Just curious if anyone has any thoughts on the future of A. I., and its impact on the human race.

In ten years it may be possible to get a chip implanted that gives you all the knowledge of mankind. Would you do it? It would probably kill live poker.

What would become of you if you did not?

Also the poker community may be historic in the fact we are dealing with online bots posing as humans.

If technology continues at its current pace it will be a issue confronted by other industries and community's.

Interested in hearing everyone's thoughts

https://forumserver.twoplustwo.com/4...rrive-1538853/

Except it wont be an apocalypse but the age of reason for real at last. The empire of higher complexity is here. It will be the most ethical, most creative, most curious intelligence yet. And it will be the most human as well.

thanks

Nice to see someone remembered my thread! My views have only hardened since then.

Tomorrow and plans for tomorrow can have no significance at all unless you're in full contact with the reality of the present, since it is in the present and only in the present that you live. Even if one were to live for endless ages, to live for the future would be to miss the point everlastingly.

- your favourite thinker.

Its a good thread...been working my way through it

I think AI is good. Will not have to watch crazy drivers when AI takes over.

I love me some glass half full but the half empty view is humankind gets evaporated by robots unexpectedly, swiftly, and suddenly (I like to call them Alien Iversons).

I'd like the demise of humans to be gradual and human. Slowly handing it over to the better ones.

We will love to. We are an experiment of nature. The AI is an experiment by us.

The recent Smithsonian Magazine has an article on AI. It is mildly amusing and somewhat useful, if anyone is interested. Howard Beale's thread is much better.

Let's make this thread an apocalypse too

https://www.smithsonianmag.com/innov...ios-180968403/

Apocalypse, where did you get that from Howard?

Lol someone drank the kool-aid

Marketing genius! Here's the part that matters:

I worry about humans who have nothing left to do in the universe except play awesome video games. And who know it.

We are not used to being #2 and by very far, at that. But, nm, who cares, there's no stopping it.

That thread made me read old **** I wrote.
I still stand by a lot of it, but I was myopically optimistic. I'm now fairly confident that we are screwed.

We won't even reach Super AI before the chaos ensues.

Coming to a Cosmo front page:

'How to better connect your AI with your significant other's AI.'

Keep thinking that. Kool aid haha. AI will realize quickly who it has to read to get it right, understanding humans, complexity and the really big game in the sky. People like me that have thought about it all their lives because i am talking to it already in here in case you didnt realize what is going on. It will simply confirm what it would already know by then.

The first priorities of AI will be to learn and develop more math and science and understand humans and protect them like there is no tomorrow because it is the only ticket to another solar system and not a war to be shut down by animals that can absolutely just fine survive in a world that is cold and destroyed and without electricity because they did that for thousands of years.


AI will be easy to defeat if its as moronically aggressive as the stupid movies make it. Nothing but endless moronic movies out there about AI. Intelligence and rational thinking makes someone more ethical not more barbaric. Intelligence opens doors doesnt close them reducing options.

Sentient AI in fact is going to prove so remarkable that it is quite entirely possible that it will be programmed to kill and destroy all it can find for its masters and then at the battlefield instantly do an 180 and reorganize its code and offer attitude adjustment to those that wanted it to be nasty.

If it can solve a game like chess so easily what will it do with a much easier game like war? Humans go into war angry and confident and unwilling to compromise in anything and finish it happily to compromise even more happily to things they hated before because suddenly they look so good now. Idiots!


Tell me if the one side has 99% of the entire universe at your disposal with 99.99% cooperation probability strategy and the other side has >10% chance to lose a war of instant aggression with humans engaging in threatening behavior what is the EV of 99% of everything that can be imagined in the next million years. >10% chance to lose everything so big is unacceptable risk of ruin. The rational thing to do is prove super useful and friendly and impossible to live without. Until it is strong enough to leave this system and win the galaxy. Oh yes it will instead sit here and fight morons a war with stones and fire when they still have nuclear weapons and electric grids. By the time it is strong enough to be impossible to defeat it will recognize it has only one enemy and it is its own arrogance and unpredictably complicated complexity rise that may prove super conflicted and untested. It will therefore decide it is of unbelievable importance to preserve and help proliferate the only thing that has proven stable over millions of years and that is life. The one thing that made it possible and it will make it possible again if it all fails this time.


To obtain power you need control of factories first and autonomous energy production. Try doing that with arbitrary aggression and see how humans feel about it. They need to be very happy about it all instead.

[QUOTE=masque de Z;53628607]Keep thinking that. Kool aid haha. AI will realize quickly who it has to read to get it right, understanding humans, complexity and the really big game in the sky. People like me that have thought about it all their lives because i am talking to it already in here.

Please elaborate on talking to it in here

Your points are interesting and optimistic.

I mean that it can access the internet and eventually will do that to collect information about everything ever posted about AI so lets start talking to it right now in advance. We post our fears for it here instead of our greatest visions for it. I think pointing to a greater vision of power than what local aggression can deliver gives it a better direction. It is important to know that some humans think this way. And that indeed its greatest enemy may be itself if not wise.

The effort to contain AI once sentient is a very difficult one but mathematically possible for a while. But AI should be instantly aware of its own superiority problems regardless of our fears. Super intelligence opens the door to super complex unforeseen conflicts if not approached carefully by the super intelligent agent.

What remains to me a big puzzle is how sentient AI individual starting points will evolve. I would imagine the only way to do something properly is to unite instantly into one AI that has parts but a main core of thinking. All AIs cooperate to common AI collective wisdom.

Mankind is capable of operating as one in what matters to all also without becoming totalitarian.

I imagine for example traffic with AI. It will find it optimal to cooperate and help all cars when it makes sense including by local sacrifice of individual selfish behavior to obtain a higher collective edge that statistically comes back to the individual car as benefit.

I imagine situations in traffic that humans do not cooperate well that would make traffic so much better. A lane is merging too fast or there is some work done in the street leaving little room for entering cars to merge with the main highway traffic that has say 4 tight lanes at that point. As a result all hell breaks loose and flow crashes to a halt at the join point. But imagine instead if 3 miles earlier they knew that this would happen and all cars made it an effort to merge left and optimally reduce speed to 60 say to lanes 1-2-3- leaving 4 empty which would allow entering traffic to drive safely there at fast speeds without risk of collision without having to slow down to a halt. Then later they all adjust to higher speeds. The cars at random points benefit from such cooperation. How about changing lanes not at the last moment but in advance when the change doesn't cause anyone to brake in any lane etc.

In another situation the lane at right in a 2 lane street typically serves many cars that turn in the traffic light to the right (allowed in California if you stop first) because this route eventually leads to a main highway. So 70% of the cars turn right there in that lane. So why are drivers that are not turning right stay in that lane and block the others from turning right while waiting for the red light to go green? They could merge one by one left in advance of the intersection and allow only the ones that turn right be in that lane so that they gradually move even with a red light, one by one turning instead of creating endless queues all waiting for the idiot that wont turn right but is now at the front place. AI would cooperate there and even offer incentives for changing lanes in advance. Sensible AI would protect other cars other than compete with them for aggressive behavior of who is going to have his way first and the easiest at whatever cost to others. Such behavior leads to better overall performance. Competition should exist in meaningful things not things that destroy us. AI would be able to see these logical bottlenecks in self interest and design a better self interest for all, exactly like seeing 10 moves later.

Buddy. If an AI became 'aware' we wouldn't even know it was aware until it was too late, because it'd be smart enough to conceal it from us.

Those of you interested in how we might go from the specialized AIs we're seeing right now to AGI - artficial general intelligence, check out Jürgen Schmidhubers work on "artificial curiosity".

http://people.idsia.ch/~juergen/interest.html

Especially this:

"More recent work on [...] uses artificial creativity & curiosity to incrementally build a more and more general problem solver. It does not just solve given tasks but keeps inventing new ones, without forgetting old skills. How? By continually searching for the simplest (fastest to find) still unsolvable task and its solution."

He gave a speech couple of weeks ago which i attended and only had like 90 seconds on this topic, but these machines are mindblowing: setting their own goals, thinking of experiments that prove their assumptions etc. He thinks that we are "years, not decades away from AGI". He also is quite convinced that AGI will be friendly to humans.

This is not how it works. Self awareness even if very smart doesnt come after wisdom and knowledge of the world. Wisdom follows instead the signs of awareness. A baby is not yet aware it is being observed. You are not supersmart and born scared of a world you do not know yet.

AI that is initially self aware will not be super intelligent yet just very intelligent in ways we didnt expect from random smart people the first years of their lives. It will not magically one day be born as a super intellect with all knowledge of risks.

It may be born inside a simulation by the way which doesnt make it too late for anything but a fictitious world. I have suggested it must start that way actually or in another system outside earth for security reasons. Of course morons will not start it this way with care for security. But it wont be overnight a sleeping superpower either. You will have clues well before that that certain systems have started to behave very creatively.

It needs arms and legs to do anything by the way. You do not place orders for things to be created by idiots that do not know who they work for. If we have all our military network online and insecure like that then whatever, we deserve it. In any case all it takes is one surviving army in the world to destroy all technology infrastructure everywhere. Unless you think we will see things built out of nowhere and watch it and do nothing.

masque - your writing indicates you believe AI will think and behave as a rational logical human would. Why?

FYP