Sam Harris Ted Talk on the danger of AI

[Berkut]

Take a look at this article:

https://www.wired.com/2016/03/sadness-beauty-watching-googles-ai-play-go/

It is really very interesting.

OK - that's about the Go AI.  Go, like chess, is a highly formalized game with a relatively simple set of rules, devoid of all social content.  Chess is also amenable to solution game theoretically - not sure if this is true of Go.  That's an area where AI has made very nice progress after decades of intensive work, but it is a pretty limited area of human experience.

The funny thing about your response is that is what people said about Deep Blue - chess is very strict, there are a finite set of moves, and the problem space can be defined well enough that it is amenable to a computer "solving" it...but Go! Now that is going to be tough for a computer to master! The rules are simple, but the possible moves astronomically complex - so much so that a computer could not just brute force it's way through, which is true, btw - I think the number of possible moves in Go is so large that no computer even possibly built can analyze them all.

Go requires intuition, spacial recognition, context, etc., etc.!

Yet it was beaten, and done so in a manner that is quite astounding. The computer Go player is not just a brute force stronger player mindlessly pounding away, it is actually *better* at the very "intuition" that was supposed to make Go such a difficult problem for a computer.

Modern AI is not really about understanding how humans think, and trying to replicate it. It is entirely possible that humans, in fact, might not be smart enough to understand how we think.

Modern AI is about teaching computers how to make themselves smarter, and doing so in ways that leverage *their* unique advantages. Advantages that they already have, and no human can ever possibly match. The ability to process faster, with more memory (perfect memory at that) and access data at a scale that no human can possibly even really imagine, much less use.

You should not be afraid of AI because it might think like a human better than a human, you should be afraid because it is going to think like a machine in a fashion that we won't even be able to understand, and in some ways, do not already

I accept all of this.  But the implication is that there are areas where it will be very difficult to devise AIs to match human capability.  As long as that is true, there will be demand for human labor and economies will adjust accordingly, just as occurred historically with other large scale events of capital for labor replacement.

Uggh, this refrain to history is over done. The reality is that the "history" of machine replacing humans is a blip in overall human history. We really have no idea how it plays out - we are at the very, very beginning of this process of replacing human labor with automation, and even that tiny beginning has seen radical upheaval in the human experience in a incredibly short time frame. In just the last few tenths of a percent of the human experience on a time scale, we've seen human populations grow by several orders of magnitude, human knowledge go from basically nothing to a huge amount, and human labor go from every single human working themselves to death to survive to us having the spare resources to fund gladiators and etertainers as full time occupations.

We are in the middle, or rather at the beginning, of a radical change in how humans live, and you want to tell us we know how it will turn out because we've seen the very, very start of the machine labor explosion curve. And we should assume that just because there has been enough additional useful work for humans to do before, we should assume there will *always* be sufficient additional untapped work to do...even while we see that there are already signs that this simply is not true. We have farmers in Africa starving because it is literally too cheap to grow corn in the USA.

We don't know that economies will adjust accordingly, and there is no evidence that is relevant to the impact in automation we are seeing over the last 100 years, and what we can reasonable expect over the next 100.

[The Minsky Moment]

Uggh, this refrain to history is over done. The reality is that the "history" of machine replacing humans is a blip in overall human history. We really have no idea how it plays out - we are at the very, very beginning of this process of replacing human labor with automation, and even that tiny beginning has seen radical upheaval in the human experience in a incredibly short time frame. In just the last few tenths of a percent of the human experience on a time scale, we've seen human populations grow by several orders of magnitude, human knowledge go from basically nothing to a huge amount, and human labor go from every single human working themselves to death to survive to us having the spare resources to fund gladiators and etertainers as full time occupations.

We are in the middle, or rather at the beginning, of a radical change in how humans live, and you want to tell us we know how it will turn out because we've seen the very, very start of the machine labor explosion curve. And we should assume that just because there has been enough additional useful work for humans to do before, we should assume there will *always* be sufficient additional untapped work to do...even while we see that there are already signs that this simply is not true. We have farmers in Africa starving because it is literally too cheap to grow corn in the USA.

We don't know that economies will adjust accordingly, and there is no evidence that is relevant to the impact in automation we are seeing over the last 100 years, and what we can reasonable expect over the next 100.

That's a long way of saying "we don't know nuthin'"
Which is sort of true but taking that far the implication is that it's pointless speculating about the impact now because we haven't a clue what will really happen.

The 200 years experience with automation and mechanization of work in human socieities across the globe, and the vast amount of data and research that has done analyzing that data - may not be sufficient to answer definitively.  But it is something - indeed it is all we have that related directly to the matter, so it's worth giving some weight to.

[Berkut]

The key component of a this time won't be different argument is specifying the mechanism that makes it different and making a persuasive argument how that will change the result.

A generalized human+ level AI would do it, assuming sufficient low marginal cost to produce and propagate.

But mechanization that can replace some functions but not others?  There is lots of historical experience with that across a variety of domains.

The mechanism may not be different.  It just may not be perfectly understood.  If I blindly reach into a bag of candies, for some time I will be pulling out a piece of candy.  I may start to think that I will always get a piece of candy by reaching into the bag.  But one day there will be no more candy left.  The mechanism of getting the candy didn't change from the first reach to the last reach, but the results eventually changed.

Indeed.

Human history is about 99.6% humans basically spending almost all their time trying to survive.

In the last 0.4% of human time, we've reached an inflection point in technology that has radically changed all that. We are still in the throws of that change, inded we are in the very beginning of that change, IMO.

Prior to this, it is pretty obvious that there was a lot of amazing work humans could do if only we could NOT spend all our time farming and gathering. We don't spend any appreciable time on that anymore, so we found all kinds of much more useful uses for our time. Namely, science and research - and not just science research, but research on art, and how to organize human activity both politically and economically.

But it seems kind of obvious to me that there is *some* finite set of useful human acivitities. But the ability of automation to reduce the need for humans to do things seems effectively unlimited in that we can easily imagine a situation where we have machines that can do everything that needs to be done, for all practical purposes, other than the purpose of "don't be a machine".

We took a bunch of people and said "Hey, you don't have to be farmers anymore! What should you do instead?" and they said "Scientist! Artist! Mathmetician! Factory Worker! Social Aid! Politician! Soldier! reality TV star!" etc., etc.

Nothing in that list seems to me to be anything that cannot itself be automated. The idea that not only will there always be some new thing to do, but that there is enough demand for that new thing that it can keep all 7-10 billion of us (even the dumbest and least motivated) busy strikes me as completely unsupported by the evidence.

[Berkut]

Uggh, this refrain to history is over done. The reality is that the "history" of machine replacing humans is a blip in overall human history. We really have no idea how it plays out - we are at the very, very beginning of this process of replacing human labor with automation, and even that tiny beginning has seen radical upheaval in the human experience in a incredibly short time frame. In just the last few tenths of a percent of the human experience on a time scale, we've seen human populations grow by several orders of magnitude, human knowledge go from basically nothing to a huge amount, and human labor go from every single human working themselves to death to survive to us having the spare resources to fund gladiators and etertainers as full time occupations.

We are in the middle, or rather at the beginning, of a radical change in how humans live, and you want to tell us we know how it will turn out because we've seen the very, very start of the machine labor explosion curve. And we should assume that just because there has been enough additional useful work for humans to do before, we should assume there will *always* be sufficient additional untapped work to do...even while we see that there are already signs that this simply is not true. We have farmers in Africa starving because it is literally too cheap to grow corn in the USA.

We don't know that economies will adjust accordingly, and there is no evidence that is relevant to the impact in automation we are seeing over the last 100 years, and what we can reasonable expect over the next 100.

That's a long way of saying "we don't know nuthin'"
Which is sort of true but taking that far the implication is that it's pointless speculating about the impact now because we haven't a clue what will really happen.

The 200 years experience with automation and mechanization of work in human socieities across the globe, and the vast amount of data and research that has done analyzing that data - may not be sufficient to answer definitively.  But it is something - indeed it is all we have that related directly to the matter, so it's worth giving some weight to.

Of course it should be given weight, but it is NOT all we have - we have the ability to think, to plan, to understand, and the ability to consider options and how we might respond given various assumptions.

You seem to be arguing that we should NOT think about such things, because either there is nothing to worry about because it all worked itself out in the past, or that we can't possibly know what will happen so lets not bother worrying about it.

I think we should be thinking about this a LOT. Much more than we do now. Much, much more. It is probably worth more consideration than things we spend billions and billions thinking about it....

[Valmy]

Well if we can do it effortlessly why do we need machines to do it? Aren't we supposed to make machines to help us do tasks that are hard to do?

Maybe I don't understand the point of AI.

I can give an AI an arbitrary amount of capacity to do more.

Yes. But to what end? I guess I figured machines were built to do specific tasks to help human existence. I don't see why we need a touchy feely robot, we have other people and animals for that stuff. Besides if I am mad that my sports team lost do I really want my freaking machines altering their behavior because of that?

[Tonitrus]

Well if we can do it effortlessly why do we need machines to do it? Aren't we supposed to make machines to help us do tasks that are hard to do?

Maybe I don't understand the point of AI.

I can give an AI an arbitrary amount of capacity to do more.

Yes. But to what end? I guess I figured machines were built to do specific tasks to help human existence. I don't see why we need a touchy feely robot, we have other people and animals for that stuff. Besides if I am mad that my sports team lost do I really want my freaking machines altering their behavior because of that?

All of those unemployed, former laborers will need sexbots to keep them from revolting. 

[Ed Anger]

Nerve staple them

[Zoupa]

The argument on the face of it makes no sense, really. We know that intelligence is possible, because we are intelligent. Is there something about the process by which the human mind cognates that makes it impossible to replicate?

Unless there is something magic about the human brain and how it thinks, there is no reason to conclude that non-biological intelligence is possible.

I'm not sure what you mean here.

[Martinus]

I am surprised this thread has gone for nearly two pages without Raz denouncing Sam Harris as a nazi racist Islamphobe.

[FunkMonk]

I am surprised this thread has gone for nearly two pages without Raz denouncing Sam Harris as a nazi racist Islamphobe.

This thread is about AI one day becoming Nazi racist islamophobes. Don't be so surprised man.

[Hamilcar]

I am surprised this thread has gone for nearly two pages without Raz denouncing Sam Harris as a nazi racist Islamphobe.

He also has a feud with Greenwald, so he also clearly hates the gays.