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

The Coming Technological Singularity (1993)

mindstalk.net

Post reply on HN
Print view View original

151–160 of 181 posts

Re: The Coming Technological Singularity (1993)

#151

Post by p1esk »

Animats wrote: post #46
p1esk wrote: post #25

Earlier quoted context omitted.

fairly clear what they can and can't do It’s not at all clear what the next gen models will do (e.g. gpt5). Might be enough to trigger mass unemployment. Or not.

Bigger LLM models probably won't fix the underlying problems of hallucinations, lack of a confidence metric, and lack of a world model. They just do better at finding something relevant on already-solved problems.

“Just do better” is the key metric here. 99% of all human jobs are “finding something relevant with already solved problems”.

Re: The Coming Technological Singularity (1993)

#153

Post by nonameiguess »

jodrellblank wrote: post #31
WillAdams wrote: post #6

The thing which these discussions leave out are the physical aspects: - if a computer system were able to design a better computer system, how much would it cost to then manufacture said system? How much would it cost to build the fabrication facilities necessary to create this hypothetical better computer? - once this new computer is running, how much power does it require? What are the on-going costs to keep it run…

Skip a few generations and the machine will build itself. There’s no need for it to take lasers exploding tin to generate ultra Violet to etch patterns to make intelligence, humans don’t grow brains that way or spend billions on fabs and power plants to produce children. How it gets from here to there is a handwave, though.

There's no guarantee it's possible. Self-replicating molecules don't necessarily have to be hydrocarbons suspended in water inside of a semi-permeable lipid membrane, but that particular solution came about because of conditions that are needed:

- Some multipolar molecules that could attract or repel each other and initiate both hydrogen bonds and covalent bonds.

- Movement through some kind of medium not as chaotic as gas.

- A boundary between self and everything else.

Industrial processes to create machines don't lend themselves to this kind of bottom-up model as they now stand. It's not at all clear that etched silicon at all is the right medium of computation if we want it to be self-replicating. Coming up with something else is a hell of an ask and a hell of a handwave, though. It would also entail getting rid of the entire impetus of Moore's law and the reason futurolists like this ever thought there would be a singularity in the first place. Other than transistor density on a silicion die, I'm not sure any other technology has ever shown consistent long-term exponential growth at some reasonably fixed exponent.

Re: The Coming Technological Singularity (1993)

#154

Post by lazide »

dh77l wrote: post #43
lazide wrote: post #30

Earlier quoted context omitted.

Eh, I’m not talking about people’s opinions. I’m talking about evolutionary functions, and how much more likely it is to prefer something that has fun and just looks like it’s doing something, instead of actually doing something. Aka manipulation vs actual hard work. Do you have any concrete proposals, besides ‘it will get better’? Actual competency is hard. Faking it is usually way easier. It’s the same reason the ‘…

You could ask FDR and Churchill that after the fall of France and it wouldnt be too useful what they said cuz it took them almost 3 years before they openly said victory = end of the nazis and nothing else. So dont just sweep the fact that things take Time under the carpet. Its not healthy cause its like looking at tree shoots in the ground and saying but why does that not look like a tree yet. Finding gold in an une…

What are you saying exactly? The more things change, the more they stay the same?

Re: The Coming Technological Singularity (1993)

#155

Post by exe34 »

Retric wrote: post #85
trashtester wrote: post #59

Earlier quoted context omitted.

AlphaFold/AlphaProteo are direct examples of how AI can allow us to bypass experiments when doing science. More importantly, though, is that insufficiently powerful AI has been a limiting factor in robotics. That seems to be coming to an end now. And once we have humanoid robots powered by superhuman intelligence entering the workforce, the impact will be massive. Quite possibly mostly for the bad for most people.

You’re overstating what AlphaFold can do. Without validation the predictions aren’t dependable, but it can cut down on what’s worth investigating. There’s definitely a point where models get good enough you can trust them, but super intelligence doesn’t mean it can both come up with a model and be able to trust it without validation.

I don't know what this article (0) says specifically, but I've been to a talk where they showed a giant grid of dishes where a robot like the one in the picture was busy doing things to the samples. yes, they don't make the glass themselves and have to wait for that in the post, but really, given the tools, these devices could very easily chew through a lot more work than any human can do in a day/week/month/whatever without getting tired or distracted.

I don't expect the first ASI to conduct novel physics experiments, but I can easily believe that after optimizing the supply chain for everything else, it just happens to have stashed a bunch of magnets and whatnot to do these in its spare time.

(0) https://www.nature.com/articles/d41586-024-00093-w

Re: The Coming Technological Singularity (1993)

#156

Post by throw4950sh06 »

ben_w wrote: post #143
shermantanktop wrote: post #137

Earlier quoted context omitted.

There are incredible technological barriers to humanoid robots who have equivalent skills and stamina. Keeping old factories running seems a very weak reason to do that, when our industrial base regularly retools production methods and brings in new equipment when old machines wear out. If what you are saying is that many factories cannot run with humans running around fixing things, I agree. But that’s pretty differ…

Yes indeed. Even just 3.5 years ago, seemed like everyone was saying that humanoid robots were a dead end or an unnecessary part of Isaac Asimov's vision of the future or similar. I think much of the current interest is because Musk watched some scifi, ordered the Optimus project, and loads of others decided it would be a mistake to bet against him. I put them in the same category as 3D printers: they can do anything…

Sometimes you need to produce megatons of a thing, but sometimes you need to produce a million different things. I bet on the humanoid robot in that case.

Re: The Coming Technological Singularity (1993)

#157

Post by ben_w »

falcor84 wrote: post #149
ben_w wrote: post #125

Earlier quoted context omitted.

> The industrial revolution didn't really change anything about land. I didn't say otherwise. I said the industrial revolution changed what wealth meant. We don't pay for rents with the productive yield of vegetable gardens, and a lawn is no longer a symbol of conspicuous consumption due to signifying that the owner/tenant is so rich they don't need all their land to be productive. And indeed, while land is foundatio…

> enough food to not starve for 4.43 seconds I'm having real difficulty reading this unit of measurement. Let me see if I can get this right - a typical person can survive indefinitely on 1600 calories. Let's say that these are provided by rice (which isn't sufficient for a long-term diet, but is good enough for awhile). 1600 calories of rice is about 8 cups/24h and there are about 10000 grains in a cup, so is it tha…

Sounds about right, but I don't have unit conversions and I'd count anything less than lifetime sustainable as gradual starvation.

Re: The Coming Technological Singularity (1993)

#158

Post by LinuxAmbulance »

aithrowawaycomm wrote: post #34

> To date, there has been much controversy as to whether we can create human equivalence in a machine. But if the answer is "yes, we can", then there is little doubt that beings more intelligent can be constructed shortly thereafter. I find it bizarre how often these points are repeated. They were both obviously wrong in 1993, and obviously wrong now. 1) A nitpick I've had since grad school: the answer to "can we cre…

Far too much of anything singularity related depends on "And then magic happens".

As if intelligence and / or consciousness arises spontaneously once there's sufficient resources to support it. Which is ludicrous, on the face of it.

Let alone bootstrapping intelligence.

Re: The Coming Technological Singularity (1993)

#159

Post by Terr_ »

amanaplanacanal wrote: post #61
auggierose wrote: post #55

Earlier quoted context omitted.

I don't think it makes sense to include all of that in the calculation. A human doesn't need all of that, they just need 900 calories. You can just eat berries, no need to cook anything, for example.

A human cannot live on just berries. Even if they could, where are those berries gonna come from?

> Even if they could, where are those berries gonna come from?

By that logic, the "true cost" of boiling a cup of water for my tea somehow involves the Big Bang and rest of the formation of the observable universe up until now.

It's kinda-true, but not in a useful way.

Re: The Coming Technological Singularity (1993)

#160

Post by Retric »

trashtester wrote: post #131
Retric wrote: post #112

Earlier quoted context omitted.

> using actual experiments That’s the bottleneck the model was trying to avoid in the first place. The goal of science is to come up with models we don’t need to validate before use, and it’s inherently iterative. Nanbots are more sci-fi magic than real world possible. In the real world we are stuck with things closer to highly specialized cellular machinery than some do anything grey goo. Growing buildings from loca…

> That’s the bottleneck the model was trying to avoid in the first place. Some real world validation is always needed, but if the validations that are performed show high accuracy, the number of experiments will go down a lot. > Nanbots are more sci-fi magic than real world possible. Let's revisit this one in 10 years.

> revisit their one in 10 years

The underlying physics isn’t changing in 100 years. Individual components can be nanoscale within controlled environments, but you simply need more atoms to operate independently.

Post reply on HN