PhD Level expertise:
There’s more: https://lemmy.world/comment/18699894
🅱️lue🅱️e🅱️ry
to be fair, that’s a hard word to spell
Can confirm
interesting…
I got the correct number of times.
guy selling stuff, says his stuff is the best stuff.
more news at eleven.
“GPT-5 is the first time that it really feels like talking to an expert in any topic, like a PhD-level expert.”
Yeah, feels like. Not actually examples of thinking and doing things at that level.
“These systems, as impressive as they are, haven’t been able to be really profitable,” … “There is a fear that we need to keep up the hype, or else the bubble might burst, and so it might be that it’s mostly marketing.”
That’s the painful truth. No profit, a lot of hype and a market in a 2008 financial crisis bubble.
Awesome. My only critique is that microwave ovens actually work really well in their niche. I can’t say the same for LLMs.
I chuckled
Salesman gonna sell.
Altman is quite good at it actually. Remember when he was saying how scared he was of his own AI. Or calling for increased regulation because their models are just sooo good that government has to nerf them.
It helps that the media propagates everything he says as if it is truth when he’s obviously lying like 80% of the time.
He’s another Musk, who’s cars have been running completely driver-less from US coast to US coast since 8 years now.
I know too many PhD’s for that to impress me
Right? I still remember the bollocking I got from a professor in front of a class about the awful state of classroom equipment, all because the man couldn’t find the PHD (push here, dummy) button to turn the computer on…
You can tell this is marketing fluff, because GPT could already provide “PhD-level expertise” - just in a hit-and-miss fashion that you couldn’t rely upon without some other form of verification. So how is this different?
At carbon footprint levels approaching the US military coming soon!
You’re on the Internet. Should we shut that off as well?
How are you any better with coming online to shit post and look at memes.
The datacenters used to run websites and social media are not any better.
Are you trying to say that making a series of http requests to view a website is even remotely equivalent in energy usage compared to running inference with an llm model???
Yes since you’re not the only one doing it. The impact on climate is substantial in both cases and people are absolute hypocrites for not realizing this. You can’t come online and shit post all day about AI without realizing you’re using services that also contribute to The same if not worse impact
Great Whataboutism.
Great that you know it’s true and this is all you can say
They have stolen more PhD level work to dump into the training model?
Don’t need to steal a lot of journals made deals.
Part of what makes these models so dangerous is that as they become more “powerful” or “accurate”, it becomes more and more difficult for people to determine where the remaining inaccuracies lie. Anything using them as a source are then more at risk of propagating those inaccuracies which the model may feed on further down the line, reinforcing them.
Nevermind the fact that 100% is just statistically impossible, and they’ve clearly hit the point of diminishing returns some time ago so every 0.1% comes at increased cost and power. And, you know, any underlying biases.
Just ridiculously unethical and dangerous.
Geriatric senile PhD on too many painkillers whose area of expertise was a pseudoscience like phrenology before it was rejected, maybe.
