Fun fact: The consensus between professional go and chess players is that all new AI systems (alphago, etc) have really revitalised the game and introduced incredible amount of new strategies and depth.

You're getting downvotes but honestly I agree. Who cares about board games? We should've moved on from this once we "solved" chess and Go. There are more important things and it's not remotely surprising that a computer can beat a human when there's a simple, abstract optimization problem to throw computing power at. Make it creative...now that's a challenge worthy of the top AI talent.

Solving the game comes before solving for fun. If we create an AI that can win, then we can hamper the AI in fun ways, or give it an altered objective function that maximizes the players fun.

Yes indeed. AI research will only take a real step forward when it learns how to be creative instead of just very good at optimising simple formal systems like board games.

if making games is a game...

You can also just download pre-trained models. Get those set up and then install Sabaki (https://sabaki.yichuanshen.de/) and connect it to your KataGo... instant (ok, a few hours probably if it's your first time setting it up) superhuman Go AI. There's even an npm package you can use to process SGF files and automatically score moves as good/questionable/bad + generate variations that were better choices: https://github.com/9beach/analyze-sgf/blob/master/README.en-...

(Edit: Misread what the other poster was trying to do, but I'm leaving this here as a reference for anyone else who just wants to use KataGo on their own machine on their own Go games.)

Awesome! Thanks! Checking it out now.

It creates useful Go knowledge. What else could be asked of it?

Here's what Lee Sedol said when he retired:

> With the debut of AI in Go games, I've realized that I'm not at the top even if I become the number one through frantic efforts. https://en.yna.co.kr/view/AEN20191127004800315

He'd been playing Go professionally for 24 years. I never said he ragequit. He's too great a man to do something like that. Lee instead apologized for his losses, stating "I misjudged the capabilities of AlphaGo and felt powerless" while emphasizing that the defeat was his own and "not a defeat of mankind". I imagine being the Hector of humanity is quite a burden to bear. His professional ratings then took a dive for a few years https://www.goratings.org/en/history/ before he announced his retirement. To this day he remains the only human being who's ever won a single game against AlphaGo.

I don't want to pull back the curtain too much, but surely DeepMind foresaw the possibility of AlphaGo winning and then Lee Sedol losing confidence or interest in the game, which would generate a load of bad publicity for them.

So it would make sense for DeepMind's contract with him to contain a clause requiring him to continue playing go professionally for a few years (but not necessarily put much effort into it), as well as the standard non-disparagement clauses.

In fact, I wouldn't be surprised if AlphaGo was programmed to throw the forth game after securing the win with the first three of the five games. That gives Lee some bragging rights, and makes for a more hopeful story than "Computer stomps likeable human".

yeah but still because of it[1]

[1] https://www.theverge.com/2019/11/27/20985260/ai-go-alphago-l...

Knowledge isn't knowledge without an explanation: https://www.ted.com/talks/david_deutsch_a_new_way_to_explain...

Why not have a seat, take that chair over there.

much more refined to think a spam of "POG!" stands for Player of Games when reading twitch chat

No he is not, but AIs are not allowed in the competition the story centers around.

I thought the protagonist wasn't nearly as talented as the culture AIs (even the ones that are not all that powerful)?

For Adama ( http://www.adama-lang.org/ ), I am using a mutable tree with two copies and then built transactions such that I can emit deltas on the parts that change. I have all the benefits of immutability without the cost PLUS I have a cheap undo/redo log.

Definitely for his later books, but Player of Games came out three years before the first Sid Meier's Civ game.

What cool is that Paradox's Stellaris, a civ-in-space game, definitely takes pages from Ian's Culture series.

Its such a crime - of war and all else, its like a blindspot of imagination. That a man would do such a thing - to what is essentially family, as tactics.. the horror..

I'll make a minor contribution to the discussion by mentioning the Culture ship normally referred to as the Mistake Not..., which is shorthand for

"Mistake Not My Current State Of Joshing Gentle Peevishness For The Awesome And Terrible Majesty Of The Towering Seas Of Ire That Are Themselves The Milquetoast Shallows Fringing My Vast Oceans Of Wrath".

Unsure if this name is also a sarcastic stab that the lack of gravitas in ships' names, but regardless it's very sad that Banks died young :(

I think splendor gets more interesting if your opponents are also trying to be strategic. You can see what color chips they are picking up, which lets you know what they are aiming for, which influences what card you want to aim for or reserve. Mid game, you can see what colors other people are missing and try to corner colors to give you room to breath and pick up cards. You can also see the set of colors people are holding to see which of the 4 final bonus point cards are being fought over.

I like the game for what it is. I'd say, surprisingly strategic.

With a smaller cache, a less efficient branch predictor and only SSE for SIMD, I'd be curious to see the benchmark too but I'd be surprised if it was close.

I don't know if the RAM bandwidth being much lower would have an impact on CPU benchmark though.

Brilliant. "Absolutely No You-No-What" is fantastic.

"Openly available" is a strong constraint that's mentioned explicitly.

Standard RL algorithms will converge to optimal play versus a fixed opponent, but will not find an optimal policy via self play.

One intuitive way to see this is that a sequence of improving pure policies A < B < C < etc. will converge to optimal play in a perfect information game like chess, but not necessarily in an imperfect information game like rock/paper/scissors where Rock < Paper < Scissors < Rock, etc

I always interpreted the end reveal as showing that control was highly confident of both the outcome of the game and of how Gurgeh got to that outcome.

It's been a while but I am pretty sure that the ship lied when saying that it got overwhelmed and did so only because it was confident he was on the right path but needed to get there in a specific way which wouldn't have worked quite the same if the ship intervened

I think the main character can be so strong at the game by the end because of his immersion in Empire culture. The ships AI would likely be at least as strong with the same experiences. Plus, as you mention, the ships AI is not smartest AI around.

First time they mention it is page 10:

> one of the strongest and most widely-used programs is Stockfish [81].

Here's the citation, note the date:

> [81] The Stockfish Development Team. Stockfish: Open source chess engine, 2021. https://stockfishchess.org/.

They mention the version number only once, further down, and don't point out that it's out of date since February 2018. All other 11 mentions of it don't have the version number, like in that sentence:

> In Chess, PoG(60000,10) is stronger than Stockfish using 4 threads and one second of search time.

Yup, "In chess, we evaluated PoG against Stockfish 8, level 20 [81] and AlphaZero."

Not going to pick it up again.

My reaction is perhaps summarized by the old quip: "This is not a novel to be lightly tossed aside. It should be thrown with great force.".

It wasn't any one thing but I eventually hit a point where continuing just wasn't worth it for me. There's too much else on my to-read shelf which is less likely to seriously annoy me.

> read next?

Stephen R. Donaldson: He's more known for The Chronicles of Thomas Covenant, but you might prefer The Gap Cycle; it's more typical SF.

Also, the Night's Dawn trilogy by Peter F. Hamilton. Perhaps Dan Simmons? And of course anything by Charlie Stross.

Player of Games (thought that was obvious from context).

I have experimented with that, too, but in my case it also multiplies the number of potential actions. If I have 7 actions per timestep, grouping them into 3-timestep blocks means I now have 777 = 343 possibilities to choose from.

From what I understand, the OpenAI Dota 2 AI has a long-term strategy module which was mostly trained by imitating 60,000+ replays played by human professional teams. My problem with doing that for the Borderland competition is that I don't have any data source for replays of someone playing the game really well. You control 3 units simultaneously and it's 2 teams against each other, so I'd need 6 dedicated volunteers playing the game for many hours to create a reasonably-sized corpus of human replays. And who says that those people are good at it?

I'm in the process of attempting just that, with limited success. In my case, I trained a classifier that takes the current surroundings of the player unit and tries to predict that we'll gain an advantage in this segment of the game. I split the game into segments based on when the HP relationships between teams change. And gaining an advantage then means that you take more HP from the enemy team than what you and your teammates lost.

The classifier has on average 90% accuracy which seems good. I then use the likelihood predicted by this classifier to compute the weight with which I want to train each action and if I want to train it positively (by pulling its likelihood of being chosen up) or negatively (pushing the likelihood of that action down).

However, what this model cannot correctly represent is the fact that whether or not a given situation will turn out to be good or bad in the long term is highly dependent on how you play. So if I train this with replay data, I will score the situations in relation to how well those (outdated) AIs could take advantage of them.

Next up, I'll try to fix this issue by introducing a graph-like stochastic structure. The basic idea is that I encode "from this state S if I take action A, then I can reach state T with P percent likelihood" into yet another neural network. If I then identify a state which is really beneficial in the sense that I can reliably convert it into an advantage, then I can use this graph to back-propagate that knowledge so that I get "from this state S, action A takes me to state T, then action B takes me to state U, and U is great".

That should allow me to train with historical data to identify which transitions are possible, and then I can combine that with realtime data about the desirability of each state. So basically I'd do A* pathfinding over the graph of possible states to identify which actions are needed to bring me from my current situation into the closest "I will surely win" situation. Except that the graph is memorized by an AI because the real state-space is huge: 15x15 fields with 6 units + 5 environment states => roughly 11^(15*15) states

My recollection is that by the end of the novel its clear that Gurgeh was never competitive with the ship, although he might have been competitive with his security drone (although even that isn't clear, since <spoilers> imply that the security drone is a better game player than it pretends to be).

To me it felt like the whole point of the novel was that Gurgeh was a piece in an even larger game and he didn't even realize it. So the idea that the people playing the "bigger" game couldn't compete in the smaller game seems silly, and I think they mention that they used Gurgeh instead of an AI to make it appear fair to the inhabitants of the planet.

Remember the "ambassador" aka Zakalwe? They had boots on the ground and even set up interactions to prevent him getting to friendly/embedded with the "host" culture.

The whole thing was about to blow up anyway, so they brought in the Player of Games, to do it in a style that would prevent any recovery. Gurgeh was not there to defeat the empire, he was there to defeat the whole idea of the game being "holy".

He was the Jesse Owen shipped to hitlers olympics.

At least that is what the ship (SC) wants him to think.

It does, but IMO it's probably worth reading The Player of Games or Use of Weapons before it anyway. With the exception of perhaps Surface Detail, none of the Culture books rely on any others. Consider Phlebas gives a good view of The Culture from "outside" (the perspective of the Idrians) but is quite slow.

I started with Consider Phlebas because I wanted to see why people raved about the Culture series. I found it kind of tedious and slow, and although I finished it, I wondered what all the hype was about. I'm thankful that I picked up the second book though (Player of Games) - because I couldn't put it down; it was fantastic. I've stuck with the series since then (Excession was another highlight). I'd would like to revisit Consider Phlebas at some point, I think I might enjoy it more now that I have more context for the story.

I enjoyed Consider Phlebas a lot, but it is very different from the rest of the series.

I couldn't get through the cannibal part of Consider Phlebas, was soo weird.

Use of Weapons and Consider Phlebas are the worst of the series IMO. I powered through Consider Phloebus just because I knew people loved the series, but there's really no reason to start there

Consider Pheblas wasn't the best in my opinion. Player of Games is my favorite, but some of the other later ones are good.

I enjoy the settings and the concept of the Culture more than the plotlines.

Consider Phlebas is easily the worst of the series. My top 3 in no particular order are Use of Weapons, Player of Games and Excession.

If you find Consider Phlebas too slow, try Excession before giving up on Banks.

One of the reasons that Contact (and Special Circumstances) have (some) humans[] around is for intuitive leaps. I can't say I recall which Culture book this is mentioned in, but it is, in one of them.

[] let's go with "human" as a general term for Culture biological citizens, it is probably a bit incorrect, but gets the point across.

The last page of the book gives it away: (MASSIVE SPOILER OBV) The security drone who came with Gurgeh to Azad was the same drone he meets during the introduction chapters who was "rejected" from SC and offers to let him cheat (though it was wearing a disguise at the time). Then, after he cheats he basically gets blackmailed into going to Azad and conveniently this "non-SC" drone comes with him in a very "non-SC" ship that claims to have its weapons removed but doesn't. At some crucial points the security drone influences Gurgeh to play the best he can, such as when he takes him on a tour of the slums and the Culture-educated Gurgeh gets so furious at the mistreatment he witnesses that he absolutely crushes his opponent in the next match.

They mention in one of the final chapters that the minds wanted the Azad empire to become a better place since it was really shitty to its citizens. However, they couldn't just invade and impose laws because they're the Culture, and the Azad empire kept claiming moral superiority because they had this one thing (the Game) that they thought the Culture couldn't match. The Minds knew Gurgeh was talented enough to get far enough in the tournament that the Azad Empire would be seriously shaken, because if this single foreigner can beat so many of the best and brightest in the Empire at the thing it claims to do best then what could the entire Culture do? This turns out to have been correct, at the end of the book the Azad empire starts to collapse because they no longer trust their leadership, who have been proven to be incompetent at the very thing they claim to do best. Beaten by a human btw, not even by one of the god-machines that the Culture also has. Having predicted that this would happen, the Minds set out to manipulate Gurgeh into going to Azad to play the Game and by doing so bring about regime change. The Minds and/or SC were playing a much higher level game than Gurgeh all along, he was merely one of the pieces they used to play.

Yes! And also, "why is a weapon?", "how is a weapon?" and "what isn't a weapon?"

Really, Culture is a great setting for discussing a lot of interesting philosophical questions and topics.