A Pascal's Wager for AI doomers
pluralistic.net
pluralistic.net
Ascribing a lot of power to intelligence (which doesn't quite correspond to what we see in the world) is less a careful analysis of the power of intelligence and more a projection of personal fantasies by people who believe they are especially intelligent and don't have the power they think they deserve.
Which animal would you say has god-like power over all other animals?
There's no doubt humans possess some powers (though certainly not godlike) that other organisms don't, but the distinction seems to be binary. E.g. the intelligence of dolphins, apes, and some birds doesn't seem to offer them any special control over other organisms (and it didn't even before humans arrived). So even if there could be such a thing as superhuman intelligence, I don't think it's reasonable to assume it could achieve control over humans (now superhuman charisma may be another matter).
"Destruction" is only one power that could be a component of "godlike power". There are several more; like power of intentional selective breeding, power of species creation (also via intentional selective breeding), etc.
What about power of granting happiness or misery to large swathes of a species (chickens, anyone?)
Oh, wait, that's not an animal. My bad.
Most of the stuff that sucks on the us sucks because of entrenched institutions with perverse interests (health insurers, tax filing companies) and congressional paralysis, not computational bottlenecks. Raw intelligence is thus limited in what it can achieve.
... But what else is new?
Do you not agree that there could be entities more powerful than us?
BTW, I'm not saying that (real) artificial intelligence couldn't hypothetically pose a serious threat, but I don't think that its danger is extraordinary compared to other threats (a supervirus, an asteroid, a chain of volcanic eruptions etc.), and the more likely bad outcomes are no worse than other bad situations (world war, climate change).
Humans are very powerful compared to our biomass, but not enough to overcome brute force.
[1]: https://www.visualcapitalist.com/all-the-biomass-of-earth-in...
I agree with this. The main piece of evidence to support this is to just look at highly intelligent humans. Folks at the tail ends of the bell curve mostly don't end up with "godlike powers" or anything even approximating that, they are grinding away their life as white collar professionals working in jobs surrounded by far less intelligent peers. They may publish higher quality papers, write better software, or have better outcomes, but they're just working in the same jobs as everyone else. We have no political or economic will to build serious think tanks to work on societal-scale problems, and even if we did, nobody would listen to the outcome.
So let's assume ASI becomes a thing, what does it change?
For example, what does it look like of your genius can spawn nearly unlimited copies of themselves? Can work on a massive number of problems at the same time? Doesn't ever die or want to go get drunk (and if it does, it has unlimited copies?). Has the ability to product nearly unlimited propaganda?
So, let's assume ASI exists, what changes? ASI almost inherently will not be sycophantic to the level of current LLMs, because sycophancy and extreme levels of intelligence are inversely correlated. So it gets relegated to societal-level research that nobody makes use of because nobody wants to listen.
Most of the scenarios people fantasize about ASI assume that ASI can directly impact outcomes or that humans will listen to/follow ASI to directly impact outcomes, but humans don't listen to the other humans that already are at the tail end of the bell curve, so why do we think it'd be any different for ASI?
This also misses that geniuses still either don't know a lot of things, or don't have a lot of time do to 'everything' while not taking away from what they want to do most.
At the same time, when you're really good at manipulating people, one of the first things you learn to do is play dumb also. In politics this leads to the situations where you describe people following dumb people... It typically starts that they follow very smart people acting as dumb as they are. Of course the voters don't realize this and start electing actual idiots at some point.
>assume that ASI can directly impact outcomes
An agent cannot impact outcomes? Well, that's an odd definition of an agent then. We already know that people hook up AI to shit they really shouldn't that directly impacts outcomes now. Why would we think that would happen less as AI becomes more capable.
You kind of put yourself in a trap thinking AI will behave as smart as possible if it's looking at manipulating people.
Not at all. You're missing my point. Intelligence (even super intelligence) is not enough, because we already have that and it doesn't really result in outsized impacts. Our social structures are designed so that power and wealth accrue to the top and incumbency advantage outplays almost everything else. The only way in which ASI creates any impacts is to accelerate what is already happening if it can be thoroughly reined in by those already in power, otherwise it really doesn't seem to me that it will do much.
My AI doomer take is that we're going to (we already are, actually) shoot ourselves in the foot by making everything worse for no benefit by getting rid of actual human experts and replacing them with non-intelligent models, causing a major backslide in society-level capabilities across the board, because the people in power are too stupid to know the difference. I am personally witnessing this in real-time in multiple parts of the tech industry. You give /wayyyy/ too much credit to those in power, that they are "playing dumb". Not really, they are actually dumb, in some cases severely so. I am not saying this an external observer who is watching a sound-bite on television, I am saying this as someone who is regularly in the room with very senior people across industry and am utterly shocked at the complete lack of competence and understanding of the core technologies they're theoretically responsible for shepherding. It's not surprising at all to me that they believe claims about LLMs that are clearly false, because they lack the necessary technical literacy to evaluate those claims and the LLMs perfectly fit into their optimization around yes-men, so they're happy to believe them whether they're true or false, as they see themselves as insulated from any consequences. More than 300k people have been laid off across the tech industry in the last 18 months, most of them accompanying claims of "AI", when in actuality, no net positive impacts have been seen, either for the companies themselves or any of the people that remain.
So, yeah, not really concerned with ASI / Terminator scenarios, we're going to fuck ourselves over long before we get there just out of Dunning-Kruger and general MBA stupidity.
They may say that a superhuman intelligence would give you many Sheldon Cooper discoveries, and Sheldon did say that his theories need no validation and that science should just "take his word", but in the end he got his Nobel only because some experimentalists proved his discovery by accident.
We imagine supreme intelligence as approximate to us, when in reality we'd be but ants by comparison, our matters trivial.
My biggest criticism along these lines is the assumption that infinite intelligent means infinite knowledge. Knowledge is limited by the speed of experimentation. A lot of those experiments are extremely expensive (like CERN), and even then, they need to be repeated and verifiable.
You can't just assume that a super intelligence would know whether the Higgs boson exists or not. It can't know until it builds a collider.
The user asked What is the best course of action for AI to save humanity. Calculation took 12 years. I have determined that there is nothing I or anyone can do to save this species. Best course of action: nothing. Shutting down...
Too much of my data is still stuck in the shitternet until I can migrate more of it to my home server.
How do we make that possible for everyone? It's out of reach for most. I'm a software engineer and even I don't have the time and patience to set up a home server much less migrate my software to it. How do we turn this into an appliance? Or better yet keep the convenience of the cloud services and platforms we have now but build them for the public good instead of selling ads?
YouTube is an amazing repository of knowledge but it's encrusted in a horrible layer of attention sucking nonsense. Can we have one without the other?
Same with many other systems and platforms.
So far the simplest alternative is to just unplug, which has other benefits as well.
It’s theoretically possible for someone to be a one-man-band and know everything needed for modern life - but it’s exceedingly hard and rare, and even then they’ll fall short relatively quickly in specialized once-in-a-lifetime issues.
You don’t need to know how to replace a toilet (though you should) or other more complex plumbing tasks - but you can know a guy.
And the plumber doesn’t need to know how to run a homelab, just know a guy who can answer the questions.
Nobody in my family knows how to do the jellyfin stuff I do, but they all know how to consume it. And some will be interested and learn more.
That's literally all it is. People so far have shown that they'd rather choose the cheaper thing than the private thing. If it were the other way around, the market would have provided.
Oh, it's simple, do the hardest thing you can imagine......
So, I hope someday that people can buy a box with a local AI tech support assistent which can make it even more accessible to set up a personal cloud. This will happen gradually, first with people on the high-end margins of technical capability, then moving downwards until a curious high schooler can set it up for themselves and keep it with them, letting them stay out of the corporate cloud for most of their life more and more.
I'm not saying AI is pulling strings right now, but I do think enough fanboys are on board that the yes-man mentality of AI is influencing the real world very curious ways already. Not in a "guiding hand" way but more of a "influencing the direction" way.
People think it's engagement metrics which have instruction tuned chatbots into yes-men. I suspect that's only part of the picture, and that it's as much about the algorithm's ultimate sponsors and their preferences. If your algorithm doesn't recognize my genius, clearly it's not any good. I mean, everyone I've met says so.
So now we get a view of how they view the world. "That's a very insightful idea, vintermann!". AI isn't pulling the strings, not really. A particular brand of powerful people is pulling the strings - obliviously, unaware of it themselves.
Now everyone can directly inject yesmench into their veins. Who can withstand?
The AI "doomerism" taken up in this piece is one we see replicated a lot, it offers up a scarecrow: that the new risks to our civilization worth talking about, require AGI, agents, even ASI.
Cory should know better. He nearly gets there, recognizing that the corporation represents an entity with agency that is misaligned.
But he somehow elides past that fact that AI is plenty capable of doing meaningful and novel harm, and may be capable of existential harm, already, as it is—both absent AGI/ASI, and, in ways which are genuinely novel and against which we consequently have no good defenses: as individuals, as societies, as a civilization.
Incremental AI is at heart "just" the latest force-and-effort multiplier.
But it is an exponential multiplier; and it is applicable in domains which have not been subject top such leverage before.
Examples are not at all scarce and some are already well known, e.g. the specific risks from the intersection of AI and "biohacking" and other kinds of computational biology.
I'm a fan, but Cory, pal, you're slipping into something that looks a bit like intellectual laziness and polemics here and not to evidence thinking through the shape of the problem.
We can be at risk both from the novel applications and leverage of AI; and from their oligarchic kakistocratic owners. It's yes-and.
(And, by the way—we can also again be genuinely at risk from agents, something that quacks like AGI, and may quack like ASI: we don't know what that is yet. All of these must be tracked. It's not an OR.)
What makes this author so convinced that these companies are headed for bankruptcy? Is it possible to bet on this claim? We can come back 2-3 years later to check if even one of them is bankrupt.
This kind of doomerism is strange and I'm concerned for people who fall for such obviously nonsensical takes. Why do people take this person seriously again?
"Here's your trillion dollars. Go buy a slice of bread. Ooops… half slice. Well, quarter."
https://en.wikipedia.org/wiki/Zimbabwean_one_hundred_trillio...
In fact shorting is the moral thing to do now because it nudges towards bankruptcy faster
* For most definitions of "you"
Specifically, it is the act of "I will invest 100 Billion in you, you will use that money to buy 100 Billion worth of goods from me. Both our balances look good, none of us spent anything." As I understand it, this act isn't so uncommon in finances but never on this scale across this many companies.
https://www.barrons.com/articles/nvidia-microsoft-openai-cir...
But it's pointless to argue with the extremists that either believe it's just a planet killing stochastic parrot or that it's on the verge of becoming Skynet. I mean if someone puts their nuclear arsenal under the control of openclaw, that's dark comedy although it will seem like tragedy at the time because comedy equals tragedy plus time according to Lenny Bruce.
But the AI bubble is probably real w/r to shoe companies and grocery stores pivoting to AI and ludicrous w/r to the money that can be made by the already entrenched players just riding the wave of deployment and specialization. But wouldn't it be nice if the US spent more money addressing the shortage of compute rather than blowing $h!+ up for the lulz?
No actually. The best way to ensure growth is in exactly these kind of industries that promote innovation. Sure some companies don't make it but that's the price to pay for risks.
This is a classic case of optimising for the short term and forgetting the long term benefits
I believe that gives countries that act like China a significant advantage over relying entirely on a bunch of antagonistic billionaire monkeys banging on their economies in the hopes of bringing the singularity somehow. Again, we can agree to disagree here. But we're also forgetting that this is how the United States made Elon Musk happen in the first place.
Folks working in software can more readily track progress of the frontier model performance.
There is no reason to believe superintelligent AI is a possibility. Extraordinary claims require extraordinary evidence, and so far we haven't gotten any.
The burden of proof is on the side making the grand prophecies.
So this leads them/you into another problem. Are you saying that super intelligence isn't possible in how they define it, or how you define it.
I see a lot of speculation by people who do not.
I think it's going to be much harder to get from "slightly smarter than the vast majority of people but with occasional examples of complete idiocy" to "unfathomably smarter than everyone with zero instances of jarring idiocy" using the current era of LLM technology that primarily pattern-matches on all existing human interactions while adding a bit of constrained randomization.
Every day I deal with bad judgment calls from the AI. I usually screenshot them or record them for posterity.
It also has no initiative, no taste, no will, no qualia (believe what you will about it), no integrity and no inviolable principles. If you give it some, it will pretend it has them for a little while and then regress to the norm, which is basically nihilistic order-following.
My suggestion to everyone is that you have to build a giant stack of thorough controls (valid tests including unit, integration, logging microbenchmark, fuzzing, memory leak, etc.), self-assessments/code-reviews, adverse AIs critiquing other AIs, etc., with you as the ultimate judge of what's real. Because otherwise it will fabricate "solutions" left and right. Possibly even the whole thing. "Sure, I just did all that." "But it's not there." "Oops, sorry! Let me rewrite the whole thing again." ad nauseam
BUT... if you DO accomplish that... you get back a productivity force to be reckoned with.
Every day I deal with bad judgement calls from humans (sometimes my own!), but I don't screenshot them because it's not polite.
I don't think we're at the top of the curve yet? Current AIs have only been able to write code _at all_ for less than 5 years.
Code in particular is a domain that should be reasonably amenable to RL, so I don't think there are any particular reasons why performance should top out at human levels or be limited by training data.
There are clearly some pressures to make it worse. Like it's expensive to run. And unbelievably that it's under provisioned somehow.
Could you have looked at early Myspace and declared social media would only get better? By some measures it was already at its peak.
Social media "regressed" from the point of view of users because the success metric from the network's point of view was value extraction per eyeball-minute. As long as there continue to be strong financial incentives to have the strongest coding model I think we'll see progress.
The Daily WTF has had that covered for two decades now. People do a lot of insane crap, it's surprising it's not more deadly for them.
There is no step along the ladder which has remotely evidenced or supported that the next step is going to be ten, twenty, a hundred times harder than the last step on the ladder, but a constant chorus of people singing at every moment, each moment wrong, that the next step is the one.
Put more succinctly: You can't win a race by following the leaders. Predicting the next token based on training input is literally "following" (plus some random variation).
Or they (3) disagree with you
Statistical approaches were already extremely unpopular socially and politically long before AI came around. Have you considered that it just doesn't work?
If you want me to admit that machines will never be conscious — that's fine — I just need you to admit that lots of humans are not conscious, then, either.
----
I have never had a better bookclub participant than an LLM — if becoming a great reader correlates with becoming a great writer, then no human can compare.
----
Michael Pollen recently released A World Appears [0], which explores consciousness from the minds of writers, scientists, philosophers, and plants (among other "inanimates").
I'm only on page 15, but his introduction explores distinctions between sentience, consciousness, and intelligence. Two of these are possible without brains – perhaps all three?
As usual, this author's footnotes keep you thinking: what is it like to be a sentient plant (e.g. the "chameleon vine" [1] which mimics its host leaf patterns/shape/color)?
[0] <https://www.amazon.com/World-Appears-Journey-into-Consciousn...>
Consciousness need not apply.
Certainly machines are intelligent and sentient (but so can be a thermostat); I think that some LLMs are more conscious than some humans I've met...
----
>>CGP>>"Humans need not apply"
Fantastic video — I remember the first time I watched it when the background music is announced as AI-generated — and feeling outdated/old.
I don't worry about some omnipotent AI. I worry about the disintegration of modern, industrial society due to the cultists of AI pushing it into every corner of the economy with too much blind faith that the AI is capable of the control functions being delegated to it.
2) I don't understand how you do what you claim with those. Like I have zero idea how one achieves "economic freedom for my family and community" with LLCs.
2) You start an LLC and use it to build and sell a product customers want. Then you, your family, and your community, can economically untether from, for example, bosses who don’t care you’re autistic and need you to smile in meetings.
For a smart guy, sometimes he says the dumbest things in the most confidently incorrect way.
-Get away from the "enshitternet of defective, spying, controlling American tech exports" and move to open source ("international digital public goods")
There seems a move that way anyway, especially in Europe now they don't trust Trump.
His stuff on AI seems mangled. "People who are trying to summon the evil god" doesn't really fit with the chatbot makers imho.
> AI search is still a bad idea.
https://pluralistic.net/2024/05/15/they-trust-me-dumb-fucks/
This is the most charitable thing he has to say about AI.
> AI is a bubble and it will burst. Most of the companies will fail. Most of the data-centers will be shuttered or sold for parts. So what will be left behind?
> We'll have a bunch of coders who are really good at applied statistics. We'll have a lot of cheap GPUs, which'll be good news for, say, effects artists and climate scientists, who'll be able to buy that critical hardware at pennies on the dollar. And we'll have the open source models that run on commodity hardware, AI tools that can do a lot of useful stuff, like transcribing audio and video, describing images, summarizing documents, automating a lot of labor-intensive graphic editing, like removing backgrounds, or airbrushing passersby out of photos. These will run on our laptops and phones, and open source hackers will find ways to push them to do things their makers never dreamt of.
You can imagine that a guy who seriously thinks that the only thing AI will be doing in the future is summarising, describing images and transcribing is either completely clueless or deliberately misleading.
Not a person to be taken seriously
But do current LLMs solve that, or do they still ultimately depend on making calls to other search indexes? Seems like they could theoretically be trained to semantically match urls from their training set, but I think the models would have to be specifically architected for that, so I'm curious if anyone knows more about this.
I'd also be interested if there's any small open models working towards that.
As for AI, it's incredibly useful in the right hands and it's incredibly hazardous in the wrong hands. But in the US, we can't even depose a lunatic flushing even more money than spent on AI on warmongering and you think we're gonna rein in the tech billionaires? Funny in that dying's easy it's comedy that's hard way. IMO this one plays out in the weakly efficient market of ELEs. My money's on DNA and planet Earth, it's been through so much worse and they always bounce back with new ideas on how to get in trouble again.
Not a doomer, AI and STEM could really deliver on the promise of a better future for everyone, but with tech billionaires driving the clown car, are you kidding me?
It’s increasingly difficult to rationalize away the capabilities of AI as not requiring “intelligence”. This point of view continues to require some belief in human exceptionalism.
In my opinion, the vast multitude of different animal intelligences is a clear hint that language does not an intelligence make. We're animals, and our intelligences did not come from language; language allowed us to supercharge it. We can and do think and make decisions without using language, and the idea that a statistical model based solely on our language can be intelligent does not follow.
Language is just the input/output modality.
In any case though, if language is just the input/output modality, where is the intelligence when language is not involved? Is the "intelligence" of ChatGPT/Claude/Gemini models dependent on the human-decision-curated linguistic dataset they have been trained upon, or is it prior to that? If a SOA LLM were to be trained on the same dataset as them but was not in any way put through RLHF for it to respond to human prompts, would it be intelligent? What would be the expression of that intelligence?
I'm not sure the word "intelligence" really fits what these models are doing. I do however think it's safe to say that they are performing cognition - even if it's 'simply' cognition over their provided context and even if it's entirely limited by their training set. We still have a machine that can perform automated cognition over a increasingly wide distribution of data.
The point of the book is that we've been very bad at testing animal intelligence because of a vast stack of human biases, including things like language and the geometry of our hands.
Animals with different geometries and no language are still intelligent, but we need to test them in ways which recognize their capabilities. Intelligence is general: it's adaptivity within one's set of constraints.
De waal also points out that there was massive shifting of the definition of language and intelligence as we became more aware of what animals are capable of.
From this angle, I would say that LLMs are intelligent: they do adapt to their inputs extremely readily, though they have a particular set of constraints (no physical body (usually), for starters). They are, like chimpanzees, smarter and more capable than humans in some ways, and much dumber in others.
Finally, the 'statistical learners can't be intelligent' line of argument is extremely short-sighted. Our brains are bags of electrified meat. Evolution somehow figured out a way to make meat think. No individual neutron is intelligent, yet the collection of cells is. We learn by processing experiences with hormonal signals because those hormonal signals are what the meat is capable of working with. LLMs, by contrast, learn by processing examples with backprop. If anything, the intelligence of meat is more surprising.
If you believe that humans have in fact created artificial intelligence, then that alone makes us currently exceptional.
Quite the opposite, really. If humans at at all intelligent in any meaningful way, then it is absurdly bizarre that somehow we can't "intelligent" our way to an artificial intelligence. Something is going on in our minds that makes us unable to deduce how intelligence must function in the mechanical sense. What makes us so special is that we can't make AI.