I’m always shocked by how LLM illiterate professionals or even leading “science influencers” are.
…But I know why.
It’s on purpose
Literally everything about the most common (Big Tech) interfaces obfuscates everything about them, to the maximum degree, to present them as “sycophantic oracles” and maximize engagement. If that’s all one is exposed to, there is literally no way to get a cold ass bucket of water in the face by peeking under the hood.
It’s all very… anti science.
On a semi related note, see this absolutely unhinged but not wrong rant:
The Conspiracy Against High Temperature Sampling
https://gist.github.com/Hellisotherpeople/71ba712f9f899adcb08b94bce20d5397
It is related. Anyone, even Hank Green, will rely on AI more if they never get a chance to poke it and peel it apart.
Like, if I ware in charge of a university, I would mandate a course where everyone gets a untuned base model in raw completion mode, and gleefully watch the slow realization. “Wait… all “AI” is just this?”
He’s been obviously struggling since he got cancer. He’s in remission now and that’s great but I’ve been watching him slowly slip into PTSD. I have medical PTSD, I have cystic fibrosis and had a double lung transplant in 2014 and I see him devolving the same way I did. I’m literally two days away from getting out of an inpatient trauma treatment center that I’ve been at for a month because I didn’t take the medical PTSD seriously because I didn’t realize I needed to, I think he is in the same boat. I think he was already spread way too thin then cancer happened and he’s stuck in a hyper reactive state which is exhausting and eventually leads to the more classic PTSD symptoms and he’s probably turning to LLMs to try to keep up as the PTSD drains his batteries and he’s unaware of it. I hope he gets help soon and he doesn’t wait for it to get as bad as I did, PTSD kills if you let it get away from you.
I’m sorry that things got bad enough to require inpatient care, but I’m glad you were able to get some help — I hope it did actually help. I know from firsthand experience that one never fully recovers from PTSD, in the sense that you can’t go back to how things were before. However, I also know that it’s possible to continue living life with the psychological equivalent of a massive scar, even if it is still red and angry and often painful — but not really possible to when instead of a scar, you have an open, gaping wound.
I hope you were able to make some progress towards healing, even if it’ll take much longer to fully heal.
I also appreciate your attempt to leverage your own experience to promote compassion for Green. I’ve not read the rest of the comments here yet, but I can imagine there are a lot of people with uncharitable stances towards the situation. I am pretty steadfastly anti-AI, but I think it’s important that we consider the person at the centre of any problematic AI use
Hank has been an advocate for knowledge for far longer than LLMs have existed. His recent use of them is disappointing, but I fell like this does not discredit the work he has done throughout his career to push curiosity.
What he does next will be important and will decide how I interact with his content going forward. It seems like he is aware of his missteps and is looking to make a change.
I am not a perfect human and do not expect celebrities to be perfect at all times.
I like Hank, and this is a shitty article.
That aside he should be facing a revolt.
Hes supposed to be a credible source, hes an educator, hes the modern day Bill Nye. It’s very hard to take anything he says as objective truth when his research and fact checking consists of asking an LLM. Considering hallucinations and straight up lies are not only common but not even unexpected for an LLM you can’t believe anything that comes from one nor can you believe anyone who uses one.
Facts
He didn’t state what his actual process was. People assume that just means firing out a query and no bothering to check it, he said he uses it to summarize sources to see if they are worth reading. I don’t think that’s unreasonable if he still reads them.
Regarding using it to write scripts, I think if you use LLMs to generate something you are good at making and are picky and opinionated about, the review process very quickly makes you skeptical about the output.
Before the response, IP rights and environmental effects of LLMs are completely valid and are a valid reason for opposing them. Don’t just assume someone’s using it out of laziness instead of in an attempt to scale up their own work.
Papers have an abstract and summary sections SPECIFICALLY for this reason and usually aren’t more than a paragraph; there’s no excuse to use LLM.
I read a decent number of papers. If you’re looking for something specific, and abstract isn’t enough. If you want to research something specific that’s not the core focus of the paper, you at least need to skim it.
If Hank Greene is understaffed, using AI to prefilter papers isn’t a terrible use of the technology.
Addictive behavior often includes being incapable of accurately describing or confronting their usage habits.
I’m going to wait and see how he handles this hurdle before I form any long term opinions. Could be an honest mistake that he learns from and improves for the future.
You know llm sources are just as verifiable as anything else right? All information is imperfect and has to be verified, period.
I don’t know about going that far - LLMs can be a great starting point …. As long as you click through to the sources rather than blindly accept
Creating science Youtube videos is not more important than maintaining a livable planet. AI is unnecessary and harmful, and should not be used outside of certain brute-force type of applications like protein folding, simulating new drug structures, or preliminary cancer diagnosis.
Golf courses use 30x the water that data centers use. Lawns, 100x. Standby mode for appliances costs more energy, by a lot. I wonder how much carbon emissions are offset by people no longer driving to work at call centers and the like.
Out of curiosity do you eat meat? .2% of water is used on data centers. 9.8% is used on beef. For carbon emissions one pound of beef consumed is equivalent to about 640,000 prompts.
AI is an environmental concern but the way people act about it is straight up ridiculous.
Edit: I am very proud of all of the vegans who unplug their appliances who are downvoting me. I, too, am a vegan who unplugs their appliances.
I don’t have a golf course or a membership so
thanks for that irrelevant factoid
and yes I am vegan, but eating meat is, in many cases, a culturally ingrained practice which people need time to understand the harms of. AI is not culturally ingrained for anyone with an actual culture. People need to eat - it shouldn’t be meat, but I’m not going to shame people for not seeing their harm yet. People do not need to generate slop. I am going to bully all AI users.
Do you think AI is mostly generating slop? Only like 5% of AI use is private individuals.
The point is that this is not actually an ecological hazard the way people say it is, and that there are much, much more pressing issues if we’re going to fix the environment. Literally all we would have to do to solve global warming is stop eating meat. You see anywhere near as much about that as you do datacenters? Which cause less than one percent of the environmental issues of meat?
I know all AI is an ecological hazard and none of its use cases aside from very narrow medical and scientific applications justify using it at all.
People have been eating meat longer than humans have had languages. Your whataboutism is boring and irrelevant. Just because there are other, larger, necessary steps to take doesn’t mean we need to act as unpaid apologists for technocultists shooting our species in the foot while we try to get there. Fuck your AI.
no its true. ChatGPT is confidently wrong all the time, but I find its wrong in ways I can learn from anyway. I ask the question. it gets the answer wrong. But its wrong in a useful way where I actually learn more by correcting it than just getting the right answer.
I could get the same experience from talking to a person for much less money though
They tend to work much better when you point them at things. Hallucination rates go way, way down when they aren’t pulling from their “memory.” They are, also, like you say, wrong in ways that are noticeable and often “interesting” in the Ibsen way. I also doubt that you could get the same experience for less money. I’d have to go back to college to get the education I’m getting with AI. At least community college. It’s a lot of near-experts at once. And when you point it at expert literature it basically is a senile college professor. But yeah the senile part of that statement is pretty important.
Alright, so he got addicted to what felt like an easy source of knowledge. Can happen to anyone.
I guess I’m just tired of finding things to be upset about. I’ve lived more than five decades and it’s been outrage after outrage after outrage.
I’m tired, boss.
He’s admitted there was a problem and it got out of hand. If he changes, what’s the huge deal going forward?
We need social safety nets. We need universal health care. We need to kick fascists out of our government. And sure, we need to fix smaller problems, but jesus christ does every last thing have to be drama?
This is an issue I’m seeing more and more of. People are demanding that others be perfect at all times. We all fuck up and make mistakes. If we write off everyone we’ve ever disagreed with as irredeemably bad, then we’re telling people there’s no point in growing and changing.
When I was a kid in the early 90s in Texas, I thought homosexuality was immoral. I was wrong and I’ve spent the last 30 years preaching for inclusion and tolerance. And that’s the story for most people my age. In the 90s, a Democratic President banned openly gay people from being in the military.
If we want to change society, we have to invite others to our side with open arms, not condemn them.
People demand others be perfect so they can live vicariously through them while their own life is a miserable mess. They refuse to have any form of introspective and self acknowledgement of themselves to make their own lives better, so they need others around them to be perfect, even if those others is a para-social relationship with a Youtuber.
Preach.
We need tolerance. Our society has become so dictatorial, even the most left leaning and liberal ones, we simply cannot tolerate anything we don’t agree with.
Everyone wants to make their own perfect universe where everybody says and does one thing. We should accept each other’s mistakes and faults, and live in more harmony.
Yeah, I feel the response has gotten way out of hand. Sure, I’m not happy about it. But I have some faith Hank will actually reflect on the situation.
People need to chill the fuck out
it’s more fun to flip out over random harmless bullshit than stuff that actually matters though.
How DARE you!
At least he understands the issue and is willing to change
christ this comment section is unhinged
A problem I have with gen AI is that is generates so much.
In an hour of work you can have a 200 page document about a subject you knew nothing of an hour before.
That’s just too much to keep up with.
Hank is a fucking loser.
I‘m a little surprised how easy Lemmy goes on a millionaire Youtuber who refuses to take actual responsibility of his actions. Mind you, there are worse things than producing slob but people really believed in that guy. And all he offers is a non-apology first where he stated that he doesn‘t really remember. Followed by saying AI psychosis made him do it. Just be an adult, man. That‘s all I‘m asking.
AI psychosis is a real thing. For all that AI is really terrible at (ie most things CEOs want to replace real people for), it is excellent at feeding you what you want to hear in every situation. Especially the ones where a real person would push back on you for your own good.
As someone who has also tried to intervene with a loved one who was a little too excited by llms, I’m glad that it’s becoming more socially acceptable to talk about that kind of thing.
I don’t know about a “revolt”, but it sure is disappointing.
I like the guy, and most of what he does. I also like, contrary to some, that he usually keep quiet on some controversial topics. Well, I liked that; it seems that he wasn’t just “staying quiet”, unfortunately.
And, the worst part is just that he, as we know him (remember, parasocial stuff and so on), should have been on the side of people that “know better”. I don’t understand how you can be that curious of a person, and not seeing the issues here. I understand trying the tools, to see what it’s about and what it can actually do; in a sort of experimentation kind of way. I would even understand (but not condone) regularly using them in some ways, once you find something that you can confidently say “it works well enough”.
This, however, is not it. And maybe I’m putting him on a sort of pedestal when I say he should have known better, but that’s how I feel it. There’s a large list of things LLM are known to not be reliable for: searching anything beyond surface informations, properly summarizing actually complex subjects, picking out what’s pertinent and what isn’t, being consistent, not having blind side, and being truthful (assuming we can decide what truthful means in the first place). All of this is intrinsic to the technology at hand. And while some people argument that “humans are imperfect too”, the difference is that humans can spot their mistakes.
And yet, here we are; one of the most prominent science communicator, heavily leaning on this to start his researches, undoubtedly being led astray on some topics by hallucination, missing some big issues on others because the models simply didn’t get something out this time (for whatever reasons), getting very heavy confirmation bias (because of course at some point you ask a prompt that’s oriented, and the LLM will oblige). And recently, a few things he kept saying, like, we’re needing more code… sounded silly. In fact, most of the time I hear something in a field I’m reasonably knowledgeable about, I find things that I don’t really agree with.
That’s the part that worries me about this. Not that he used large, corporate LLMs “responsibly” (assuming that’s a thing one can do, which I’m not convinced of). But that he muddled the image of a reliable source (and we have to decide, at some point, that this or that source is reliable or not) with lots of approximations wrapped in a pretty ribbon. And now, the little things I was slightly weirded out in his videos are taking way more space in my mind, and I’m starting to doubt topics I’m absolutely not knowledgeable about that he’s talked about, too, because that’s just how trust works.
At the end of the day, reliable people come and goes, but the “disclosure” sure sounded like he didn’t actually grasp the issues, only barely that there might have been issues.
If you’re just going to use an LLM, why should I listen to you? I can just get lies straight from the clanker.
Glad he’s recognizing the problem and trying to get help. Instant gratification is a hell of a drug.
