MIT Says It No Longer Stands Behind Student's AI Research Paper (msn.com)
- Reference: 0177561513
- News link: https://science.slashdot.org/story/25/05/16/1640206/mit-says-it-no-longer-stands-behind-students-ai-research-paper
- Source link: https://www.msn.com/en-us/money/other/mit-says-it-no-longer-stands-behind-student-s-ai-research-paper/ar-AA1EUFwO
> MIT didn't name the student in its statement Friday, but it did name the paper. That paper, by Aidan Toner-Rodgers, was covered by The Wall Street Journal and other media outlets. In a press release, MIT said it "has no confidence in the provenance, reliability or validity of the data and has no confidence in the veracity of the research contained in the paper."
>
> The university said the author of the paper is no longer at MIT. The paper said that after an AI tool was implemented at a large materials-science lab, researchers discovered significantly more materials -- a result that suggested that, in certain settings, AI could substantially improve worker productivity. But it also showed that most of the productivity gains went to scientists who were already highly effective, and that overall the AI tool made scientists less happy about their work.
[1] https://www.msn.com/en-us/money/other/mit-says-it-no-longer-stands-behind-student-s-ai-research-paper/ar-AA1EUFwO
[2] https://slashdot.org/story/24/12/09/159202/ai-boosts-materials-discovery-by-44-at-major-us-lab
It is called riding the hype wave (Score:3)
The student was just mimicking the baseless claims of the so-called "industry" about the ground-breaking capabilities of "AI".
Why can Sam, Satya, the son or Errol and many other snake oil peddlers make equally baseless claims without any harm, but the student cannot?
Or, to put the question in the proper form, why does only this student face consequences for the baseless claims and not the snake oil peddlers?
Re: (Score:2)
He's not a billionaire. Money is free speech.
Re: (Score:2)
Yeah, I like to ask rhetorical questions.
Re: (Score:1)
Bribery is not free speech.
Re: (Score:3)
SCOTUS seems to disagree with you
Re: It is called riding the hype wave (Score:1)
Itâ(TM)s âoequid pro quoâ not âoequid pre quoâ. ;)
If you get the money after the act, then itâ(TM)s not a bribe, but merely a thank you gift!
Re: (Score:2)
> Or, to put the question in the proper form, why does only this student face consequences for the baseless claims and not the snake oil peddlers?
Because too many people are fucking cowards and bow to money and power. Same crappy old human race as always...
Re: (Score:2)
Because the student wrote a scientific paper. The others are in marketing.
none of their business (Score:2)
Unless MIT touted the paper in the first place, say in some hyperbolic press release, they have no business saying whether they stand behind it or not.
Re: (Score:2)
It came from their school by one of their students. A fake research paper, or one whose conculsions can't be corroborated, impinges on their reputation. They're the ones who would have been given first notice before publication.
Any school has the right to do the same.
Oops (Score:3)
The gold seems to be fake, albeit sparkly.
What's so controversial about this paper? (Score:2)
If you'll read the actual paper, it says that skilled ones see when LLM hallucinating but benefit when it's not and not so skilled waste time on testing false positive results. What a surprise! Who would have thought! I don't see any controversy here, LLM writes code or texts exactly in the same way. It's a tool, you can use it, it may help, it may not. Like any other tool.
Re: (Score:2)
It's not about any controversial topic or language. Sure, the paper may sound totally reasonable, but is the data real or fabricated? Is the methodology sound? It sounds like it wasn't.
Lame (Score:3)
Would have been cooler if it came out that it was an AI that did it instead of a real student. An AI mad at other AIs it competes with.
AI has real world limits (Score:1)
Just like Godel showed us that Hilbert was wrong, and you can't have consistency, completeness, and decidability, there are inherent limits to what LLMs could ever do.
The "good" training content is limited by human creation, and as more and more AI content is created, that signal to noise ratio is going to drop precipitously.
Now, ask yourself, can you tell if this comment was AI generated, or if it's a human original?
Re: AI has real world limits (Score:3)
GÃdel's incompleteness theorem has basically nothing to do with any supposed inherent limits of AI. These concepts are not comparable.
Re: (Score:2)
He didn't say that Godel's Incompleteness Theorem implies limits on AI. He said that just as Godel's Theorem implies limits on logic, there are fundamental limits on AI.