Tuesday, February 25, 2014

Comments on SETI, AI and UFOs by 'RW' and 'Nome'

Online Article:

"Alien Electromagnetic Signals Will Be Discovered by 2040" --SETI's Chief Astronomer

In my opinion, if the AIs are so far advanced then they'd know how to stay quiet, don't you think? Also, wouldn't they have figured out long ago that less advanced beings would try to eavesdrop on outmoded communication technologies. And wouldn't they have figured out where all the sentient beings are in their own galaxy? And if they are there and know about us, then why have we not been contacted in an unambiguous way? How can we know how an AI thinks? How can we know what they find important? How many of these types of questions can be answered?

The SEARCH for ET is part of our heritage of curiosity and exploration and our deep need to know if we are alone in this Universe. Will we know by 2040 that we are not alone? I hope it'll be sooner 'cos I'll be an old dude, by then!!!

Mebbe the AI is waiting for us to become AI before we can talk?
Mebbe the first AI figured out that being an AI wasn't so hot after all and it's better to experience this Cosmos as an organic being?

I like that we may not be able to tell the difference between an alien artifact and a natural one. This rings true to me. I think the technology would vanish from sight. It would alter things in the natural world but its action, its workings, would go undetected – we would just see effects.

As I read it:
SETI is shifting their search from X to Y. X was where possible
biologicals could be, Y is where machine AIs could be.
It sounds rational. I'm not sure how this XY shift connects to the 2040
date in the opening para.

The piece ends with some Wolfram quotes which help to cement the ideas
about AI, but are not necessarily correct.

In fact, what I find wrong here is this *inevitable* view of AI. I can
see how the views - AI exponential growth into the "Singularity" as well
as the programmable universe ideas - happen. They make a kind of sense.
  However, the data is not in and cannot be for a long time.
One argument against the AI boom (an AI that makes an AI that makes
another (or many) AI, with each being smarter) suggestion is that the
first AI or so will not necessarily be smarter than *groups* of humans
who have many different fields of study and knowledge.
  It took those hundreds of fields and thousands of material parts to
build that single AI. There's mining and commerce and trade and travel
and energy and the whole complex human world to negotiate. It all goes
into that AI. So, without the power to control all of it, how can the AI
"boom" the next AI?   I don't think there will be a boom. I think it will be a slow
evolution and feedback between us and machines.

I don't think it's unreasonable to look for AI life though. I'm not sure
if it ever got off the ground, but if it did then it surely had
advantages over biology and I agree it would seek-out energy and
resource rich areas that are also cool. Makes sense.
To a guy who touts the Alien AI thing, I said:

Why assume AI wants to meet us? I'd say the faster you get smarter, the
further from the neighbours you wanna be.

That, or the resources to grow AI parts drain at the same rate they do -
and suddenly the lights go out. A bit like Douglas Adam's whale that
popped-into being high in the sky and gradually discovered itself just
before it hit the ground.

Clever compaction into quantum-realms aside, the bigger a brain the more
it can store. The more it is said to know, the more storage it requires.
Very clever AIs will tend to be connected to very big data warehouses —
all literal material things. Google is not small. They have warehouses.
  The data from a single run of the LHC (Large Hadron Collider) is vast. It requires room to
hold the tape (yes, tape!) machines.

All these things require matter to build bodies (brains are parts of
bodies) and they'll require energy too. I surmise that before an AI civ can reach the crescendo they require to get off-planet, they will face severe shortage of materials. Either because they are all mined-out and gone, or because organics oppose them.

I'm thinking along lines of simulations of AI running upon other simulations of AI and the supernormal amount of storage and processing power this will all take.

All this would be silly if multiverses are real and computing happens
across them. In that case every atom in each 'verse could be part of a
cross-verse AI running in unimaginable parallel ways, but now I sound
like Wolfram. ;)

I don't really know what the energy requirements of a brain vs a CPU
are. I suspect it costs a *lot* more to keep a brain going. At any rate,
the sheer complexity of brains dwarfs the CPU race.

Putting billions of dumb CPU's together is not leading to an AI either;
the problem is in a different class to that.

Hey, I feel all kind of wrong about some of what I just wrote. I need
info that could take me years to get and then internalize. I can't give
you a good answer, but I feel that *rapidly* rising super-smart
beyond-human AI is a pipe dream. I feel that a (possibly disastrous)
merging of flesh and tech is what's going to happen first. For now.

:So, yeah - I am close to what you posted to that site. How can we
presume to know how an AI would think - if an AI even exists?

I have been posting to two AUFO types as well. Here's my position on
what I'd need to accept a AUFO (Alien UFO):

For myself to accept a legitimate UFO (of any origin), I'd need to be
*convinced*. This is a complex state of mind that I cannot really
describe before the event. It's not a strict category thing. I require a feeling of being
"convinced". Depending on the situation this will be hard or easy to do.

Any UFO of any origin would have to satisfy my criteria - even though anything man-made would already have satisfied many of them.
An Alien UFO or man-made UFO event would require:

1. It would start with evidence that has few holes in it.

2. The evidence behind that evidence: the harmonious meshing of times,
places, names, faces, stories and traces would all chime a true note.

3. After some time and study by many people, the facts would fit into
the nature and reality of the world - even if it opened new avenues.

4. The ripples from this finding would inform and grow other sciences.

5. There would be more findings like it because knowledge helps us find
things; it sharpens the search.

6. The story would not go stale like bread left in a school bag, the way
so many UFO/Cryptid tales do. It would mature and connect, dovetailing
into reality and our future.

Something like that. A few UFO events that have been rationally explained, have satisfied me. They are the ones
that are shown to be mundane and the how and why behind them. I recall the oil-rig flames one and another about military flares.

In the case of the oil-rig flames event, for example, I see no reason for mystery. The observed/reported facts are well matched by mundane hypothesis and evidence. These are UFO sightings that boiled-down to "bright light from burning things": show the burning things, match their position in space and time with the sighting and close with a flourish.   No more need be said because it's not surprising that a UFO should be seen where a normal thing was at the same time and place (or the illusion/reflection of same.) Especially since the original reports are so thin on any other convincing details.

No comments:

Post a Comment