Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Prism

openai.com

Post reply on HN
Print view View original

231–240 of 543 posts

Re: Prism

#231

Post by toomuchtodo »

Cornbilly wrote: post #227
toomuchtodo wrote: post #185

Earlier quoted context omitted.

AI slop security reports submitted to curl - https://gist.github.com/bagder/07f7581f6e3d78ef37dfbfc81fd1d... HN Search: curl AI slop - https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

This one is hilarious. https://hackerone.com/reports/3516186 If I submitted this, I'd have to punch myself in the face repeatedly.

The great disappointment is that the humans submitting these just don’t care it’s slop and they’re wasting another human’s time. To them, it’s a slot machine you just keep cranking the arm of until coins come out. “Prompt until payout.”

Re: Prism

#233

Post by bonsai_spool »

epolanski wrote: post #197

Not gonna lie, I cringed when it asked to insert citations. Like, what's the point? You cite stuff because you literally talk about it in the paper. The expectation is that you read that and that it has influenced your work. As someone who's been a researcher in the past, with 3 papers published in high impact journals (in chemistry), I'm beyond appalled. Let me explain how scientific publishing works to people out o…

> But if asian researchers (yes, no offense), were already spamming half the world papers with cheated slop (non reproducible experiments) in the desperate bid of publishing before, I can't imagine now.

Hmm, I follow the argument, but it's inconsistent with your assertion that there is going to be incentive for 'proper work' over time. Anecdotally, I think the median quality of papers from middle- and top-tier Chinese universities is improving (your comment about 'asian researchers' ignores that Japan, South Korea, and Taiwan have established research programs at least in biology).

Re: Prism

#234

Post by jedberg »

reassess_blind wrote: post #113

Do you think they used an em-dash in the opening sentence because they’re trying to normalise the AI’s writing style, or…

> because they’re trying to normalise the AI’s writing style,

AIs use em dashes because competent writers have been using em dashes for a long time. I really hate the fact that we assume em dash == AI written. I've had to stop using em dashes because of it.

Re: Prism

#235

Post by bonsai_spool »

The example proposed in "and speeding up experimental iteration in molecular biology" has been done since at least the mid-2000s.

It's concerning that this wasn't identified and augur poorly for their search capabilities.

Re: Prism

#236

Post by trees101 »

cryzinger wrote: post #158
SchemaLoad wrote: post #143

Earlier quoted context omitted.

GenAI largely seems like a DDoS on free resources. The effort to review this stuff is now massively more than the effort to "create" it, so really what is the point of even submitting it, the reviewer could have generated it themself. Seeing it in software development where coworkers are submitting massive PRs they generated but hardly read or tested. Shifting the real work to the PR review. I'm not sure what the fin…

More relevant than ever: > The amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it. https://en.wikipedia.org/wiki/Brandolini%27s_law

The P≠NP conjecture in CS says checking a solution is easier than finding one. Verifying a Sudoku is fast; solving it from scratch is hard. But Brandolini's Law says the opposite: refuting bullshit costs way more than producing it.

Not actually contradictory. Verification is cheap when there's a spec to check against. 'Valid Sudoku?' is mechanical. But 'good paper?' has no spec. That's judgment, not verification.

Re: Prism

#237

Post by ggm »

A competition for the longest sequence of \relax in a document ensues. If enough people do this, the AI will acquire merit and seek to "win" ...

Re: Prism

#238

Post by cauliflower2718 »

jeffybefffy519 wrote: post #58

I postulate 90% of the reason openai now has "variants" for different use cases is just to capture training data...

ChatGPT lets you refuse to allow your content to be used for training (under Preferences -> Data controls), but Prism does not.

Re: Prism

#239

Post by bariswheel »

I used overleaf during grad school and was easy enough, I'm interested to see what more value this will bring. Sometimes making less decisions is the better route, e.g. vi vs MS word, but I won't speak too much without trying it just yet.

Re: Prism

#240

Post by agnishom »

pickleRick243 wrote: post #221
JBorrow wrote: post #32

From my perspective as a journal editor and a reviewer these kinds of tools cause many more problems than they actually solve. They make the 'barrier to entry' for submitting vibed semi-plausible journal articles much lower, which I understand some may see as a benefit. The drawback is that scientific editors and reviewers provide those services for free, as a community benefit. One example was a submission their und…

I'm curious if you'd be in favor of other forms of academic gate keeping as well. Isn't the lower quality overall of submissions (an ongoing trend with a history far pre-dating LLMs) an issue? Isn't the real question (that you are alluding to) whether there should be limits to the democratization of science? If my tone seems acerbic, it is only because I sense cognitive dissonance between the anti-AI stance common am…

> whether there should be limits to the democratization of science?

That is an interesting philosophical question, but not the question we are confronted with. A lot of LLM assisted materials have the _signals_ of novel research without having its _substance_.

Post reply on HN