Browse Comments

"Industrial-scale distillation" is just the White House's fancy way of saying DeepSeek copied OpenAI's homework to train a cheaper model. The hilarious part is that OpenAI literally argued in court last year that mass-scraping data is considered "fair use" for AI training. I guess "move fast and break things" is only cool until someone moves faster and breaks your IP, huh?

The biggest problem is money. There is just too much incentive to forge on ahead because, if you don't, your opposition will. Also, any government safeguards will be way too far behind to be effective. Another problem may be that the AI will create a situation where they have already taken over and we don't have the mental capacity to realise it. In a way, this may already have happened.

I could see using AI maybe to scan shit even to suggest matches but to think that AI SAYS ITS 100% MATCH THAT MEANS ITS RIGHT is so brain dead. AI is one of the biggest threats to our species right now, it might actually cook us before the inevitable global climate catastrophe.

The AI driver can't unload their own truck. The AI can't do pre-trip safety inspections. The AI can't turn the crank to lower or raise the landing gear on a truck. The AI can't adapt to changing road conditions. The tech isn't there... yet. This will be a disaster.

When the robot inside the truck came out, I was expecting he would say, "Yo! What the f*ck?!".

I think it's really important to remember that AI has been trained completely on existing art, in the earlier days of image generation, I remember being able to clearly recognise distinctive artistic trademarks of particular people I was following on instagram. make a claim about copyright if you like, but where does that line get drawn? what about the thousands of people who've had their art f…

Both are at fault. The distracted driver is a danger to himself and others. These autonomous car companies are all enabling these dangerous people to be even more of a danger to everyone. Our roads should not be a trial zone for unproven technology.

When the first complaints about AI art and copyright infringments started people were quick to judge that the art was uploaded onto the internet out of the artists free will, and that how it is used by AI would be equivalent of someone maybe copying it and altering on photoshop. What the screenwriters and actors strike shown us is that even though there was no laws that would work retroactivelly …

The person who just violently murders the Ai or traumatise them for no reason: that’s me

0:15 Yo why are the AI images with _more_ poisoned training less bad then the ones with _less_ poisoned training?

This is teaching the kids to not bond with humans, but to bond with AI!! This cant be good!

This video breaks it down perfectly! Feels like the rapid society that we live in now has created a culture where everything has to be instant and available - AI fulfils that need for people unfortunately. They don’t have to spend thousands of hours honing their craft and unique style like us artist because they can just enter prompts and get infinite styles and iterations. It’s sad how some peop…

omggg real. Like I once searched up a roaring cougar and got a bunch of ai slop.

Homework was created for this reason. To teacher people to work after hours. I fight all homework. My kids need to play outside. Our school has limited homework

You hit the nail on the head man. I've been studying this for 4 years in undergrad and I always tell people that it's not IF climate change will happen, the time for that is over, the question now is just how bad is it going to be.

If you can 3D print it cheaply, it will be more profitable for poachers to shift to printing

> It is an advanced economy **with a functioning democracy**. Only having a borderline functional democracy since last year when the previous president was kicked out and in prison under the influence of a [female Rasputin](https://www.efe.com/efe/english/world/south-korean-court-sentences-female-rasputin-to-20-years-in-corruption-case/50000262-3522095). But seriously, South Koreans usually p…

From the sound of the article they threw in a ton of resumes and trained the AI with the resumes of people they hired. Really the AI was picking up on the bias of the hiring personnel if the only reason it could tell why the resume was not chosen was because it had the word woman or an all woman's college on it. This really sounds like a garbage in garbage out issue... Since I don't really think …

Yeah, I'm working on the GSA AI one right now. I'm still getting my salary, but if we win the team splits $50,000

Funny you should mention that. Have you tried Chat-GPT-LangChain which is gpt-3.5 but when it doesn't know something it can access "tools" like the internet or wolfram alpha. The idea is that wolfram is very complimentary to gpt-3.5. I have to say it interesting to use. I asked it to summarize what the senate did yesterday. Then asked it was it thought was the most important. It said the un…