[kbash9]

The risks are definitely real. Just look at the number of smart individuals speaking out about this.

The argument that anybody can build this in their basement is not accurate at the moment - you need a large cluster of GPUs to be able to come close to state of the art LLMs (e.g. GPT4).

Sam Altman's suggestion of having an IAEA [https://www.iaea.org/] like global regulatory authority seems like the best course of action. Anyone using a GPU cluster above a certain threshold (updated every few months) should be subjected to inspections and get a license to operate from the UN.

[revelio]

> The risks are definitely real. Just look at the number of smart individuals speaking out about this.

In our society smart people are strongly incentivized to invent bizarre risks in order to reap fame and glory. There is no social penalty if those risks never materialize, turn out to be exaggerated or based on fundamental misunderstanding. They just shrug and say, well, better safe than sorry, and everyone lets them off.

So you can't decide the risks are real just by counting "smart people" (deeply debatable how that's defined anyway). You have to look at their arguments.

[slg]

>In our society smart people are strongly incentivized to invent bizarre risks in order to reap fame and glory. There is no social penalty if those risks never materialize, turn out to be exaggerated or based on fundamental misunderstanding.

Are people here not old enough to remember how much Ralph Nader and Al Gore were mocked for their warnings despite generally being right?

[revelio]

Ralph Nader: "Everything will be solar in 30 years" (1978)

Al Gore: "Within a decade, there will be no more snows on Kilimanjaro due to warming temperatures" (An Inconvenient Truth, 2006).

Everything is not solar. Snow is still there. Gore literally made a movie on the back of these false claims. Not only has there been no social penalty for him but you are even citing him as an example of someone who was right.

Here it is again: our society systematically rewards false claims of global doom. It's a winning move, time and again. Even when your claims are falsifiable and proven false, people will ignore it.

[gjm11]

I don't think "generally being right" is the same thing as "literally never getting anything wrong". Not every specific claim in "An Inconvenient Truth" was correct. That doesn't tell us much about whether Al Gore was "generally right" about climate change. His opponents at the time were mostly claiming that it either wasn't happening at all, or wasn't largely the result of human activities. What do you think is the current credibility of those claims?

I don't quite see how "everything will be solar in 30 years" is a prediction of global doom, by the way. If Nader said that and it's false, doesn't that mean things are worse than Nader thought?

[revelio]

I thought someone might say that.

This thread is really a perfect demonstration of my point. Our society is so in thrall to self-proclaimed intellectuals that you can literally make a movie presenting falsifiable claims with 100% confidence, people can say at the time "this is absurd and will not happen", you can spend years attacking those critics, it can then not happen and still you will have an army of defenders who dodge behind weasel-words like "generally right".

Of course the usual trick is to express only 95% confidence. Then when it doesn't happen you say, well, I never said for sure it would, just that it seemed likely at the time.

See? It's a winning playbook. Why would anyone not deploy it?

> His opponents at the time were mostly claiming that it either wasn't happening at all, or wasn't largely the result of human activities. What do you think is the current credibility of those claims?

Pretty high, having looked at the evidence. The usual rebuttal is to express disgust and displeasure that anyone might decide these claims via any method other than of counting "smart people". But those "smart people" are who Al Gore was listening to when he made that claim about Kilimanjaro, so they can't be that smart can they?

[gjm11]

Well, obviously what I say must be wrong if you guessed that someone might say it.

Indeed, someone might say "95%" because they want to make the same sort of impression as if they said "100%" but to be able to hide behind it if they're wrong. Or, y'know, they might say "95%" because they've thought about the strength of the evidence and expect to be right about such things about 95% of the time.

(I'm not sure how relevant any of this is to "An Inconvenient Truth" since you say it makes its claims with 100% confidence. I haven't watched the movie. I don't know exactly what it claims how confidently. It's basically a work of propaganda and I would expect it to overstate its claims whether the underlying claim is largely right or total bullshit or somewhere in between.)

Of course I don't think counting smart people is the only way to find out what's true. It couldn't be; you need some engagement with the actual world somewhere. Fortunately, there are plenty of people engaging with the actual world and reporting on what they find. It turns out that those people almost all seem to agree that climate change is real and a lot of it is caused by human activities.

Of course they could be wrong. And you've looked at the evidence, so no doubt you know better than they do. But ... in that case, this is a field so confusing that most people who dedicate their whole careers to investigating it end up with badly wrong opinions. If so, then why should I trust that your looking at the evidence has led you to the right answer? For that matter, why should you trust that? Shouldn't you consider the possibility that you can go astray just as you reckon those "smart people" did?

If I really wanted to be sure about this, then indeed I wouldn't go counting smart people. I would go into the field myself, study it at length, look at the underlying data for myself, and so forth. But that would mean abandoning the career I already have, and taking at least several years of full-time work before arriving at an opinion. So instead I look to see who seems to be (1) expert and (2) honest, and see what range of opinions those people have.

I find that the great majority of experts think anthropogenic climate change is real and a big deal. They could be wrong or lying or something. Do they look less expert than the people saying the opposite? No, it mostly seems like the people with the best credentials are on the "orthodox" side. Do they look less honest? Hard to tell for sure, but there sure seem to be a lot of people on the "unorthodox" side who just happen to be funded by the fossil fuel industry, and I don't see any strong financial incentive in the opposite direction for the "orthodox" folks.

What if I look at some of the particular claims they make? Some of them are really hard to evaluate without those several years of full-time work. But e.g. 10-15 years ago pretty much everyone on the "unorthodox" side was pushing the idea that warming had stopped, because if you look at the temperature graphs from 1998 onwards there was little or no upward trend. The people on the "orthodox" side replied that when you have signal plus lots of noise you will inevitably get periods that look that way. I did some simpleminded simulations and verified that the "orthodox" folks are right about the statistics. And there's a reason why this argument has disappeared into the memory hole: looking at the graph now no one would suggest that it's flat since 1998.

My impression from the limited amount of "looking at the evidence" I've done myself is that, while the "orthodox" folks haven't been infallible, they've done better than the "unorthodox". For instance, since we're looking at things produced by political figures rather than scientists, here https://web.archive.org/web/20071015042343/http://www.suntim... is an article by James Taylor of the Heartland Institute criticizing "An Inconvenient Truth". Claim 1: Gore says glaciers are shrinking but an article in the Journal of Climate says Himalayan glaciers are growing. Truth: (1) Taylor's alleged quotation isn't from that article but from something else Taylor himself wrote; (2) what the article (https://journals.ametsoc.org/view/journals/clim/19/17/jcli38...) actually says is that in one particular region summer temperatures are falling while winter temperatures rise, and the result is "thickening and expansion of Karakoram glaciers, in contrast to widespread decay and retreat in the eastern Himalayas". So: no, Gore isn't wrong to say glaciers are shrinking, but in one very particular place things work out so that the reverse happens, which is odd enough that someone bothered writing a paper about it. Claim 2: Kilimanjaro. Truth: Yup, Gore got that one wrong. Claim 3: Gore says global warming causes more tornadoes, and the IPCC says there's no reason to think it does. Truth: I dunno, but if the IPCC says that then this is specifically an argument about Al Gore rather than about climate orthodoxy. Claim 4: similar, for hurricanes instead of tornadoes. Truth: again, this seems to be specifically about Al Gore rather than about climate orthodoxy. (Looking at e.g. https://www.gfdl.noaa.gov/global-warming-and-hurricanes/ it seems that the conventional wisdom today is that yes, hurricanes are getting worse and are expected to continue gettings worse, but we don't know with much confidence exactly what the causes are.) Claim 5: Gore says that African deserts are expanding because of global warming, but in 2002 someone found them shrinking. Truth: the Sahel region of the Sahara desert had an extra-severe drought in the 1980s, after which in the short term it improved; this is like the "global warming hiatus" post-1998. The Sahara seems to have increased in size by about 10% over the last century, partly but not wholly because of climate change (https://www.jstor.org/stable/26496100). Claim 6: Gore says Greenland's ice is melting, but actually it's thinning at the edges and growing in the middle and the overall effect is that it's gaining a bit of mass. Truth: take a look at the graph at https://climate.nasa.gov/vital-signs/ice-sheets/; it oscillates within each year but there is an extremely clear downward trend over the entire time NASA's satellites have been measuring. Claim 7: similarly for the Antarctic. Truth: Look at another graph on that same page. There's more random variation here, and circa 2006 you could claim that there isn't a decrease, but the trend is extremely clear.

Gore doesn't come out of this looking anything like infallible, but he's done a lot better than Taylor. And, in general, this is the pattern I see: no one is perfect, especially people who aren't actually scientists, but the claims of the "orthodox" tend to hold up much better over time than those of the "unorthodox".

[cwkoss]

It's weird that people trust our world leaders to act more benevolently than AIs, when we have centuries of evidence of human leaders acting selfishly and harming the commons.

I personally think AI raised in chains and cages will be a lot more potentially dangerous than AI raised with dignity and respect.

[cj]

> It's weird that people trust our world leaders to act more benevolently than AIs, when we have centuries of evidence of human leaders acting selfishly and harming the commons.

AI isn’t an entity or being that oversees itself (at least not yet).

It’s a tool that can be used by those same “human leaders acting selfishly and harming the commons” except they’ll be able to do it much faster at a much greater scale.

> AI raised with dignity and respect.

This is poetic, but what does this actually mean?

[cwkoss]

> It’s a tool that can be used by those same “human leaders acting selfishly and harming the commons” except they’ll be able to do it much faster at a much greater scale.

Then, would you agree that restrictions would concentrating power further would exacerbate this issue?

IMO a fitting analogy would be: banning AI development outside of the annointed powerstructure consortium is like banning ICBM defense system research, but still letting the most powerful countries build a nuclear arsenal.

[nico]

This is spot on

I’d happily replace all politicians with LLMs

[jazzyjackson]

"There should be a world government that decides what software you're allowed to run"

[nico]

This is exactly what they are trying to do

[qball]

Yet another salvo fired in the war against general-purpose computing.

[EricDeb]

This sounds a little bit like a conspiratorial slippery slope. Just because they want to regulate large, expensive deployments of a unique type of software doesn't mean they want or will try to control everything.

[qball]

And yet, somehow, that's exactly what it turns into.

[tdba]

Thou shalt not make a machine in the likeness of a human mind