Open Contests
There are currently no open contests, but you can keep an eye out for new ones coming soon!
Past Contests
-
Distillation Contest
Distillation is the practice of “taking a complex subject and making it easier to understand.” Distillation is more involved than summarizing; a distillation also communicates new examples and applications of the original research.
-
Mini-ELK Contest
An advanced AI knows something that you don’t. How can you ensure that the AI accurately reports that knowledge?
-
Edit Your Source Code Contest
You’ve just been given access to change your own source code. You’re free to change it in any way you’d like. You can set up new functions to make more accurate predictions about the world. You can increase your processing speed. You can manipulate your own reward signals. But, the changes you make might have unforeseen consequences.
-
Global Warming Contest
People frequently express concern about climate change as one of the largest existential risks and potential sources of suffering for humans and other animals in the coming decades.
In addition to the threats posed by climate change, there are other neglected large-scale problems. A few issues Effective Altruism focuses on include present-day suffering (global suffering and inequality, animal cruelty) and new risks from rapid technological development (manufactured viruses and bioweapons, capabilities of artificial intelligence).
Yet, the relationships between these existing and emerging problems are underexamined.
The Global Warning Contest hopes to create more awareness for all of the problems listed above and reward students for thinking critically about the future.
AIMS Contest Series
Artificial Intelligence Misalignment Solutions.
The Artificial Intelligence Misalignment Solutions Contest series was created in order to provide UC Berkeley students with a tangible opportunity to engage with the Alignment Problem. Working on the Alignment Problem generally means working to ensure that AI produces outcomes that reflect the coder’s intentions and values. When the AI does not reflect those values, we consider the AI to be “misaligned.” Alignment is considered a challenge because it is
for coders to create rules and control mechanisms that lead AI to do exactly what we intend for it to do, especially when the AI is advanced.
Why is it important to find solutions for misaligned AI?
“In the near term, the goal of keeping AI’s impact on society beneficial motivates research in many areas, from economics and law to technical topics such as verification, validity, security, and control. Whereas it may be little more than a minor nuisance if your laptop crashes or gets hacked, it becomes all the more important that an AI system does what you want it to do if it controls your car, your airplane, your pacemaker, your automated trading system or your power grid. Another short-term challenge is preventing a devastating arms race in lethal autonomous weapons.
In the long term, an important question is what will happen if the quest for strong AI succeeds and an AI system becomes better than humans at all cognitive tasks. As pointed out by I.J. Good in 1965, designing smarter AI systems is itself a cognitive task. Such a system could potentially undergo recursive self-improvement, triggering an intelligence explosion leaving human intellect far behind. By inventing revolutionary new technologies, such a superintelligence might help us eradicate war, disease, and poverty, and so the creation of strong AI might be the biggest event in human history. Some experts have expressed concern, though, that it might also be the last, unless we learn to align the goals of the AI with ours before it becomes superintelligent.”