Hacker News Reader

Comment by acdha | Hacker News Reader

I agree having grown up on sci-fi about that but increasingly wonder whether we’ll get enough AI to change the cost dynamics for sending fragile meat-bags to a harsh environment instead of robots. Also, I suspect climate change is going to force NASA to direct an increasing percentage of their work to earth, but I hope the symbolic potential keeps us from looking completely inward.

View on HN

Strong enough AI, probably. Until then, light speed dictates that even remote robotic exploration would work better with a command&control facility at the scene.

How about in geosynchronous orbit? So the control facility doesn't have to deal with the fine particles.

That is what I expect how things will be done. High orbit saves you not just from the fine dust, but also all the Δv you'd need to land people and get them up later.

Oh, no argument that it’d be better in terms of quality but I’m wondering how it’d look from a cost perspective. Say we could make a self-driving rover which could handle basic maneuvering, avoiding / recovering from hazards and storms, etc. – at some point there’d be a trade off between having many rovers running for years for less budget than a human crew would need, and depending on the budget climate it might be better to have a cheaper option that doesn’t fail to deliver any scientific results if a big dependency isn’t funded.

You also don’t need smart rovers. You just need dumb rovers with a bunch of sensors and a radio, a few satellites and then a local command server with all the processing power centralized.

You run an AI on that with a bunch of safeguards, and constant human albeit speed-of-light-delayed monitoring that can correct erroneous decision making.

Stop poking holes in my desperate attempt to save the future of manned space exploration! :(.

(If we keep reasoning this way, humanity will forever be stuck on this ball of mud, eventually wiring our heads to computers, and then the closest we'll come to exploring the universe is browsing through pictures snapped by ancient probes.)

My very far future prediction is that either:

A. We retreat in the comfy matrix whilst we let robots rule the roost. If the matrix could simulate.. say.. Star Trek style exploration for you in a more interesting, comfortable and challenging way than reality, why would you want the watered down real version?

B. We digitize ourselves, but (most of us) prefer reality. To travel to a different world you get beamed over, into a vat-grown android body.

Human bodies are much too fragile for long distance space travel. Why add massive amounts of extra required resources (cryo-G couches, extra fuel due to those couches, bigger ship for those couches, redundancy, etc.) when you can just launch a robotic factory at much higher Gs, land (always perilous), let it bootstrap itself into a forward base, vat-grow a couple of android bodies and then you get to beam into the frontier :)

For the not-so-far future I see us putting a few bases on the moon and a very scrappy base on Mars, all manned by warm-blooded humans.

You’d probably want some intelligence in both places to improve the odds if a survivable mishap or dust storm breaks communications, but yes, keeping it in orbit does have some really nice properties for a longer running program.

Reply on news.ycombinator.com