Speeding up space coloniization

https://en.wikipedia.org/wiki/DeepMind
*AlphaZero* is a computer program developed by the Alphabet
-owned AI research company
DeepMind , which uses an approach
similar to AlphaGo Zero 's to
master not just Go , but also chess
and shogi
. On December 5, 2017 the DeepMind
team released a preprint introducing AlphaZero, which, within 24 hours,
achieved a superhuman level of play in these three games by defeating
world-champion programs, Stockfish
, elmo
, and the 3-day version
of AlphaGo Zero, in each case making use of custom tensor processing units
(TPUs) that the
Google programs were optimized to make use of.[1]
AlphaZero
was trained solely via "self-play" ...

Humans are still in control, but the future looks bleak for humans. This
development has taken place only recently, 2017, SF writers may not have
fully digested what has happened.

Humans could develop AI in a controlled way, may be embedding Isaac
Asimov's three laws of robotics in all computer systems. This is however
unlikely to happen as AI seems to be part of military plans of many
countries.

The great philosophical question is, will colonization of space take place
with Humans and AI working in tandem or will AI decide that they could do a
better job on their own. We know that organic life has purpose on Planet
Earth. Does it have purpose in the cosmos?

Regards,
Selvaraj

On Fri, 13 Jul 2018 at 19:13, 'Combs, Mike' mikecombs@...
[spacesettlers] wrote:

>
> > It looks as if we are rapidly moving beyond tele-robotics to AI-robots.
> The reality seems to be stranger than fiction.
>
> Id have to respectfully disagree. My observation is that AI-robotics is
> advancing at a much slower pace than what SF might have inclined us to
> expect.
>
> I remember when I was very young, I came across a newspaper article
> talking about how we would have robotic chauffeurs, gardeners, police
> officers, judges, etc., and by what decade to expect each of them in. The
> decades projected were the 80s, 90s, and 00s. (I remember sealing the
> article in a time capsule to see how accurate the predictions proved to
> be.) With the exception of the chauffeurs, I dont think an argument can
> be made that the anticipated levels of advancement have happened in the
> time frames envisioned. And even in that case, the implementation is
> imperfect and a long way from being in mass use.
>
> On the other hand, I think a lot could be done with human-guided
> tele-robotics. If we could just have a few advancements in feedback
> systems, we could be doing things like mining the moon or building things
> in HEO without having to wait on human-level AI (which might still be
> several decades off). Now there might be lots of simpler tasks that
> short-term AI would probably be perfectly adequate for, but I think running
> the entire operation would require at least human-level AI. So were
> either talking about largely automated facilities with at least a small
> number of human directors and technical trouble-shooters, or facilities
> where automation and tele-operation work in concert. And only the latter
> could save money by entirely eliminating the requirements for life-support
> and human transportation back and forth.
>
> > It is interesting to note that science fiction writers have not
> anticipated the growth of AI.
>
> I find this a puzzling remark. Human-level AI has been a part of SF from
> the very beginning. And we are still very far away from human-level. On
> the other hand, I think it is correct to say that SF writers never
> anticipated that orbital habitats will have so many advantages over
> planetary colonies that they will likely dominate the future beyond the
> Earth.
>
> Regards,
> Mike Combs
>
> *From:* spacesettlers@yahoogroups.com [mailto:
> spacesettlers@yahoogroups.com]
> *Sent:* Thursday, July 12, 2018 10:09 PM
> *To:* spacesettlers@yahoogroups.com
> *Subject:* Re: [EXTERNAL] [spacesettlers] : Speeding up space
> coloniization
>
> It looks as if we are rapidly moving beyond tele-robotics to AI-robots.
> The reality seems to be stranger than fiction.
>
> It is interesting to note that science fiction writers have not
> anticipated the growth of AI.
>
> Regards,
>
> Selvaraj
>

AlphaZero

is a computer program developed by the

Alphabet
-owned AI research company

DeepMind
, which uses an approach similar to

AlphaGo Zero
's to master not just

Go
, but also

chess

and

shogi
. On December 5, 2017 the DeepMind team released a preprint introducing AlphaZero, which, within 24 hours, achieved a superhuman level of play in these three games by defeating world-champion programs,

Stockfish
,

elmo
, and the 3-day version of AlphaGo Zero, in each case making use of custom

tensor processing units

(TPUs) that the Google programs were optimized to make use of.
[1]

AlphaZero was trained solely via "self-play"
...
Humans are still in control, but the future looks bleak for humans. This development has taken place only recently, 2017, SF writers may not have fully digested what has happened.
Humans could develop AI in a controlled way, may be embedding Isaac Asimov's three laws of robotics in all computer systems. This is however unlikely to happen as AI seems to be part of military plans of many countries.
The great philosophical question is, will colonization of space take place with Humans and AI working in tandem or will AI decide that they could do a better job on their own. We know that organic life has purpose on Planet Earth. Does it have purpose in the cosmos?
Regards,
Selvaraj
On Fri, 13 Jul 2018 at 19:13, 'Combs, Mike'
mikecombs@...
[spacesettlers] <
spacesettlers@yahoogroups.com
> wrote:

> It looks as if we are rapidly moving beyond tele-robotics to AI-robots. The reality seems to be stranger than fiction.
Id have to respectfully disagree. My observation is that AI-robotics is advancing at a much slower pace than what SF might have inclined us to expect.
I remember when I was very young, I came across a newspaper article talking about how we would have robotic chauffeurs, gardeners, police officers, judges, etc., and by what decade to expect each of them in. The decades projected were the 80s, 90s, and 00s. (I remember sealing the article in a time capsule to see how accurate the predictions proved to be.) With the exception of the chauffeurs, I dont think an argument can be made that the anticipated levels of advancement have happened in the time frames envisioned. And even in that case, the implementation is imperfect and a long way from being in mass use.
On the other hand, I think a lot could be done with human-guided tele-robotics. If we could just have a few advancements in feedback systems, we could be doing things like mining the moon or building things in HEO without having to wait on human-level AI (which might still be several decades off). Now there might be lots of simpler tasks that short-term AI would probably be perfectly adequate for, but I think running the entire operation would require at least human-level AI. So were either talking about largely automated facilities with at least a small number of human directors and technical trouble-shooters, or facilities where automation and tele-operation work in concert. And only the latter could save money by entirely eliminating the requirements for life-support and human transportation back and forth.
> It is interesting to note that science fiction writers have not anticipated the growth of AI.
I find this a puzzling remark. Human-level AI has been a part of SF from the very beginning. And we are still very far away from human-level. On the other hand, I think it is correct to say that SF writers never anticipated that orbital habitats will have so many advantages over planetary colonies that they will likely dominate the future beyond the Earth.
Regards,
Mike Combs

From:
spacesettlers@yahoogroups.com
[mailto:
spacesettlers@yahoogroups.com
]
Sent:
Thursday, July 12, 2018 10:09 PM
To:
spacesettlers@...m
Subject:
Re: [EXTERNAL] [spacesettlers] : Speeding up space coloniization

It looks as if we are rapidly moving beyond tele-robotics to AI-robots. The reality seems to be stranger than fiction.

It is interesting to note that science fiction writers have not anticipated the growth of AI.

Regards,
Selvaraj