A team from the University of Glasgow have won $100,000 after placing second in an Amazon programming competition.
A team from the University of Glasgow has placed second in an Amazon programming competition, with each winning a share of the $100,000 prize. The competition saw teams from different universities build multimodal conversational agents that assist customers in completing tasks requiring multiple steps.
The reigning champions from last year’s competition, team GRILL from UofG, composed of 8 students and 1 lecturer, faced off in the final against four other teams from the University of Pittsburgh, Virginia Tech, University of California, Santa Cruz, and NOVA School of Science and Technology. The latter came away with first place, winning the $500,000 first-place jackpot.
The final stage of the competition involved responding to user feedback. Amazon users could interact with the university taskbots by saying “Alexa, let’s work together” on devices with Amazon Echo enabled. After saying the command, customers received a brief message informing them that they are interacting with an Alexa Prize university taskbot before being randomly connected to one of the participating taskbots. During this final phase, university teams implement their latest innovations and adjust their approach based on customer feedback.
Success in the challenge required teams to address difficult AI obstacles. The challenge required the fusion of multiple artificial intelligence techniques including knowledge representation and inference, commonsense and causal reasoning, and language understanding and generation.
Before the competition, first-year PhD student and team leader, Sophie Fischer, told The Glasgow Guardian: “We find it exciting to use the latest advances in Large Language Models to create experiences for users that are fun and engaging by bringing in new types of rich interactions and teaching them new knowledge and skills. The Challenge is a fun rollercoaster because the time from research to real-world deployment is shorter than in most research projects.”
