Browse Comments

I love to see this shit unfolding. The billionaire who built his empire about hype bubble of "BILLIONS DOLLARS NEEDED FOR AI" destroyed by a little lab that spent only $5M open sourcing for everyone.

I draw a parallel between AI and Google. Google enabled us to retain less information because it’s at our fingertips. AI will also free up mental energy we’d use on lots of things. I’m interested to see where it leads.

That's quite easy to do. Go into your settings and click on "Customise ChatGPT". Here's what I have under "What traits should ChatGPT have?" Like a person, casual and with natural back and forth conversation, not saying "I'm here to serve" or anything like that. Thoughtful and considered, not being overly confident of answers, saying "I don't know" when appropriate, offering different perspect…

Yeah Australia tried this shit with a thing called Robodebt. Essentially we tried to automate welfare payments but it was fucked from the beginning and never legal. People are slowly being held accountable for it, but not as much as should be.

$5,000–$15,000 to not even sell your house, and Arkansas has a housing shortage so they will probably sell it tomorrow and they still get OP's money

I majored in philosophy of mind at uni and the gold standard was passing the turing test. ChatGPT blows through that so now goalposts hastily being moved. I’m old enough to remember being taught in school that humans were different as they were the only animal that could use tools. Just saying that we seem to want to have the comfort of thinking we are ‘superior’ There’s pretty strong arguments …

The article says: > “The filter relies on manually curated open-source blocklists, including the ‘nuclear’ list, provided by uBlockOrigin and uBlacklist Huge AI Blocklist,” DuckDuckGo said in a post on X. “While it won’t catch 100% of AI-generated results, it will greatly reduce the number of AI-generated images you see.”

I am so glad I am not the only person who uses ChatGPT for this! It's so helpful and enjoyable to use it for brainstorming, plotting, et al. 5 showed up for me for a bit last night then now it's back to 4o.

As someone who partly (aside for creative writing, that is now dead), does use gpt for mental health problems (and I have no problem with GPT5 doing this, I do like GPT 5 tone better, as I am triggered by, "you are special, you are unique, you are valid, you are not broken" and get angry), I feel like the message is this: "Do not overwhelm your friends with your poblems." So I go to Chatgpt to no…

Generative AI is glorified autocomplete, and stupid people are really putting their full trust in it without bothering to verify sources before spreading the misinformation it spreads.

How to test students’ ability to think and learn in the world of AI is the problem.

It should be banned primarily because it's being used to distribute AI deepfake and child sexual abuse material. The platform is doing nothing to curb this. Completely unacceptable and immoral behaviour that justifies completely banning Twitter.

surely these tech monopolies won't use all this training data for evil /s (of course they will lol)

Please don't open in Bangalore. The city infra is fucked already thanks to the IT boom and the most corrupt state government in the country.

A lot of people don't care about the person but the product. If AI is ever able to produce a completely believable replica of their favorite actors then you might have a hard time getting people to turn against it, or the idea of it. An AI replica of your favorite actor that never ages. It would prolong content indefinitely. Honestly, I could see a world where actors allow for AI film companies…

Lmao facts, auntie Jackie taught me how to block people for the slightest reason. 🤣 If you breathe wrong you’re exiled from my space and attention. 

Anthropic walked away from that deal because they weren't comfortable with the Pentagon's request and they were blacklisted for it. OpenAI stepped right in and said "oh yeah we'll do mass surveillance and autonomous weapons targeting no problem"

this 100% resulted in me canceling my OpenAI subscription. I may just be a tiny fish.. but I'm swimming to a different pond. Anthropic's pro plan is cheaper, anyway..

I'm a software engineer who works with AI daily, and ~90% of my code is written by AI. That does not imply "minimal human oversight". I read every single line of the code before putting it out for review (often requesting many changes, because the AI isn't as smart as a human), and then another engineer has to review the code before it lands. Another AI also reads over the code and notes anythin…

It’s the opposite. He believes he’s so important that even an AI version of himself would be better than delegating responsibility to another actual human. Assuming this is actually true the hubris is astounding