[Ygg2]

> Weird demo, suboptimal science, but nevertheless you've now got homeless people stuffed with anti-aging drugs.

Anti-aging drug. Not anti death drug. We don't keep more lab rats than we need. Not to mention lab rats aren't known for their quality of life. You aren't going to wait thousand years. You'll find a way to induce aging. Then run a battery of tests.

> It might be automated, but then there's no longer a meaningful distinction between rich and poor. A genuinely fully automated economy, all it takes is one person with a von Neumann replicator to decide everyone should have one

Yeah, no. First that is not necessary for full automation. Second. It's a replicator, not a magic entropy defying system. Energy for it has to come from somewhere and they aren't free.

[ben_w]

> Anti-aging drug. Not anti death drug. We don't keep more lab rats than we need. Not to mention lab rats aren't known for their quality of life. You aren't going to wait thousand years. You'll find a way to induce aging. Then run a battery of tests.

We've already got literal lab-rats, if that's what someone is planning to do. Human trials are pretty pointless if you don't do them realistically. (Not that this means nobody will do them, the Tuskegee study happened, but it was also low-value in addition to being unethical).

> Yeah, no. First that is not necessary for full automation.

It's a sub-set of what's necessary for full automation, as full automation requires anything that a human can do, and we can already do "build robot".

If machines cannot make robots, people will be paid to make robots, and then it won't be fully automated.

> Second. It's a replicator, not a magic entropy defying system. Energy for it has to come from somewhere and they aren't free.

Entropy doesn't need to be defied, magic is un-called-for. We are an existence proof of this.

Giant fusion reactor in the sky that will, if left to its own devices, probably give us gradually increasing power for about five times longer than our atmosphere will last. And it's only "probably" because there's a reasonable chance Earth gets ejected from the solar system over that time scale.

And before you say it PV is also a thing that we can do and thus a thing that must be fully automatable in any economy deserving of the description "fully automated".

But it doesn't need to last that long; if such a thing takes a year to make a copy of itself, then even limited to the surface of the Earth it would be able to make the last doubling, 4 billion units, if the construction had an energy budget of 247.7 GWh: https://www.wolframalpha.com/input?i=%28%286000km%29%5E2*pi*...

28.276 megawatts on average for a year is considerably more than we use to reach adulthood, even in countries with high per-capita usage. Biologically speaking, it's about 15700 times the energy consumption we need to reach adulthood (and the disparity is even more severe for, say, a dog which reproduces significantly younger than a human), and we get that energy and those materials by eating plants or animals that ate plants, which is also a clearly sufficient source of both materials and energy that this planet can provide without violating entropy or being magic.

[Ygg2]

> We've already got literal lab-rats, if that's what someone is planning to do. Human trials are pretty pointless if you don't do them realistically.

Yeah, and there is a gulf between works in mice and works in humans, as anyone reading science journals will tell you. Now, a human model. That's much closer to the real deal.

> It's a sub-set of what's necessary for full automation.

Not really. You are going for a holistic approach when a piecemeal bootstrap is much more likely.

It's a very theoretical solution to a problem that can be solved in a much messier but available way. E.g. Warp drive vs Nuclear power generation ships.

> Giant fusion reactor in the sky that will, if left to its own devices, probably give us gradually increasing power for about five times longer than our atmosphere will last.

You mean the sun? Sure, but that's an extremely unstable source of power that will have us relocate Earth(lings) first, if we want to continue to "use it".

> Entropy doesn't need to be defied, magic is un-called-for. We are an existence proof of this.

Magic is an apt comparison because it's an arcane, theoretical construct that has little to do with reality. Looking at the Wikipedia entry, it's an overkill for trivial purposes, by the time you construct a few, let alone, give everyone a copy, you'd probably exhaust Earth and nearby resources.

[ben_w]

> Now, a human model. That's much closer to the real deal.

Only if you don't shoot yourself in the foot in the process.

> Not really. You are going for a holistic approach when a piecemeal bootstrap is much more likely.

Yes really, and tautologically regardless of if it's piecemeal or sudden.

> Sure, but that's an extremely unstable source of power that will have us relocate Earth(lings) first, if we want to continue to "use it".

The sun is more stable than Earth's orbit and we're using it already. And self-replicating mechanisms ("life") have been running on it for billions of years before we came along.

> Magic is an apt comparison because it's an arcane, theoretical construct that has little to do with reality. Looking at the Wikipedia entry, it's an overkill for trivial purposes, by the time you construct a few, let alone, give everyone a copy, you'd probably exhaust Earth and nearby resources.

I'm looking at one right now: myself. Specifically, my fingers as I type this, because all life meets the criteria of a VN machine.

[Ygg2]

> Yes really, and tautologically regardless of if it's piecemeal or sudden.

You don't have to do it all from scratch. First variation can be built by humans, then the rest can be maintained by machines long term. It's like bootstrapping a compiler versus having compiler write/build itself and hardware.

> I'm looking at one right now: myself. Specifically, my fingers as I type this, because all life meets the criteria of a VN machine.

You aren't a Von Neumann replicator. Or at least not a useful one. No human can construct hammer, chairs and PCs given sequence of DNA. Unless you have to learn it yourself, which defeats the purpose, or you have to raise a new one from scratch for 18 years.

Previous statement indicated that they are necessary for full automation, implying they are useful when it comes to generating artifacts useful for humans.

> The sun is more stable than Earth's orbit and we're using it already. And self-replicating mechanisms ("life") have been running on it for billions of years before we came along.

Sun is stable? Could have fooled me. How are the solar flares?