Like Morphy and Fischer Alpha Zero is done.

You were wrong then and you're also wrong now on a completely different front...let's call a spade a spade.  Until AZ plays in an official match against the latest  version of Stockfish, not just some informal games tacked onto their Stockfish 8/9 experiment, they still are hiding the truth, just like their first backroom "experiment" .  AZ proved it had +50 Elo over an early release of Stockfish 9 (Jan 2018).  Stockfish 10 has a similar increase in Elo.  AZ and Stockfish 10 have not had any official engagements.  Stockfish 9 won numerous games as white once it was run as designed, with databases.  Stockfish 10 is untested.  That is the actual provable state of affairs as things sit.

"Today, Ford proved that a 1994 Taurus is the fastest car in the world...in numerous races against top of the line Formula 1 racers, the Taurus won every heat.  The Formula One racers engines all died within minutes of starting, due to using 86 Octane regular gas with Techroline from Chevron, which was the conditions chosen by Ford for this demonstration.  Ferrari and McLaren/Honda protested in vain that it was not fair to run the races on a closed track owned by Ford's CEO.  The press hailed the results as controversial and click-worthy, and passed them right along to the public without a lick of scrutiny or understanding of what they were reporting."

Google/Alphabet got what they wanted with their press release, the mindshare of the masses.  The people in the chess world and in the know in the development community know that these results are probably real, but still dubious.  Ergo this quote from the Chess.com article:

"The results leave no question, once again, that AlphaZero plays some of the strongest chess in the world."

Wow, really definitive statement there.

The members of the DeepMind team that are avid Chess players have their own agenda beyond Google's corporate agenda, and so will keep attempting to aggrandize themselves by using unverifiable backroom results, until they actually have a version that flat out beats the best chess engine.  Then they will emerge for a real match and let the accolades wash over them  ...it's a coward's way of competing without actual risk.  If they never can beat Stockfish, or some future version of Leela, or whomever it is, they will take the same out that the OP did, and claim they just didn't care enough.  Until you take your rocket car to the salt flats of Utah and do an official run timed by 3rd parties, you can claim you have the fastest land vehicle 'til the cow's come home.  It's just posturing.

Meanwhile, OP, your incessant arguing that the DeepMind team didn't care a whit about chess is obviously laughable at this point.  They wouldn't have released more dubious claims if some one or someones on the team were not heavily invested.  In Anna's visit to DeepMind, you can see this investment first hand in a clip where a DeepMind team member (who looks amazingly like the Candidate Master chess player on the team) is being told by his colleagues "you've been trying to do this for 2 years", and the giant smile on his face.  Oh...and then there's the fact that Google volunteered their London HQ, where DeepMind sits a floor or two below for the London Chess Classic, coincidentally dovetailing with the timing of this press release.  Nah, they never cared at all...

I admitted I was wrong. You should admit your a jerk.

"you're"

I'm just reflecting your own attitude back, read the thread again...and, your admission was just preemptive, because you knew you were going to have to eat crow.  The difference between you and I is that you are dismissive of everyone that disagrees with your overblown conclusion in this thread, where I am just dismissive of you (and the troll/lapdog that jumped in to help you).

Don't make me have to go back and quote the dozen or so one liners in this thread that make you look quite the a$$hat at this point...*that* would make me a jerk.  But I will if you ask for it.

Get a life.

I'll take that as a jerk's way of saying "no, I'd rather you didn't point out the many times I said things with 100% certainty and [more importantly] ridiculed anyone that would think otherwise, which then turned out to be wrong".

Of course I will accept Leela's results, for the obvious reason that Leela is playing in official competitions...

It's not my conclusion to make, that's kind of the whole point.  If A0 competed, there would be no question.  If you are assuming that I do not think A0 could possibly be stronger than Stockfish, you haven't been reading very closely.  Whether it's true or not is not even directly relevant to the discussion at hand, which is Google making press releases about unscientific and unofficial results.

Must be nice to live in your world where you can just make up definitions to fit your narrative.  Sentence 2 does not support/prove your conclusion in sentence 1.  Paragraph does not follow, either.  Something actually "in the spirit" of science does not use the phrase "in the spirit" to support itself ...nice paradox, though.

Oh, they did it in a 1000 game match in early 2018, just published in December. AZ won again.

I guess you missed the part where I said "official match", not "we downloaded the latest version, configured it ourselves, tested it in a back room vs. A0 until we were satisfied, then told the world what our unverifiable results were" ...

Until that issue is corrected, it really doesn't matter which version they use.  They did a  couple of official matches for Go, so...what is the reason for not doing so for chess?  Not like it's at all expensive.   They sponsored the London Classic and touted A0 there, which cost them far more expense and effort than an official match between computers.  This is the reality:

- They tested A0 vs. Stockfish behind closed doors

- They made a press release characterizing their testing as if it were a real match (presumably with Stockfish team participating for the mainstream public, but they just omit the reality)

- They spent more money and manpower sponsoring a high profile chess tournament and touring the press around their facility than it would have taken to just play a real match against Stockfish, which would have answered the questions many people have voiced definitively

So, if A0 can really wipe the floor with Stockfish 10 straight up with both programs configured to run at their best as they are designed to (opening book + tablebase), it seems like there is zero reason for not doing so officially...it's cheaper and better publicity than the course of action they have already taken, so it stands to reason that there is something they are not willing to risk.  Does that mean A0 is not already better than Stockfish 10, no, but maybe it's too close right now and they aren't willing to take a public loss even if the odds are in their favor.

You make these absolute statements that are plainly false all the time.  Maybe it's a translation thing?  Or maybe it's just your ego...

So, apparently, you'll say anything to prevent being wrong.  Now you're saying that IBM's research was "academic" when Deep Blue was a promotional campaign for the IBM brand, but Stockfish research is "commercial" when it's open source and makes no significant money?  Your logic circuits are fried, it seems.  The ratio of "commercial" to "academic" for Google is pretty much the same 20 years later as IBM's ratio was.  They both sponsored expensive chess events for brand publicity/prestige...the only real difference is, Deep Blue beat Kasparov in an official match, where Alpha Zero has not beaten a single opponent in any official match...

As for A0 playing "any game", yes, this is somewhat true from Google's overall point of view, though chess is still seen as the most prestigious "intellectual" game for the western world at least if not arguably the whole world...but it's not true for the Deep Mind team.  The Deep Mind team has a prominent member who is/was a candidate master and has a clear and present drive to see A0 do well in the chess world.

They also didn't make any public matches with AlphaGo Zero and AlphaZero for Go anymore. Why? Because it has no great value and and is of no interest to them.

Engine vs engine games are not interesting to the public. Go was special, because the strongest available opponents for AlphaGo were humans, and five years ago the best of them were way, way stronger than any Go engine. What remains now is to make an engine a few ELO stronger. DeepMind doesn’t sell software, they do research.

Shoo.  I already covered your points many moons ago on this thread.  On the former point you are making my case for me ("anymore" is not relevant) , and on the latter you are repeating a point I made on page 1 (AlphaZero is not a consumer product and Google is not directly trying to make any money from it).