AlphaGo vs Lee Se-Dol pobeda

[Demian]

Prvo sam postovao u Roboti ali mislim da zasluzuje svoj topic. I definitivno nam trebaju Nauka:Vesti

Nije bas robot, ali je znacajan korak za nauku - ne znam gde da smestim. Potreban nam je topic za novosti

Naime juce je Googlov AlphaGo pobedio legendarnu sampionku Lee Se-Dol u korejskoj igri "Go". Danas je zablezio jos jednu pobedu i trenutni skor je 2-0 za Ai razvijen i naucen putem neuralnih mreza. Mozda ne zvuci kao nesto posebno, obzirom da odavno imamo programe koji su pobedili velemajstore saha, -  ali jeste. Sahovski programi se oslanjanju na "brute computing force", milione memorisanih partija za trazenje odgovorajuceg poteza, gde covek jednostavno ne moze da parira .... to se jednostavno ne moze primeniti na Go.

Igrace se jos 3 meca, ali cak i da Lee pobedi to ne umanjuje ovaj neverovatan milestone u istoriji vestacke inteligencije. Prvo, niko nije ocekivao da ce AlphaGo pobediti. Drugo, cak i da se pripise sreci/ljudskoj gresci prva pobeda, potezi koje je mreza napravila u drugoj partiji su sve osim ocekivani. 

  Quote

"Yesterday I was surprised but today it's more than that — I am speechless," said Lee in the post-game press conference. "I admit that it was a very clear loss on my part. From the very beginning of the game I did not feel like there was a point that I was leading." 

  Quote

"We're very pleased that AlphaGo played some quite surprising and beautiful moves, according to the commentators, which was amazing to see." 

 

  Quote

The close nature of the game appears to offer validation of AlphaGo's evaluative ability, the main roadblock to proficiency for previous Go programs. Hassabis says that AlphaGo was confident in victory from the midway point of the game, even though the professional commentators couldn't tell which player was ahead. 

  Quote

But AlphaGo could also open up new avenues for the game. Members of the Go community are as stunned with the inventive, aggressive way AlphaGo won as the fact that it did at all. "There were some moves at the beginning — what would you say about those three moves on the right on the fifth line?" American Go Association president Andy Okun asked VP of operations Andrew Jackson, who also happens to be a Google software engineer, at the venue following the match. "As it pushes from behind?" Jackson replied. "If I made those same moves…" Okun continued. "Our teachers would slap our wrists," Jackson agreed. "They’d smack me!" says Okun. "You don’t push from behind on the fifth line!"

"We’re absolutely in shock," said Jackson. "There’s a real question, though. We’ve got this established Go orthodoxy, so what’s this going to reveal to us next? Is it going to shake things up? Are we going to find these things that we thought were true — these things you think you know and they just ain’t so?"

 

Potezi koje je program odigrao nikada nisu do sada bili koristeni. Ako nekoga interesuje sledeci mecevi prenose se live na:

https://www.youtube.com/channel/UCP7jMXSY2xbc3KCAE0MHQ-A

Detaljnije objasnjeno:

http://www.theverge.com/2016/3/9/11185030/google-deepmind-alphago-go-artificial-intelligence-impact

 

Reddit megathread > https://www.reddit.com/r/baduk/comments/49g1bv/alphago_and_lee_sedol_mega_thread/

Edited March 10, 2016 by Demian

[Lucky]

Pričao sam sa nekim likom koji cepa Go i koji mi je objasnio da kao postoje još više "legendary" igrači od ove i kao njih nema šanse da dobije! :) Sad otkud znam, ova deluje kao dosta ozbiljno a i da nije, svakako je sjajan poduhvat.

[Demian]

Anywayz 3-0 je za Ai, challenge je pripao masini ... odigrace se i naredne dve, sporta radi.

http://www.theverge.com/2016/3/12/11210650/alphago-deepmind-go-match-3-result

[nikola kralj]

Prvo sam postovao u Roboti ali mislim da zasluzuje svoj topic. I definitivno nam trebaju Nauka:Vesti

Nije bas robot, ali je znacajan korak za nauku - ne znam gde da smestim. Potreban nam je topic za novosti

Naime juce je Googlov AlphaGo pobedio legendarnu sampionku Lee Se-Dol u korejskoj igri "Go". Danas je zablezio jos jednu pobedu i trenutni skor je 2-0 za Ai razvijen i naucen putem neuralnih mreza. Mozda ne zvuci kao nesto posebno, obzirom da odavno imamo programe koji su pobedili velemajstore saha, -  ali jeste. Sahovski programi se oslanjanju na "brute computing force", milione memorisanih partija za trazenje odgovorajuceg poteza, gde covek jednostavno ne moze da parira .... to se jednostavno ne moze primeniti na Go.

Igrace se jos 3 meca, ali cak i da Lee pobedi to ne umanjuje ovaj neverovatan milestone u istoriji vestacke inteligencije. Prvo, niko nije ocekivao da ce AlphaGo pobediti. Drugo, cak i da se pripise sreci/ljudskoj gresci prva pobeda, potezi koje je mreza napravila u drugoj partiji su sve osim ocekivani. 

"Yesterday I was surprised but today it's more than that — I am speechless," said Lee in the post-game press conference. "I admit that it was a very clear loss on my part. From the very beginning of the game I did not feel like there was a point that I was leading." 

"We're very pleased that AlphaGo played some quite surprising and beautiful moves, according to the commentators, which was amazing to see." 

 

The close nature of the game appears to offer validation of AlphaGo's evaluative ability, the main roadblock to proficiency for previous Go programs. Hassabis says that AlphaGo was confident in victory from the midway point of the game, even though the professional commentators couldn't tell which player was ahead. 

But AlphaGo could also open up new avenues for the game. Members of the Go community are as stunned with the inventive, aggressive way AlphaGo won as the fact that it did at all. "There were some moves at the beginning — what would you say about those three moves on the right on the fifth line?" American Go Association president Andy Okun asked VP of operations Andrew Jackson, who also happens to be a Google software engineer, at the venue following the match. "As it pushes from behind?" Jackson replied. "If I made those same moves…" Okun continued. "Our teachers would slap our wrists," Jackson agreed. "They’d smack me!" says Okun. "You don’t push from behind on the fifth line!"

"We’re absolutely in shock," said Jackson. "There’s a real question, though. We’ve got this established Go orthodoxy, so what’s this going to reveal to us next? Is it going to shake things up? Are we going to find these things that we thought were true — these things you think you know and they just ain’t so?"

 

Potezi koje je program odigrao nikada nisu do sada bili koristeni. Ako nekoga interesuje sledeci mecevi prenose se live na:

https://www.youtube.com/channel/UCP7jMXSY2xbc3KCAE0MHQ-A

Detaljnije objasnjeno:

http://www.theverge.com/2016/3/9/11185030/google-deepmind-alphago-go-artificial-intelligence-impact

 

Reddit megathread > https://www.reddit.com/r/baduk/comments/49g1bv/alphago_and_lee_sedol_mega_thread/

 

Pričao sam sa nekim likom koji cepa Go i koji mi je objasnio da kao postoje još više "legendary" igrači od ove i kao njih nema šanse da dobije! :) Sad otkud znam, ova deluje kao dosta ozbiljno a i da nije, svakako je sjajan poduhvat.

https://en.wikipedia.org/wiki/Lee_Sedol

Ovaj Lee Sedol mi opasno lici na lika a ne na ribu...ali u sv sl nadam se da alpha nece u skorije vreme da evoluira u neki skajnet i da, nakon sto zagospodari planetom, pocne sve da nas tera da igramo go.

 

 

 

 

Edited March 13, 2016 by nikola kralj

[re1]

Lee Sedol i jeste lik, a ne riba :)

Trenutno najbolji igrac, bar na rankinzima, je Ke Jie, koji je izjavio ovo: http://www.shanghaidaily.com/national/AlphaGo-cant-beat-me-says-Chinese-Go-grandmaster-Ke-Jie/shdaily.shtml

Tako da kontam da ce sledeci match da bude Ke Jie vs AlphaGo-a.

Takodje, AI trenutno igra poteze koji ce mu uvek dati samo marginalnu prednost nad protivnikom do kraja partije, tako da jos niko ne zna koliko je u stvari jak.

Edited March 13, 2016 by re1

[peja_]

igra poteze koji mu daju najvise sanse da pobedi, bilo to sa 0.5 poena razlike ili 50 poena razlike, a ne ovo gore

[bolkonski]

A i igra je poreklom iz Kine!!!

[papara]

 

Ovaj Lee Sedol mi opasno lici na lika a ne na ribu...ali u sv sl nadam se da alpha nece u skorije vreme da evoluira u neki skajnet i da, nakon sto zagospodari planetom, pocne sve da nas tera da igramo go.

daleko je to od skajneta. ovo je brute force search tree koji feeduje poteze u dobro obucenu neuralnu mrezu koja ocigledno sa velikom pouzdanoscu ocenjuje pozicije u korist jednog ili drugog. ali ce i olupati ovog sto tvrdi da je nepobediv. inace, googlov team je rekao da im je sledeci zadatak ai za starkraft 2 :)

[re1]

Lee Sedol beats AlphaGo in Game 4

Edited March 13, 2016 by re1