Hacker News new | past | comments | ask | show | jobs | submit login

I think the difference is that humans can fix problems and adapt to changing conditions a lot better than machines currently.



Humans plus Earth-like environment. Remember that micro environment has to go with them everywhere, and they die within minutes without it functioning within relatively narrow constraints.


But there'll still be humans. Just not necessarily physically at the mine or factory. All you really need is a robotic chassis that has sufficiently good AI to use hands and follow basic instructions, perhaps helped by a guy in a VR suit back on earth or in a nearby spaceship. Perhaps aided by simulations to ensure low latency - things in space should be easier to simulate without the outside world interfering much.


> perhaps helped by a guy in a VR suit back on earth or in a nearby spaceship.

Definitely not "back on earth". Lag is an issue with VR as-is with mere milliseconds of it, and you'd be calling for minutes of latency there.

And "in a nearby spaceship" would, definitionally, involve getting humans most of the way there; might as well get them all the way there at that point.


I think it's interesting to explore whether remote-control VR lag can be combated with simulations. If you can simulate the other side well enough, you should be able to manipulate the simulation until you've got a recording that you're satisfied with and then send it. This is kind of how Mars rovers are steered today.

Obviously doing this for robots on Earth is tricky and largely pointless, given it's probably easier to just reduce the latency directly. For remote mining on asteroids or something there's not a whole lot going on in space. You're not going to get a random pedestrian or cyclist suddenly popping out in front of your robot. It's far more practical to model and simulate it entirely in e.g. a game engine, and then record it.

At least, that's my guess where things will go.

The point about the spaceship is not distance related but that building life support structures into everything is quite expensive, whereas building them once in a spaceship and then extending the human presence to all surrounding structures would be a lot cheaper.


> If you can simulate the other side well enough, you should be able to manipulate the simulation until you've got a recording that you're satisfied with and then send it. This is kind of how Mars rovers are steered today.

True, but that has limitations, right? Meteorite impacts, electromechanical failures, and such still do happen. Plus on Mars there's still dust flurries and such, and then you've got Titan with its "dummy thicc" atmosphere blowing stuff around even harder, though obviously asteroids wouldn't have those issues.

With the Mars rovers we're basically in a position of "well we'll run them as long as we can, but if they get stuck in a rut they're hosed". Spirit and Opportunity would probably still be running and collecting science if we had human crews nearby able to hop over and fix them up.

> The point about the spaceship is not distance related but that building life support structures into everything is quite expensive, whereas building them once in a spaceship and then extending the human presence to all surrounding structures would be a lot cheaper.

Right, but my point there is that "surrounding structures" implies you're basically pretty much there anyway. You don't necessarily need life support structures in the mining equipment itself; you just need some spot for the ship to dock for a sec while the maintenance crews don their spacesuits and do their repairs.

Even if we did build human-habitable areas in those "surrounding structures" (e.g. break rooms, workshops, bunks), as long as there's a supply chain to provide air/water/food (and the first two could likely be extracted from the asteroid itself) I suspect it wouldn't be that much of an additional cost in the grand scheme of things, especially when weighed against the increase in productivity.

Plus, you gotta think on the flip-side, too: would it make sense to haul all those life support structures around on a moving ship? If you're sending crews someplace often enough it seems reasonable to have those on-site and thus not have to haul all that mass with you (then again, depending on the length of the trip such facilities might be necessary anyway).




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: