4 ms·
> I wonder what Gödel would think about modern AI. As we have found, learning is not a special ability of humans, and can, in fact, be simulated. That on its ow
by kevindamm 3y ago
> I wonder what Gödel would think about modern AI. As we have found, learning is not a special ability of humans, and can, in fact, be simulated. That on its own doesn't grant eternal life, though AI certainly has more chances of survival, not being bound by our fleshy meat sacks and all.
Instead bound by substrate and power grid that are both inefficient and minimally redundant. I'd wager more on the meat sacks.
- namaria 3y agoSystems that can reproduce themselves are way more likely to endure than systems that need to be held together from the outside. In 5000 years I think it's more likely that there will be people than that there will be machines.
- ben_w 3y agoWhile the only currently known Von Neumann replicator is organic life, my bet is that a robotic one will be made at the latest by the end of century unless some major global catastrophe happens first. Might turn out that organic life is pretty optimal already and we still dominate in numbers, or not, that kind of specific is really difficult to guess at even once the first example gets made (look at all the people looking at each version of GenAI, or even video game graphics 30 years before GenAI, and arguing that no further improvements are possible; or in the other direction, Concorde was both the first and the last supersonic passenger liner).
- namaria 3y agoYou got me thinking, could we distinguish engineered vs emergent life? I mean, I at look those animations of proteins in the cell and... it looks pretty robotic to me. Maybe it's projection in terms of how we interpret the data. But I really like that the answer to a most ancient question such as "what is life" is building up to something like "it's all nanobots controlled by molecular computers"...
- ben_w 3y ago> could we distinguish engineered vs emergent life? I think we can tell the difference between engineered and evolved (but note that simulated evolution is still evolution) — oddities like the path of the recurrent laryngeal nerve are cited as evidence against an intelligent designer, and I expect it isn't the only such oddity: https://en.wikipedia.org/wiki/Recurrent_laryngeal_nerve https://en.wikipedia.org/wiki/Recurrent_laryngeal_nerve > But I really like that the answer to a most ancient question such as "what is life" is building up to something like "it's all nanobots controlled by molecular computers"... If you've not already watched it, I recommend this video by A Capella Science: https://www.youtube.com/watch?v=ObvxPSQNMGc&t=198s https://www.youtube.com/watch?v=ObvxPSQNMGc&t=198s
- namaria 3y agoFlaws are not proof of non intentional emergence vs intentional design. We design flawed systems all the time. Besides, it would be a very naïve stance to think the universe as it is today was drawn complete and definitively. There are a lot more nuanced and intricate creationist worldviews and they have been espoused in a clear headed way by some of the most fundamental characters in the history of science.
- ben_w 3y agoIt's more the type of pattern in the errors rather than the mere existence of them. On the other topic, although the mere existence of errors would count against a supposedly all knowing and all powerful designer, that combination is not found in most religions, and not even all versions of Christianity.
- namaria 3y agoHey that combination is also not found anywhere else in this discussion.