Chimpanzees granted 'legal persons' status to defend their rights in court

[Aeson]

You can have an evidence based prediction of the next 20ish years (I and HC's position) or a faith based prediction.

Your's is a faith based position where you place your faith in some nebulous (and I'm guessing so far uncalculated) statistical analysis of past events in CS or related fields and believe that it describes the timing of future events in those fields, even though this specific event is dissimilar. You're basically saying we're > X from an unknown destination, without knowing anything useful about the destination or how to get there.

My view here is an agnostic one where I choose not to believe in predictions about understanding of concepts which aren't currently understood.

I admit there is faith involved in my view as well, as there is in any view. In this case my faith is in the veracity of the claim "we do not know enough about how consciousness/human-learning actually works to judge how close we are to being able to replicate it on artificial platforms". I think it's a pretty solid claim, but nothing is 100% and I'm willing to cover the difference with faith.

Something shockingly new is much more likely to come out of genetics/biology/etc than artificial intelligence/computation/etc in the next 20ish years. This is because we can look at the history of the two fields and look at what has come out and what hasn't.

We may not need anything shocking in CS. We may (or may not) already have the hardware, maths, and language required. Or we may not.

The thing that obviously is missing from the equation is the understanding of how consciousness/human-like learning functions. That is as much a genetics/biology question as it is CS.

Modeling it after we understand how it functions is likely going to be just an eventuality. At that point you can make useful predictions, since you know where we're going and what's involved. But right now it's simply an unknown destination. You can't very well predict when we'll get there, or even how far away it is.

[Aeson]

I don't expect it to impact the layperson, unless we find extraterrestrial life.

What are your predictions about when/if we'll find extraterrestrial life?

[Jon Miller]

Your's is a faith based position where you place your faith in some nebulous (and I'm guessing so far uncalculated) statistical analysis of past events in CS or related fields and believe that it describes the timing of future events in those fields, even though this specific event is dissimilar. You're basically saying we're > X from an unknown destination, without knowing anything useful about the destination or how to get there.

No, it isn't faith based. It is based on that in the past, in every part of human advancement, we have known something about where we were going before we got there.

Often the fundamentals were studied and known to experts far before the shocking advance (to laymen) occur. The advances didn't come out of left field.

You believe there is something special about CS/intelligence/etc that means that it won't follow the rules. This could happen.

Just like aliens could show up in a spaceship. Or Christ could return.

But those things are faith based, we don't have evidence to make such predictions.

Unlike the advance of science for the last several hundred years.

JM

[Jon Miller]

What are your predictions about when/if we'll find extraterrestrial life?

I don't have any because I am not interested in making faith based predictions in that field. The difference from real 'AI' is that that field is not predicted to be scientific like real 'AI' is. It is more luck based and not based on human scientific development.

As a Christian of the evangelical persuasion, I predict that if the traditional view of Christ's return will happen then it will happen within 100 years.

If it doesn't happen within 100 years, I think a different understanding will be necessary because of the environment condition we put the earth into or because of advances in biology/genetics (and at the time scale of 100 years, possible CS like real intelligence of 'machines').

JM

[Aeson]

No, it isn't faith based. It is based on that in the past, in every part of human advancement, we have known something about where we were going before we got there.

Yes, it's faith based. You just aren't able to understand what you actually are relying on when you make your claims as to what will happen. You want to pretend it's hard science, but it's just pulling numbers out of your ass about issues you (and to be fair, all of humanity) are still in the dark about.

Often the fundamentals were studied and known to experts far before the shocking advance (to laymen) occur. The advances didn't come out of left field.

So are you saying that all advances and discoveries are always and always will be predictable before they are made?

You believe there is something special about CS/intelligence/etc that means that it won't follow the rules. This could happen.

There is a way which human-learning is unquestionably special. (Though not completely unique.) That you deny there is anything about it that sets it apart from the vast majority of other advances shows very clearly how much you are relying on faith to come to your conclusions.

Human-learning by definition would be a self-improving and self-expanding (as in capabilities) algorithm. Even lower order learning algorithms would be self-improving and self-expanding to some extent. We can look at our own development and see that our specific algorithm started with relatively very simple entities and progressed to what we are today.

So as development towards human-learning progressed it could help out with or even take over development. It's possible that even a very rudimentary learning program could handle [some of|much of|the rest] of development on itself. As it improved, so would how much it could handle and the speed of development. Judging how fast this would accelerate is of course impossible, but that acceleration by definition would have to be there at some point. It leaves the door open for development which only requires a very simple seed.

Another human-learning is special is that we have a very good historical analogy to draw off of. Human-learning developed from a (relatively speaking) rudimentary learning capacity. By accident. By looking at the speed of generations, communications/storage losses, time expenditures, mutation risk (for extent) ... we can see that a similar learning capability as what we started with, running on a computer, could develop much, much faster than we did as a species. The vast majority of our species' time and effort has been simply to survive. A computer platform wouldn't have to worry about that at all. Instead of thousands of generations taking 100,000 years to go from apes to modern humans, they could pass in a few years, or even faster depending on how complex it ends up being, and how much processing power we want to throw at it.

We would also likely be able to give some expert systems that would help speed it up or keep it on the desired course. Likely we could start it out with > ape learning capability. If just we understood how learning capability works.

Which brings us to the point ... we don't understand how it works. We don't know if that's a multilayered problem that we'll strip away at over decades that will require materials/maths/processes that won't be developed yet, or if it's something that can come from an "aha" moment and be implemented with modern technology.

Just like aliens could show up in a spaceship. Or Christ could return.

But those things are faith based, we don't have evidence to make such predictions.

My claim is highlighted. You are the one making a prediction about the timing of an arrival at a currently unknown destination. (You also are the one who indirectly suggested a time-frame for alien contact.)

[Jon Miller]

My claim is highlighted. You are the one making a prediction about the timing of an arrival at a currently unknown destination. (You also are the one who indirectly suggested a time-frame for alien contact.)

If we are getting there via science, then yes I can make a prediction based on the fact that we are not anywhere close. Scientifically.

I think that we had some miscommunication if you believe I suggested a time-frame for alien contact.

JM

[Aeson]

I don't have any because I am not interested in making faith based predictions in that field. The difference from real 'AI' is that that field is not predicted to be scientific like real 'AI' is. It is more luck based and not based on human scientific development.

As a Christian of the evangelical persuasion, I predict that if the traditional view of Christ's return will happen then it will happen within 100 years.

If it doesn't happen within 100 years, I think a different understanding will be necessary because of the environment condition we put the earth into or because of advances in biology/genetics (and at the time scale of 100 years, possible CS like real intelligence of 'machines').

JM

It was a rhetorical question

The similarities between someone finally understanding consciousness, and alien contact were amusing to me. Both are things that have been considered and studied for millenia ... both with essentially no progress. Maintaining a predictive timetable for either is rather silly.

[Aeson]

If we are getting there via science, then yes I can make a prediction based on the fact that we are not anywhere close. Scientifically.

We do not know what separates us now from what is required to achieve it. It may be something simple (or "quick" at least) or it may be something very arduous and difficult. It could even be (functionally) impossible. Putting a timetable on it with how little it is understood is simply guessing.

[Jon Miller]

So are you saying that all advances and discoveries are always and always will be predictable before they are made?

I am saying that before some scientific advance/discovery 'gets lose and greatly impacts the world' it is always studied and is predictable for some portion of experts before it is made. This is because before applied research gets done (which can cause impact), basic research gets done.

In this subject the basic research is still lacking.

Note that this is different than non-scientific discoveries, like aliens arriving at earth or similar such things.

JM

[Jon Miller]

We do not know what separates us now from what is required to achieve it. It may be something simple (or "quick" at least) or it may be something very arduous and difficult. It could even be (functionally) impossible. Putting a timetable on it with how little it is understood is simply guessing.

I can say that scientifically we are not close to a real AI, because scientifically the basic research doesn't exist yet.

Now maybe the time scale from the basic research to the real AI will be very short. And maybe the basic research will have some huge advance this summer.

But very short has still been greater than 5 years.

So I can still say that it won't happen within 5 years.

And even more, I can say that it won't happen accidentally or unexpectedly to the experts.

JM

[Jon Miller]

Once more, I am just talking about what I expect to happen.

It very well may be that Intelligence doesn't follow normal behavior, like aliens arriving or Christ returning.

But those are things to worry about from a faith based perspective.

Not from a scientific basis (and a non-faith based view should stick to the scientific basis and not assume different behavior).

JM

[Jon Miller]

I am passionate about this subject because I have read things by/about Hawking and Musk and so on and have not been able to reply.

JM

[Aeson]

I can say that scientifically we are not close to a real AI, because scientifically the basic research doesn't exist yet.

"Close" is getting "closer" it seems ... You've gone from "worrying about AI is stupid" to "not within 20 years" to "could happen in 5 years". At this rate we'll have AI in no time!

Now maybe the time scale from the basic research to the real AI will be very short. And maybe the basic research will have some huge advance this summer.

But very short has still been greater than 5 years.

So I can still say that it won't happen within 5 years.

Are you accounting for the self-development aspect of such an advance?

Let's say it was 5 years (or 20, doesn't really matter) with X processing power devoted to it. What happens if more resources are devoted to it? The nature of the issue (at the point "learning" begins) lends itself to faster development with more processing power.

And even more, I can say that it won't happen accidentally or unexpectedly to the experts.

The only way you can say that is if you know how it will happen, or at least what separates us from it happening. You can't give an accurate description of either of those things ... only guesses. Because you don't know what consciousness actually is. (None of us do yet.)

[Aeson]

Once more, I am just talking about what I expect to happen.

It very well may be that Intelligence doesn't follow normal behavior, like aliens arriving or Christ returning.

But those are things to worry about from a faith based perspective.

Not from a scientific basis (and a non-faith based view should stick to the scientific basis and not assume different behavior).

JM

Even with 20 years, worry isn't "stupid".

[Jon Miller]

Let's say it was 5 years (or 20, doesn't really matter) with X processing power devoted to it. What happens if more resources are devoted to it? The nature of the issue (at the point "learning" begins) lends itself to faster development with more processing power.

Huh? You have no basis to say that.

The 5 years is if I go on the extreme side (about the fastest I think basis research has gone to applied), if you take a more representative time scale is 20 (and like with fusion, that 20 can continue for decades). That is a representative time from complete basic research to complete application.

My problem with your statements is that believing in the most ridiculous case without basis is irrational. You can use them to argue that we are just on the verge of developing anti-gravity, faster than light travel, and contact with aliens.

JM