same level and avoids a monopoly on having ai training data

Anonymous

#3 8w

Because there needs to be ethics? like any other field

Anonymous

#5 8w

The problem with that is we’re in a race with China and they don’t use ethics. They are pirating as much as they can get away with so slowing the US down with that is risky. That’s why it’s not obvious bc ofc we need ethics and people should be paid for their property but slowing them down is a huge risk

Anonymous

#4 8w

Not in practice. Often the “expert” is biased or corrupt or doesn’t know what they’re talking about. Exhibit A would be the exactly situation this thread is about.

Anonymous

#2 8w

Yeah ik that too. I still don't think the answer is to let rich capitalist wannabe oligarchs "regulate" themselves here

Anonymous

#4 8w

Yeah I agree but I don’t know what the solution should be here bc govt regulations dramatically slow growth. What do you think the solution should be?

Anonymous

#2 8w

I'm not a capitalist. (More specifically, I’m not pro-growth at the expense of human lives and well-being and quality of life.) So I'm okay with slower growth and overregulation that gets undone over time, personally.

Anonymous

#2 8w

No regulations

Anonymous

#5 8w

It’s not about ethics training an LLM is a transformative work, also cs majors pirate all the time because Blu-ray > Streaming

Anonymous

#3 8w

That's what we already have and it’s shit (in terms of actual quality and research, how ai is being used, how companies are trying to monetize it, and how few people in the field can now get any kind of entry-level job). Like, all these problems are distinctly not going to be solved by continuing to do what we're doing

-1

Anonymous

#4 8w

Idc if entry level jobs get killed all I care about is cs and entry level cs jobs going away is just fear mongering, most of yall can’t get jobs bc yall don’t want to move to the big cities

Anonymous

#4 8w

Oh then I think that would be a completely separate conversation bc I’m talking about AI regulations in the US as it stands, and you’re talking about AI regulations in an alternate universe where the US is fundamentally completely different from how it is.