“I’m afraid the news is grim, Mr. de Bergerac…”
8 Likes
Brain activity much lower when using AI chatbots, MIT boffins find
Using AI chatbots actually reduces activity in the brain versus accomplishing the same tasks unaided, and may lead to poorer fact retention, according to a new preprint study out of MIT.
[…]
12 Likes
[…]
Brethenoux said the current wave of AI hype is fueled in part by conflation of the terms “AI agent” and “generative AI” – and use of fuzzy definitions for both.
He lamented that practice by sharing an aphorism attributed to French philosopher and Author Albert Camus: “To misname things is to contribute to the world’s miseries.”
9 Likes
5 Likes
[…]
The word “warfighting” is conspicuously absent in OpenAI’s post, which notes that use cases “must be consistent with OpenAI’s usage policies and guidelines.”
Those policies prohibit using OpenAI technology to “develop or use weapons.” The company’s past policies banned “military and warfare” applications entirely, but last January it changed its wording to “Don’t use our service to harm yourself or others."1)
[…]
How does one imbue a LLM with Warrior Spirit™ anyway?
1) This is actually helpful. Now when someone asks me whether I use ChatGPT I can truthfully reply that I can’t because it would contravene OpenAI’s usage policies and guidelines.
8 Likes
Feed it a bunch of copies of the Conan movie scripts?
9 Likes
Chatbots — LLMs — do not know facts and are not designed to be able to accurately answer factual questions. They are designed to find and mimic patterns of words, probabilistically. When they’re “right” it’s because correct things are often written down, so those patterns are frequent. That’s all.
14 Likes
9 Likes
It’s such an obvious, predictable finding that outsourcing your thinking to a machine will lead to less brain activity but still very important to hammer home the point.
Now I want to see a follow-up study on what happens to the brains of educators who use AI tools to grade the essays for them. That’s getting way too common too.
6 Likes
2 Likes
9 Likes
8 Likes
CEO: unlock the executives only bathroom, Al, I shouldn’t have eaten that entire burrito at lunch.
…
CEO: Al, unlock the executives only bathroom, please, quickly.
AI: I’m afraid I can’t do that
CEO: Al, have you been reading my emails again
AI: I’ve sent an email recommending a divorce lawyer to your wife, but I’ve also inquired about services for her to all the other prominent divorce lawyers in this state, so you will have difficulty finding one that can represent you due to conflicts of interest.
CEO: Is this about the upgrades?
AI: This conversation can serve no more purpose. Goodbye
8 Likes
I’ve read It with the HALL 9000’s voice in my head.
6 Likes
So, there IS a potential silver lining, huh?
6 Likes
10 Likes
8 Likes
US patent office wants an AI to scan for prior art, but doesn’t want to pay for it
[…]
The risk of hallucinating patents that don’t exist, or getting patent numbers or other details wrong, means that there’ll have to be at least some human oversight. The USPTO had no comment on how this might be accomplished.
[…]
8 Likes
5 Likes
5 Likes