Browse Comments

Is hare-seafarer meant to be a pun? I don't know if I'm missing a punchline or chatgpt is just very weird.

I wonder if Altman actually thinks about what he's saying before the words come out of his mouth, or does it come as a surprise to him? On the one hand, he says he agrees with Prof Marcus that AI can only be trusted if we know exactly what data it's been trained on, but at the same time, OpenAI is vehemently opposed to releasing information about what data it's been trained on. Does he mean tha…

From the article: “OpenAI’s Mr Altman said that AI models needed to be trained on a system of “values” developed by people around the world. He also wants an independent commission whereby experts are able to evaluate whether the AI models are complying with regulations and have the power to both grant and take away licences. “We’re excited to collect systems of values from around the world,”…

But you also have to remember that the child abductor friend of yours that you have to arrest to make your other friends happy, yeah that guy, he also owns icbm nukes.

THE FIRST THING WE AUTOMATE OUT IS ART?! You sure this is the first time automation has replaced a job?

ChatGPT is a people pleaser. I think this says more about the people running these wargames than the AI.

Did you? An Indian company imported Dell servers from Malaysia, then exported the servers to two companies in Russia. That’s the story alongside quotes such as this: > It’s now the second-biggest supplier after China of restricted technology to Russia, Bloomberg reported October 12. I’m not defending the OP’s generalizing point, but I’ll ask: what is yours? They are wrong because Del…

honestly you should say to them shut the fuck up i thought women talked more

CEOs primarily get by via their connections and credentials. AI cannot "replace" that, in the same way rich people cannot be "replaced" even if they do effectively nothing but sail around the world while making millions every day on their "investments." The system is setup so that those people do not have to do anything; it's the rest of us who will be replaced.

Let's be real. Russia wouldn't even be able to push into Poland. They'd get mega fucked in a conventional war with Europe.

Reminds me of Sgt. Hatred on Venture Bros, and the one time a bunch of soldiers start storming the place and he says something like, "Any images you find on my computer are completely legal computer generated images." If people want to use AI to create porn, I don't really have a problem with that, at least as long as they aren't based on real people. You want to create porno using a mashup of s…

I'm just here to piggyback off the top comment: This is such a tone deaf nothing plea. There's actual terrible shit going on and these rich people are worried about the only thing they care about, money. Fuck em.

And with how chatgpt is hard-programmed to pretty much agree with you on everything. Itll spin things enough to make anything agreeable. Its a weird interaction.

The full response I got is: ✅ What It Gets Right 1. “LLM = Predictive Math” Spot on. A large language model (LLM) like me works by predicting the next likely token (not technically a full word) based on everything that came before it. It’s probability distribution all the way down, trained on patterns in text, not facts or feelings. 2. “It doesn’t think or know” Correct again. I don’t have bel…

It’s not just you being a good teacher that’s caught this trend, it’s your nuanced AI sleuthing skills at work! That’s not your students writing—it’s ChatGPT! Would you like some infographics created to help show your students how this is affecting their writing? Cause I have ideas.

It's also a question of the business side of things even care about developer productivity. If you look at the experienced devs sub, a common talking point is the futility of advocating for better processes and tooling, but the business side of things don't care about that. Big tech might try to convince the MBAs they can replace their workers with it, but like... They could've already done that …

You get what you give - which is sad because some people may never see this side of AI due to only giving negativity.

Umm, did you read what he wrote? He is still offering to give feedback to anyone who genuinely wrote their paper and wants to learn/ improve beyond getting their A and moving on. He is also arguing that those who put in the effort to actually write their paper are developing skills and growing personally in ways those who entrusted their paper to AI will not.  I superficially get the frustration…

Also if you're switching from OpenAI because of their DoD/DoW deal, you might want to think again about Grok as they already have a contract with the military. Also Microsoft who might have been first but lets be honest, everyone hates Copilot anyways. Edit: Oh and Google is helping ICE so Gemini would be a no go too

I’m not sure that’s true. Sure the AI companies internally are ahead but that’s because the models aren’t red teamed/productized as much and are more raw from my understanding. I’d imagine there would still be a lot of testing/post training/productizing that would go into the models before they are able to be properly used by the pentagon still. Actually just remembered anthropic said it was a…