Earlier this year, we wrote about launching INFER’s forecasting tournament with the effective altruism (EA) community. From April to July, eleven EA student groups from universities across the globe competed to earn the top forecasting prize. The students were drawn to the tournament to learn more about forecasting as an aligned philosophy within the EA community, develop new skills, connect with other EAs, and partake in friendly competition. Since it was a tournament of firsts (many of the participants were new to forecasting, and it was the first tournament run on INFER!), we wanted to share some highlights.
In all, nearly 100 participants generated 1,100+ forecasts and 1600+ comments on 16 qualifying questions. Those questions covered a variety of topics, including microelectronics, the tech industry, cryptocurrencies, and AI policy (an EA priority area). Other tournament questions asked about the safety of autonomous vehicles, U.S. Congressional action on algorithmic decision-making, and funding for AI companies.
We want to congratulate the team who took the first place honor: EA Delhi! The student group from the University of Delhi in India said they attribute their win to their close collaboration throughout the tournament. The team frequently hosted group forecasting sessions, where they talked through their perspectives on questions, gathered relevant data, and came to a rough consensus on questions together before submitting their individual forecasts.
EA Delhi had the best performance on any single question (the Algorithmic Accountability Act of 2022), which helped them rise quickly in the rankings. Many members of the team, like Samuel Sadrach, expressed having favorite questions that they actively monitored – one example was a question on whether the French data protection authority (CNIL) would announce a fine against Clearview AI by 31 July 2022.
“I spent some time reading older reports (translated by google translate) on various companies that had been fined by CNIL to get an outside view on the timelines in which CNIL tends to fine companies,” said Sadrach.
What else did the winning team learn along the way? Team member shreya appreciated the conversation that happened both on and off INFER. For others looking to get better at forecasting, shreya suggests: “Be open to ideas of different people. We can learn a lot about a problem and ways to approach it by talking to people.”
The EA tournament prizes totalled $2,500 and were awarded to the EA groups from the University of Delhi (1st Place), Aalto University in Finland (2nd Place) and The University of St Andrews in Scotland (3rd Place). The University of Edinburgh in Scotland won Honorable Mention for their team’s activity throughout the tournament.
In addition to recognizing teams, INFER recognized several individuals for outstanding accuracy. Their performance was rewarded with an invitation to the Pro Forecaster Program, INFER’s group of paid forecasters. We want to extend a warm welcome to our new Pros: tango (Delhi), Revmag (Delhi), samuel-da-shadrach (Delhi), sanyer (Aalto), and efosong (Edinburgh)! A handful of other tournament recipients won several of INFER’s individual forecasting rewards available to the broader community.
Once the tournament ended, one thing became clear to many participants – they wanted to continue forecasting! Just like other forecasters who join INFER from different places, backgrounds, and perspectives, EA tournament participants said they enjoyed finding common ground with a community that seeks to better understand possibilities and opportunities for the future.
“I absolutely love the rush of determining what is unknown,” said Sunishka Sharma of EA Delhi. “It feels like calculated astrology!”
We want to thank all of the EA teams for their valuable contributions, especially our tournament co-sponsors, EA Cambridge and UChicago EA, who helped to organize and recruit participating teams. We look forward to continuing to partner with EA student groups to forecast for a better future.
See below for the final tournament rankings:
Are you interested in forming your own team on INFER so you can forecast together? You can create a team here.