Author Topic: 1st Chess , then Go , now US Air Force , when D&D ?  (Read 13192 times)

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #40 on: December 07, 2017, 05:06:22 PM »
here We GO


A generalized version AlphaZero just got put out.
It learns on it's own, without previous human sources.
It beat the best a.i. versions of Shogi in 2 hours , Chess in 4 hours , and "AlphaGo Zero" in 8.

Fancy paper ---> https://arxiv.org/abs/1712.01815
News writeup A ---> https://venturebeat.com/2017/12/06/google-deepmind-repurposed-its-alphago-ai-to-beat-the-best-chess-and-shogi-bots/
News writeup B ---> https://www.theverge.com/2017/12/6/16741106/deepmind-ai-chess-alphazero-shogi-go

I don't pretend to understand ... but ... anything "quantize-able" (to paraphrase PLZ) about d&d, it'll get'r done.  Maybe soon too.
Your codpiece is a mimic.

Offline Archon

  • Sr. Member
  • ***
  • Posts: 334
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #41 on: December 08, 2017, 09:09:33 AM »
Interesting. I wonder how much these systems invent their own internal meta-game, and if that effects how good they are at dealing with some more human-intuitive scenarios? Independent of the different strengths and weaknesses of the human and computer player.

The problem with getting a AI to play D&D is that D&D is fundamentally poorly defined. Both in rules (which often require rulings or disambiguations) and in objectives (which are kinda ??? in computation terms). Once computers have empathy and natural language and speech ect. down (i.e. notime soon), they will be able to take a crack at D&D.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #42 on: December 08, 2017, 04:50:13 PM »
Stuff like i.p.proofing and anything simple math like a +1 blanket ability vs a +2 small use ability , is definitely fair game for a.i.

But yeah, so much of the game is semantics, english language logic + usage quirks, leadership, divinations, etc.


Chess based sites now have up longer write-ups about the AlphaZero games.
Tossing so much 1000+ year tradition aside, for tactics unimagined until these games happened.
Your codpiece is a mimic.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #43 on: January 06, 2019, 03:37:56 PM »
Star Talk with Neil DeGrasse Tyson, had a good one.

His astro co-host Charles Liu said (~paraphrasing) :   "I do open book exams, I do open internet exams.  We have to be smarter than a google search."

I found this tweet :
Charles Liu @chuckliu 26 Nov 2018
Are my exams really open internet? Ask my students... https://www.nationalgeographic.com/tv/watch/57842686b7482a6c6684e8d534b838cb/ …
links to season 5 episode 3.


I'll pat myself on the back then ... because I'm at least as smart as his freshmen (the googling part).
 :D
Your codpiece is a mimic.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #44 on: July 19, 2019, 05:05:29 PM »
Elon Musk's white paper about Neuro-link to the brain.

Pretty picture on page 7 is a rat/mouse with the 3072 channels connected.
https://assets.documentcloud.org/documents/6204648/Neuralink-White-Paper.pdf
Your codpiece is a mimic.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Your codpiece is a mimic.

Offline altpersona

  • Legendary Member
  • ****
  • Posts: 2000
  • #78
    • View Profile
    • You are here
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #46 on: September 22, 2019, 03:16:42 PM »
already surpassed w. 72?
The goal of power is power. - 1984
We are not descended from fearful men. - Murrow
The Final Countdown is now stuck in your head.

Anim-manga still sux.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #47 on: September 24, 2019, 04:58:50 PM »
Maybe, sort of, not yet, idk.

“To our knowledge, this experiment marks the first computation that can only be performed on a quantum processor,” the Google researchers wrote.
Bolding is mine.  Curious milestone. 

This site claims to have snagged a copy of the original paper :
http://www.spaceref.com/news/viewsr.html?pid=52862
... very dense with jargon and detail.

Subtleties are mostly lost on me (at this point),
but Math had a proof a few years back, that couldn't
be hand checked rather had to be computer processed.
I presume this is on the same conceptual trajectory.
Your codpiece is a mimic.

Offline altpersona

  • Legendary Member
  • ****
  • Posts: 2000
  • #78
    • View Profile
    • You are here
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #48 on: September 24, 2019, 05:13:28 PM »
yeah, the whole thing is up in the air
The goal of power is power. - 1984
We are not descended from fearful men. - Murrow
The Final Countdown is now stuck in your head.

Anim-manga still sux.

Offline Keldar

  • Legendary Member
  • ****
  • Posts: 1032
  • What's this button do?
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #49 on: February 29, 2020, 04:34:50 AM »
When will then be now?  Soon. 
(soft paywall)

Ok, not soon.  But they are trying to make a computer play D&D.  Hopefully they don't teach it to play CN.

Offline Nanashi

  • Sr. Member
  • ***
  • Posts: 358
  • it means "he who has no name" in a foreign tongue.
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #50 on: February 29, 2020, 11:46:55 AM »
The fundemental problem with teaching AI to play dnd is that, since there is a strong luck based element, it's not solvable like other games.

Offline awaken_D_M_golem

  • Epic Member
  • ****
  • Posts: 7639
  • classique style , invisible tail
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #51 on: February 29, 2020, 02:54:46 PM »
Your codpiece is a mimic.

Offline Keldar

  • Legendary Member
  • ****
  • Posts: 1032
  • What's this button do?
    • View Profile
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #52 on: March 01, 2020, 02:56:10 AM »
The fundemental problem with teaching AI to play dnd is that, since there is a strong luck based element, it's not solvable like other games.
Which is a good reason to work on it.

Offline oslecamo

  • DnD Handbook Writer
  • ****
  • Posts: 10080
  • Creating monsters for my Realm of Darkness
    • View Profile
    • Oslecamo's Custom Library (my homebrew)
Re: 1st Chess , then Go , now US Air Force , when D&D ?
« Reply #53 on: March 01, 2020, 05:14:24 AM »
The fundemental problem with teaching AI to play dnd is that, since there is a strong luck based element, it's not solvable like other games.

Although luck element is a problem, it's hardly the most fundamental one with teaching AI to play D&D. First you need to deal with:
1-D&D rules are full of holes. Our human minds will just unconsciously fill them in, but an AI would go "DOES NOT COMPUTER ERROR BSOD" at the drop of a hat.
2-If you somehow overcome the above, now you need to deal with all the infinite loops and borked combos. The AI will answer every D&D problem with "I use pun-pun lol" assuming they don't find something even more OP.