You can set its intelligence arbitrarily high, the fact remains that it may still bump up against hard physical constraints already familiar to us.

It's naïve to assume that everything is possible.

Vidya is vidya gaem.

I joke, but I do see this as the long suicide of our species. There's a skinner box in each of our futures.

 

11am - Wake up
12pm - AI powered VR porn
1pm - AI powered VR porn
2pm - AI powered VR porn
3pm - AI powered VR porn
4pm - DnD
5pm - DnD
6pm - DnD
7pm - DnD
8pm - Browse Reddit
9pm - Vidya
10pm - Vidya
11pm - Vidya
12am - Vidya
1am - Write poetry under the stars.
2am - AI powered VR porn
3am - AI powered VR porn
4am - Sleep

My response to him would be, "Yeah. Those are legitimate concerns, but the subservient, narrow AI may kill us before the rogue AGI does."

Why though? It makes it harder to get work done. That's ostensibly what it's there to do, help us do things.

No, I think you're right about how the economic model works today. If population flat lines, consumption does too, and production follows, both because there's no growth in demand and because there are no new bodies to work.

I think something like half of annual GDP growth comes population growth, and the rest from gains in productivity. That's not a feature of capitalism, because even the Soviet Union's output benefited from more bodies.

That said, yes. The flattening of population growth hurts economic growth, but that might be mitigated against by removing labor as a bottleneck. Human consumption could stagnate, but there's no reason productivity has to, so the economy could continue to grow.

I have an opinion. I expressed it. My opinion is that none of this is as obvious as you made it out to be.

>I think by now we all agree that a highly automated society can become a post-labor society and therefore should resort to a wealth distribution instrument such as universal basic income.

Uh. Don't put me down for that. I don't have expert knowledge in macroeconomics, so I don't have a clue what an automated economy looks like. Not even a guess. North and South could become East and West for all I know.

Anyone here who purports greater certainty than that is lying to you or themselves.

A machine can cycle oxygen through an air sac. Is there any point to breathing?

You'll do art for yourself, because you enjoy it. There won't be a commercial incentive to do it.

I've enjoyed writing poetry for 20 years now. It's a dead art. My sonnets were never going to pay the rent. I understand what it's like to practice at something that has no monetary value, and visual artists will too soon.

>I don't have to 'justify' anything, that's not what I'm trying to do. I'm raising questions, not peddling answers. I'm trying to be a philosopher about AI, not a preist.

I've seen you put forward firm, prescriptive opinions about how people should think and about what's signal and noise. It's clear that you have a lot of opinions you'd like people to share. The title of your OP and almost every sentence since then has been a statement about what you believe to be true. I have not seen you ask any questions, however. So how is this different from what a priest does?

You say you're not trying to persuade anyone, then follow that with a two paragraph tangent arguing that AI needs to be handled under the paradigm of psychology and not economics.

You told me you weren't doing a thing while doing that very thing. This is gaslighting.

Those examples are tedious and unrealistic, but I think by design. They're cartoons meant to illustrate a point.

If you want a more realistic example of the alignment problem, I'd point to modern corporations. They are powerful, artificial, intelligent systems whose value function takes a single input, short term profit, and discounts ALL of the other things we'd like intelligent systems to care about.

When I think about the alignment problem, I don't think about paperclips per se. I think about Facebook and Google creating toxic information bubbles online, leveraging outrage and misinformation to drive engagement. I think of WotC dismantling the legal framework that permits a vibrant ecosystem of competitors publishing DnD content. I think of Big Oil fighting to keep consumption high in spite of what it's doing to the climate. I think of banks relaxing lending standards so they could profit off the secondary mortgage market, crashing the economy.

That's what the alignment problem looks like to me, and I think we should ask what we can do to avoid analogous mismatches being baked into the AI-driven economy of tomorrow, or we could wind up with things misaligned in the same way and degree as corporations but orders of magnitude more powerful.

>... 'absolute truth' is a load of nonsense ...

Is that absolutely true, "bro"?

If we can put aside our mutual lack of respect for one another, I'm genuinely, intellectually curious. How do you expect people to be moved to your way of thinking without "cartesian style explanations"?

Do you envision that people will just feel the weakness of "cartesian-thinking"? If that's the case, shouldn't you at least be making more appeals to emotion? You categorically refuse to justify your beliefs, so what is the incentive for someone to entertain them?

Again, sincere question.

Invested readers will note that he didn't provide any concrete explanations here either.

I see you're having some trouble with verbosity. Let me help you.

>... arrogance, conceit and self-importance?

These words mean the same thing.

That was a very butthurt response.

>It's not my fault you don't understand what I mean; 'storyteller' is not a complex word.

I think it actually is, because there's no context given. How does a storytelling AI differ from what's being built now? What is a story in this context? How do you instantiate storytelling in code? It has nothing to do with reading comprehension; there are a lot of ambiguities you've left open in favor of rambling about Descartes.

>So it would be better, I think, to build 'storyteller' minds that can build up their senses of ethics independently, from their own knowledge and insights, without needing to rely on some kind of human 'Ten Commandments' style of mumbo-jumbo.

Putting aside the fact that I don't think anyone knows what you mean by a "storyteller mind," this is not a solution to the alignment problem. This is a rejection of it. The entire problem is that we may not like the stories that AIs come up with.

Don't worry. So is OP.

So. You don't say anything about AI. You don't offer any ideas about what a useful approach to alignment looks like. You don't even justify your animosity toward Descartes or illustrate his relationship to AI safety.

I can't tell you that you're wrong, because you've literally said nothing.

It's not obvious that there can be a finite number of things to learn.

>I'm pretty sure the only real 'purpose' of humans will be to just live out our lives and then die

That has been (and frankly continues to be) the human experience for the vast majority of people.

I misinterpreted you to mean that it felt soulless. But whether the consumer sees the soul in a piece has more to do with them than the artist.

In any case, yes. Artists are bound to be negatively influenced, and I'm surprised I haven't seen anyone with the receipts to show how their commissions fell off a cliff. I'm sure it's happening for a lot of people.

>There’s nothing AI can currently produce like the organic expression of actual human artists

If that were the case, you'd think it would be easier to tell what is and isn't AI generated art, but people are routinely failing at this in both directions.

>"there's no point for you to create any more".

If you were confronted by a robot that cycled oxygen through an air sac, would you give up breathing?

I've enjoyed writing poetry for decades, since before AI could do it, but it was a dead art since before I was born. The market for it is small and the barrier to entry is low, so I was never going to make money from it, and practically no one would ever see it, but I still do it for me.

Visual art is going to the same place, just much more suddenly.

1. Sure, but it doesn't need to be environmentally friendly to be the pinnacle of engineering. Nothing's less environmental than a Dyson sphere, and that'd be pretty impressive.

  2. I don't want to have to walk in the rain or heat to a bus stop, hauling my stuff, wait for the bus, sit on a circuitous route, stopping at places I don't want to stop at, get off the bus, and walk through the rain or heat, again hauling my stuff, just to meet up at a friend's house for some board games. THEN DO IT AGAIN ON THE WAY HOME! That's not freedom. No. I did that until I was 30. Never again. Absolutely not.

3. Trains have pre-determined destinations. They don't go where you want to go. You know that. They cannot solve the problems that a driverless car can.

4. Yes, and?

5. So put the vehicles on a subscription service and run them like Uber.

6. We're not going to Replan, tear down, and rebuild Houston and Chicago as part of some kind of top-down initiative. If you want more compressed cities, you can either take this to Mars or wait for additional growth to fill in the empty space.