Tweets on NIPS 2017 Human Computer QA Competition

UMD Department of Computer Science @umdcs

On December 8th, at #NIPS2017, @boydgraber, @haldaume3, Shi Feng of #umdcs and @MohitIyyer (of AI2), He He (of @Stanford) , and Pedro Rodrigez (of @cuboulder) will host a Quiz Bowl competition between #Computer frameworks and top #human competitors: sites.google.com/view/hcqa/home…

Mohit Iyyer @MohitIyyer

Come see top human trivia players take on the winner of our quiz bowl shared task Friday evening at #NIPS2017! Panel on QA afterwards featuring @jekbradbury @VioletNPeng @RichardSocher @aggielaz and He He sites.google.com/view/hcqa/sche…

Jordan Boyd-Graber @boydgraber

The humans are better but so are the algorithms. Epic showdown! twitter.com/MohitIyyer/sta…

Jordan Boyd-Graber @boydgraber

Through today, I'm revealing the humans who will take on winning computer team at #nips2017 (5:45, 103AB): Raj Dhuwalia is a Jeopardy champion, a winner of 250k on Who Wants to be a Millionaire, and a two time top-5 at the national quizbowl ICT. norgesquizforbund.no/interview-with…

Jordan Boyd-Graber @boydgraber

Through today, I'm revealing the humans who will take on winning computer team at #nips2017 (5:45, 103AB): David Farris, mathematician, won national champions in HS, UG (Harvard), Grad (Berkeley). This year his team won the Karnataka Quiz Association championship in Bangalore.

Jordan Boyd-Graber @boydgraber

Through today, I'm revealing the humans who will take on winning computer team at #nips2017 (5:45, 103AB): Charles Meigs is the two-time All-Star in D2 quiz bowl (UCLA) and led the University of Maryland (go Terps!) to a 2008 national championship. hcqa.boydgraber.org

Jordan Boyd-Graber @boydgraber

Throughout today, I'm revealing the humans who will take on winning computer team at #nips2017 (5:45, 103AB): Sharad Vikram is a PhD candidate at UCSD studying clustering, is a veteran of Torrey Pines and Berkeley quizbowl, and created the database the computer team trained on.

Jordan Boyd-Graber @boydgraber

Throughout today, I'm revealing the humans who will take on winning computer team at #nips2017 (5:45, 103AB): James Bradbury is a Salesforce research scientist working on question answering. As a quiz bowler, he was on top 15 teams at Thomas Jefferson High School and Stanford.

Hal Daumé III @haldaume3

the competition of humans against machines begins in about 5 minutes. HISTORY WILL BE MADE!!! convention center 103 A+B #nips2017 twitter.com/mohitiyyer/sta…

Hal Daumé III @haldaume3

last year humans won about 345-150 ... can they pull it off again or are machines going for an upset???? #nips2017 twitter.com/haldaume3/stat…

Bert Huang 🐥 @berty38

The applause for the machines has died like our hope for a non dystopian future twitter.com/haldaume3/stat…

Taka Hasegawa @takahasegawa

#nips2017 Ousia and my friend @ikuyamada killing it in the Q&A competition pic.twitter.com/fQmvuBnYrZ

拡大

Bert Huang 🐥 @berty38

The #NIPS2017 audience here being silent when the AI answers correctly sounds like a collective “what have we done”

Hal Daumé III @haldaume3

@berty38 someone run up and unplug the laptop! we’ll show the computer who is smarter!!!!!

Hal Daumé III @haldaume3

final score this year: 465-200, machines win!!! twitter.com/haldaume3/stat…

Surya Bhupatiraju @suryabhupa

Just watched an AI named Ousia absolutely CRUSH a team of very qualified Quiz Bowl veterans at Quiz Bowl in the Human-Computer Question Answering competition track! #NIPS2017 pic.twitter.com/5S4zzbOrR2

拡大

Taka Hasegawa @takahasegawa

The AI that @ikuyamada and Ousia made won against human experts by a landslide! Congrats! #nips2017 pic.twitter.com/5kPBnxlr00

拡大

WorldQuizInfo @worldquizinfo

[USA] (12/8 NIPS207) 人工知能に関する国際会議です。その中で先ほどAIと人間チームによる早押しクイズ対決が行われ、日本のベンチャー起業Studio Ousiaが開発したAIが見事Raj Dhuwalia氏含むスタープレイヤーたちに圧勝する快挙を成し遂げました。 sites.google.com/view/hcqa/

Victoria X Lin @VictoriaLinML

Final question of the human-computer quizbowl competition — human expert scored 15! Missing the context from the last NIPS challenge, but the ensemble of 6 expert players did equally well as @KenJennings 2 yrs ago (youtu.be/xV1unG3tln0) #HumanComputerQAChallenge #NIPS2017 pic.twitter.com/r2tNkcnFr0

拡大

Ikuya Yamada @ikuyamada

Our quiz AI system won against a team of six human experts by a wide margin! #NIPS2017 sites.google.com/view/hcqa/sche… pic.twitter.com/c81JNHNd1Q

拡大

Ikuya Yamada @ikuyamada

Slides of my talk at #NIPS2017 HCQA competition is available online: slideshare.net/ikuyamada/stud… via @SlideShare

Jordan Boyd-Graber @boydgraber

@VictoriaLinML @KenJennings I would say that the 6-person strong human team was much stronger than @KenJennings (already a very smart guy). The **huge** difference was the algorithm. Leaps and bounds above where we were even two years ago.