DeepMind's AlphaZero plays chess like a tornado in the junkyard

[In situ]

A possible interpretation is that evolution is itself a simple (although inefficient) learning algorithm that learns how to make creatures well-adapted to many environments, and in time, creatures who can adapt to many environments. The most successful of these eventually use learning to adapt environments to themselves.

Notice thought you cannot trivlialy find this algorithm by study the parts. The effect is visible, i.e. unique inherent diversity, but the cause is subtle. Analogoues, you cannot trivial find the software in a computer by study the hardware, but the effect is there to see. As such both evolution and software execute in an abstract space.

 

[In situ]

Well its relevant now, because you stated AZ didnt have a telos, and if so, that makes its decision-to-learn a stunning breakthrough of deep consequence.

Please read the OP. I stated in the OP that AlphaZero has telos and it is irrelevant for the discussion. If you then bring up telos I will, reasonable, assume you mean some kind of other goal been programming into AlphaZero, and it was that second telos I asked for. You then made clear you referred to the first telos. Therefore, I referred you back to the OP. I can then only assume you never read the OP, and instead of reading the OP you persists in breaking up open doors.

 

[In situ]

Ummmmmm....SOMEbody had to "create" the robot and its "intelligence" from the gitgo.

Didn't they?

Yes, but I mention telos in the OP so it is unclear to me what your point is supposed to be.

 

[jacks]

You could argue that the tornado was given the rules of nature.

I don't think I could argue that successfully. Which rules of nature would allow a tornado to create a Boeing 747 in a junkyard?

 

[brinny]

Yes, but I mention telos in the OP so it is unclear to me what your point is supposed to be.

...the point is, that none of what you posted just came into "being" from nothing.

 

[FrumiousBandersnatch]

Notice thought you cannot trivlialy find this algorithm by study the parts. The effect is visible, i.e. unique inherent diversity, but the cause is subtle. Analogoues, you cannot trivial find the software in a computer by study the hardware, but the effect is there to see.

As an analogy it's a bit of a stretch, but I take the point.

As such both evolution and software execute in an abstract space.

Conceptually, all processes execute in an abstract space. In practice, evolution executes in a natural environment, and software executes in a computer.

 

[Larniavc]

I don't think I could argue that successfully. Which rules of nature would allow a tornado to create a Boeing 747 in a junkyard?

I’m not saying I agree with it.

I think the junk yard argument is silly.

 

[Larniavc]

Notice thought you cannot trivlialy find this algorithm by study the parts. The effect is visible, i.e. unique inherent diversity, but the cause is subtle. Analogoues, you cannot trivial find the software in a computer by study the hardware, but the effect is there to see. As such both evolution and software execute in an abstract space.

Indeed. I would look quite irreducibly complex with a specified complexity co-factor of 7K-*

 

[Speedwell]

...the point is, that none of what you posted just came into "being" from nothing.

So what? That is not what is being argued.

 

[In situ]

Interesting OP. (I love chess.) I agree that we get many interesting variations from randomness. However, I don't know if the Alpha Zero program was truly only random. As you stated it was "given the rules of chess". That seems to be like a big leap up from assembly a Boeing 747 in a junkyard. Unless the program was given detailed designs of a Boeing and then just created the jet from the parts available. This seems different to me than just randomness.

Where did humans find the blueprints for building Boeing 747's or computers? The answer is; by observing the effect of the laws of nature. Post #86 explain this in further details.

 

[Speedwell]

As an analogy it's a bit of a stretch, but I take the point.

Conceptually, all processes execute in an abstract space. In practice, evolution executes in a natural environment, and software executes in a computer.

The hardware aside, is the comparison of execution an apt one?

 

[FrumiousBandersnatch]

The hardware aside, is the comparison of execution an apt one?

Well, they're both processes, they both 'run' or 'execute'; frankly, I'm not really that keen on the comparison overall.

Any analogy that takes this much explanation is of questionable value.

 

[Chesterton]

I refer you to post #81 and post #86 for my arguments why the origin is irrelevant as a distinction.

I don't see how what you've said does away with the importance of origin. Again, I have to quibble with you tossing in suggestive words. You say:

"As such, AlhpaZero can be seen as an example for the mind to accept the idea that a process can mindlessly design things only determined by the laws of the word,..."​

Alpha did not design anything. Let it design a living 3D baby boy, then we'll have an analogy to evolution.

"...AlphaZero is a process which only been told how the world works, i.e the "natural laws" of the world it exists in."​

If it needs to be told, then origin (an agent doing the telling) is relevant.

"Living beings are evolved by evolution to survive and reproduce, i.e. "win", under the rules governing our reality. Therefore living beings directly corresponds to the winning rules AlphaZero has discovered."​

No. Surviving and reproducing are not "winning" without telos. Laws don't care. It's not analogous to playing a game with a goal. So the origin of the laws is relevant.

And I am telling you it is not a good label since there is no such thing as "atheistic evolution", you could just as well had made the distinction by saying "scientist evolution" or "polar bear evolution". It still does not make much sense to me to single out a specific group which shares properties only superficial correlates with naturalistic evolution.

Then you must object to "theistic evolution" also, yet you've used the term.

Beside the objection I already made there is another objection to the above and that is the null position. The null position is that you cannot claim the above to be either true or false. If the null position is valid then your argument is invalid.

This is why:

One cannot assumes that order is a degenerated state of design and that order cannot create design, since we know design can emerge from order and the other way around. Order was given to the world of AlphaZero in the form of the chess rules it been provided with. Since there now exists order in the mini world of AlphaZero, then it implies that design can, in principle, emerge from this order. Which is precisely what happen (see post #81 and post #86). In a similar fashion, the laws of nature impose order on the universe. This order is, in principle, sufficient to explain "DNA and consciousness and all the biological stuff".

To know if the null position is valid we first need to ask: was the the order in the universe imposed on the universe by an agent? This is a question on whether or not order can emerge from chaos. The answer is; it can.

Hence the null position is valid.

Therefore we can just as well assume the universe emerge from chaos. This implies AlphaZero can emerged from chaos and therefore your claim that "AlphaZero could not and would not happen naturally" is invalid.

I think you're bording on sophistry here. If there are laws, there is not chaos. You can't claim there are both simultaneously. And I don't how you can refer to a team of highly educated and talented intelligent agents as "chaos".

At a fast glance at this quote it seems the author might have conflated different concepts. See post #86 and maybe post #81 to understand my argument against the objection that AlphaZero did not learned something from the self play and why the rules of chess must be given as domain knowledge.

Yes, the rules must be given (by an intelligent agent). Again, you're just making a decent argument for theism. If you're trying to make an argument against creationism, I don't think it's effective because we already have clocks and gas combustion engines and all manner of machines which "do things". AlphaZero is impressively complex, that's all.

This article raised a red flags in my mind already at the start.

[snip]
Seems harsh. From what he wrote I see no reason to question his motives. He's not the only one, there are other critiques out there. I didn't want to just hand you a whole bunch of links so I picked one. You can Google for more if you're interested.

 

[Speedwell]

Alpha did not design anything. Let it design a living 3D baby boy, then we'll have an analogy to evolution.

Similar algorithms design computer chips. Real 3D computer chips.

​

Then you must object to "theistic evolution" also, yet you've used the term.

As a theist who also accepts the science of our origins, I object to it myself. There is no theistic evolution versus atheistic evolution any more than there is theistic thermodynamics versus atheistic thermodynamics.

. If you're trying to make an argument against creationism, I don't think it's effective because we already have clocks and gas combustion engines and all manner of machines which "do things.".

As I understand In situ's argument, it is not against "creationism" in general, but more specifially against the claim that random variation and selection cannot produce the biologial diversity which we observe.

 

[Chesterton]

Similar algorithms design computer chips. Real 3D computer chips.

Yes but the chips don't cry and burp.

As a theist who also accepts the science of our origins, I object to it myself. There is no theistic evolution versus atheistic evolution any more than there is theistic thermodynamics versus atheistic thermodynamics.

I see nothing wrong with either term. They simply identify whether one believes God is responsible.

As I understand In situ's argument, it is not against "creationism" in general, but more specifially against the claim that random variation and selection cannot produce the biologial diversity which we observe.

Again, Alpha produced nothing.

 

[In situ]

I’m not sure if pattern learning algorithms are strictly analogous to the random generation of mutation

The analogue to mutations are the random selection of nodes in the search tree by the Monte Carlo Tree Search (MCTS) algorithm. The algorithm does not correspond to specific details such as mutation but instead the entire process of evolution itself. The selection of nodes by the MTCS correspond to natural selection in nature, i.e. the survival and deaths of organisms in nature. Mutations are analogous to to the random pick of nodes in the search tree by MCTS. The nodes that are picked in the first round are twhen evaluating a chess position is a "parents" ready for further evaluation, further down that branch in the search tree. This correspond to a parent making offsprings with unique inherent differences.

The purpose of the pattern learning algorithms is to generate a winning solution. The analogue to a "winning solution" is a living organisms which successfully manage to reproduce. In other words MCTS "design" winners which is uses for the next round of games. The selected winner will then compete as parent, ready to give birth to new offsprings, in the next move of the chess game. And so on and so on.

Hence, while playing chess, i.e. obeying the rules of the mini world, AlphaZero discover how to play and win a chess game. This without being told by anyone how to play chess. In doing so AlphaZero develops a knowledge of winning chess moves. The minimal precondition to be able to play chess is that you know the rules. This is why the rule must be give as minimum knowledge. Analogous, the rules for evolution is the Laws of Nature. With these rules evolution then plays the Game of Life. In doing so evolution evolves or design, all by itslef, winning organisms in the same way as AlphaZero plays a game of chess.

 

[Speedwell]

Yes but the chips don't cry and burp.

If that's what you want then you need a different algorithm. Fortunately, God has given us a good one for that outcome.

I see nothing wrong with either term. They simply identify whether one believes God is responsible.

As applied to individuals yes; as applied to the theory itself it's nonsense. There is only one theory of evolution, not one for theists and another for atheists.

 

[durangodawood]

....What AlphaZero discovered was how to act in the world, i.e. how to win, or survive if you so like. There was no human involved in guiding AlphaZero on how to discover how to act in the world....

Didnt the algorithm tell AZ how to do exactly that?

 

[durangodawood]

....Alpha did not design anything....

Thats just wrong.

It designed a way to win chess, which is pretty impressive. Not only that, but many aspects of the type of play it designed are novel and astonishing to chess experts.

 

[In situ]

Playing chess is not in the same universe as assembling a Boeing 747 and assembling a Boeing 747 is not in the same universe as creating creation.

The argument in shortform is like this:

1) The junkyard tornado is analogous to the creationists claim that evolution is not viable.
2) AlphaZero's ability to play chess is analogues to evolution (see post #136).
3) If AlphaZero can play chess, then evolution must be viable.
4) 3) contradicts 1)
5) Therefore there must exists significant differences
6) What are the differences which prevent evolution from being viable?

To which of these points do you have an objection?

I've taught people to play chess but assembling Boeing 747s?

Almost everyone answering 6) has responded with telos as the difference. This answer assumes evolution must be a natural process. Therefore creationists are right and evolution does not work. However, anyone can claim theistic evolution works, therefore creationism is wrong. However, this is not the claim creationist makes. Creationists claims evolution does not work at all. Therefore the telos answer must be irrelevant as an answer.