A new tournament to be held in Fuzhou, China.

Official Chinese name: “吴清源杯”世界女子围棋赛
English name: Go Seigen Cup World Women's Go Championship (this is the literal translation, official name unknown so far)

Dates: 26th April - 2nd May 2018

Format: 28 players (7 Chinese mainland, 4 Korean, 4 Japanese, 2 Chinese Taipei, 8 European and North American, and 3 wildcards)

Players known so far:

Korea: Choi Jeong, Oh Yujin, Oh Jeonga (appointed), and Kim Cheayoung (winner of a preliminary tournament)

Japan: Fujisawa Rina, Xie Yimin, Nyu Eiko, Ueno Asami (top 4 Japanese according to their 2017 prize money)
Chinese Taipei: Hei Jiajia (top ranked), Yang Zixuan (winner of a preliminary tournament)

Prize:

prize pool: 1,200,000 CNY. Winner prize: 500,000 CNY; runner-up prize: 200,000 CNY

side event: an international AI's tournament

confirmed AI players: Tianrang, CGI, Deepzengo, Dolbaram

Also there is a game between Ke Jie and a mysterious AI program

My personal comments:

* This is what a proper 'world championship' looks like.
* Prize is higher than any current international women's tournament
* The Chinese appears to be very serious promoting Go by giving eight seats to western players, although I guess there may be some preliminary rounds among these players, otherwise why 28 players in total? We will see.

It looks like a nice event, although I am a bit surprised to see 8 'Western' entries there. I'm curious as to how they will go from 28 to 2!? You'd imagine a 12 + 4 from 8:8

There are two basic formats for a World Championship.
1 The best player from every country is invited to compete on an equal footing.
2 The best players in the world compete
I think that #1 is nice for Amateur Competition, and that #2 is best for professional. I'm not a super fan of the mixture, but look forward to it all the same.

_________________
North Lecale

I agree that this is a much more satisfactory tournament for the women of the world (the Chinese do get more places than anyone else, but I approve of that).

But are you sure about the prize? The Nihon Ki-in list the Senko winner's prize (優勝賞金) as 10 million yen, which I think is about 20% higher than the GSG Cup. Of course the total fund for the GSG must be considerably higher, and maybe that's what you meant.

I think it is also worth recording that the Chinese have been extraordinarily generous in providing games for the western professionals in general. With that in mind, I'd be a little disappointed if the GSG places for European and North American players went to Chinese players simply resident there.

We have a story on the BBC today, as part of a much wider debate about equal pay for women, in which John McEnroe is said to be paid ten times more than Martina Navratilova as a TV commentator (the BBC already has omelette on face about their women's pay). Even the Senko Cup comes in at under a third of, say, the Meijin (technically open but for men in practice). However, the Meijin prize was drastically reduced a couple of years ago. I assumed that that economic stringency, but now I wonder whether it may also be a way of releasing more funds to boost the women's prizes.

Yes you are right about the winner's prize: 10,000,000 JPY is about 20% more than 500,000 CNY. The original source is not very clear about this point. It might refer to the total pool of money or maybe the budget of the tournament. The winner's prize only account for 5/12 of the total. As mentioned in a recent post, such a balanced arrangement is not unusual in the Chinese system.

I do feel bad for any professionals without places, and maybe I'm selfish here, but I can't complain about 8 EU/NA places if it's true (although I suspect something closer to the Samsung Cup).

_________________
On Go proverbs:
"A fine Gotation is a diamond in the hand of a dan of wit and a pebble in the hand of a kyu" —Joseph Raux misquoted.

This event happens within 2 km from my home ... It's a pity that I'm not there at the time.

You live in Fujian/Fuzhou? I was working there from 2009 to 2015

_________________
I am the author of GoReviewPartner, a small software aimed at assisting reviewing a game of Go. Give it a try!

I grew up there, and now I’m working in some other place, but my home is still there

Nice!

As for me, this is place I learned Go, I used to attend lessons in a Go school (normally for children, but we were a very small group of adults taking lessons as well). Well, I did not improved that much in Go, but I certainly improved in Chinese while attending the lessons

Also, there is that museum located at that house were Go Seigen that opened in 2012 (WuShanKou, at south of NanHouJie) and that Starbuck with decoration related to Go just next to that museum, I played a bunch of games there, that was good time.

I could not find such a Go friendly environment after I moved out of FuZhou, but there is a possibility that I move back there again for work in the future.

_________________
I am the author of GoReviewPartner, a small software aimed at assisting reviewing a game of Go. Give it a try!

I am really curious about that one. Could it be LeelaZero

_________________
I am the author of GoReviewPartner, a small software aimed at assisting reviewing a game of Go. Give it a try!

No, we'll soon learn the retirement of a strong AI program and the release of its next generation, probably with a different name.

There's a press release. This mysterious AI program is said to have its own creative algorithm. No detail is given at this time, however the program "does not use the core deep learning techniques as published in the 2nd AlphaGo paper".

The new AI program has been named as 星阵围棋, ('星阵' literally meaning 'an array of stars'). But there is a formal English name called Golaxy. It is related to the AI program Abacus, but new (none machine learning) techniques are said to be used.

Perhaps djhbrown's SWIM finally got implemented?!

The "new (none machine learning)" part sounds intriguing and contradictory, I wonder if there is a typo or some inaccurate translation somewhere?

If it is using some new AI-like techniques, it would still fall in the "machine learning" category. Maybe: "new machine learning techniques"?

If it is using some old, rule-based techniques - then I don't think "new" is accurate...

_________________
Sorin - 361points.com

Maybe Go Seigen's brain in a vat? Or is that actually DeepMind's secret?

Could the new techniques be to do with hardware, not software?

Having read several reports I still don't know what exactly the new techniques are. But one thing I learn is that the developers are exploring ways to use only a small amount of computing resources (relative to AlphaGo) to achieve high level playing. So this does not sound like a hardware solution.

The game 1st golaxy vs Leela Zero game will start in less than an hour, with English commentary by Zhao Baolong 2p here: http://portal.yikeweiqi.com/online/goli ... il?id=8326. From https://www.reddit.com/r/cbaduk/comment ... be_played/ Golaxy will use 10x1080Ti GPUs and LZ 4*Titan V (apparently these are about the same power). Also Golaxy has 19 blocks, so is neural-network based, just probably not a Zero-style.