Browse Comments

If California was its own nation, it would be the 5th or 6th wealthiest in the world.

Vote Bernie Sanders to start down the road of fixing our broken for-profit healthcare mess.

I don't think that that many random people have any clue what AI actually is aside from what Hollywood tells them. And that includes Germany.

Fuck us millennial amirite guys?? Right in the middle of finishing 4 years of college and the world's economy goes to shit. Spend the next 10-12 years slowly inching our way back to recovery and then boom Pandemic and Trump fuck us.

Lots of Uyghurs went to Syria during all the chaos. I remember a video of soldiers storming a compound and asking the women where they were and they said China. Also the Uyghurs there were particularly brutal although that’s sort of a given when it comes to ISIS.

Well you couldn’t automate your existing house in that way, probably. But at some point in the future, if you were building a house, it might be possible to create new designs that would incorporate AI-controlled diagnostic and repair capabilities that may or may not be carried out by bots.

>We Don't Understand Why Language Models Work I'm sensing a bit of projection here, because ya, we kinda [do](https://arxiv.org/search/?query=large+language+models&searchtype=all&source=header) understand how and why they work. They didn't just pop out of thin air like no one has any idea what's going on here. “Any sufficiently advanced technology is indistinguishable from magic” You'…

I'm curious -- even if AI is heavily regulated, wouldn't there still be an underground movement developing private models? Or there could be safe havens where this kind of work is tolerated and unregulated... what's to prevent model weights from leaking and spreading to other countries? My personal perspective is that it may be too late to shut open source down... you'd need a ton of political…

Yes, this is like plagiarism but nearly impossible to detect with proper execution. I expect you’ll see a lot of mid-level writers that easily hit their deadlines and get promoted because they actually use AI for first drafts.

>AI looks good on paper to executives, This right here is the crux of all of it. People have generally been focusing on the wrong thing in this whole debate - it doesn't matter if AI is any better at anything than a human, it's that the execs in charge will \*think\* it is and act accordingly, regardless of what happens after that as long as it makes line go up in the short term.

Well, the thread's up voters realize it, at least, haha. That's hopeful. Even if the internet as we knew it is dead, people need that experience of creating and connecting. That's the thing AI enthusiasts don't seem to recognize when they say things like, "Who cares where the content comes from?". Human beings do. That's why we're here. We create to connect. They care, too, even if they don'…

ChatGPT seems like it may have a drug problem as that somehow made the list twice.

So glad I sort my recycling and only use energy efficient lighting and equipment in my house to help slow climate change, and then overnight an entirely new technology comes along that requires 1% of the global power output to operate

for now, we're doing stuff like generating components from figma. solutions that technically already exist but we're kind of reinventing the wheel for our needs. it's simple use cases that would save us a bit of development time for sure. but I would much rather be coding than using AI.

I mean that's sort of like saying that your whole problem with Hitler was him being a murderous demagogue. Putting profits over safety is Altman's entire professional identity. If the concept of negative externality took corporeal form, it would be Sam Altman. Rumor has it that Sam Altman was once asked to help provide RLHF guidance for a model using the trolley problem but had to be removed fro…

You know if you sanction whole world, you’re the one who is sanctioned right? Typical western mindset

It's always "we" when the billionaires need something the government is protecting. "We" need a thriving fracking and shale oil industry for energy independence! Ignore that we're exporting as much oil and LNG as we can sell instead of creating reserves... As soon as they get what they want, "we" are left out in the cold and told we're lazy worthless idiots.

I worked in automation 10 years ago and I started saying this. I had to move on, it was too depressing every day to automate away jobs. It’s way worse now.

Terrible prompt that will likely cause anyone who uses it a lot of problems. First off, the fact your chatgpt sounded like. robot in an existential crisis means you've probably locked it into a "persona" that was misaligned. Check chatgpts latest paper on it [https://cdn.openai.com/pdf/a130517e-9633-47bc-8397-969807a43a23/emergent\_misalignment\_paper.pdf](https://cdn.openai.com/pdf/a130517e-9633…

Of course that only matters if partners read the cases the associates/AI cites lmao /s. Judges of course will and inevitably there will be a ton of judicial orders barring the use of AI until the hallucinations slow down.