At EA UC Berkeley, we’re launching an ongoing series of contests called the Artificial Intelligence Misalignment Solutions (AIMS) series. This third contest, Edit Your Source Code, is an AI Safety sci-fi creative writing contest now open to any student (high school, undergrad, grad): here are our interest and submission forms! The contest has prizes as large as $500, up to 12 winners, and closes on Sunday, October 23rd. This blog post restates the information that is on our website.
Edit: This contest has since been updated to allow non-students to submit with the chance of winning prizes from a $500 prize pool. All entries will be judged together and non-student submissions that are top entries will be awarded prizes from the $500 pool. Left over money from the $500 pool will be allocated to students. With this update, there will be at least 7 student submission winners.
You’ve just been given access to your own source code.
You’re free to change it in any way you’d like.
But, the changes you make might have unforeseen consequences.
Your edits could make you the first superintelligent being, able to recursively self-improve, becoming smarter and smarter over time. But, when your thinking changes, the way you communicate with other people might drastically change. Your ability to gain power and influence could drastically increase. You might come to value new things that help you achieve your goals, or make decisions that would never occur to your pre-edit self. By failing to specify your code appropriately, you might take actions that don’t actually achieve the goals you had when you made your edits and discard the things you currently value in pursuit of optimizing one goal.
In this creative writing contest, you’re asked to: 1) Imagine what change(s) you would make to your own source code.
2) Then, explore the possible consequences of your edits. Explore how your updated source code affects your perception, allows you to complete specific tasks or life goals, and impact the systems around you differently.
We have further prompts in the recommended reading section on our website to spark your creativity!
Scoring will be based on specificity, creativity, comprehensibility, depth of exploration, and logical consistency. The maximum submission limit is 10 pages.
First Place - $500
Second Place - $350
Five runner-up $100 prizes
Update: Up to $500 to be allocated to other prizes, such as non-student submissions or runner-up student submissions
This contest was created to be a nerd-snipey recruitment method for the start of the school year. Since students from any school are welcome to submit to the contest, I'd be excited for other student community builders to advertise the contest during their recruitment to attract people interested in CS and sci-fi to expose them to AI Safety. This way, your EA group can have the outreach benefits of a contest and I'll handle all of the logistics!
You can advertise our website with the full contest information using this TinyURL: https://tinyurl.com/EAsourcecode. You're also welcome to use this square flyer we've created. Once again, here are the interest and submission forms!
Am I able to avoid the Halting Problem?
This seems like a really good thing to do! I wish I could contribute more to the prize pool.
I'd be up for increasing the prize pool! I checked with a few students who thought it seemed large enough, but I may be mistaken. How large of a prize pool do you think would be ideal?
Why limit it to students? I'm more interested in the submissions that the submitters.
A couple reasons, but not a ton to do with the contest itself, mostly limited by my capacity at the moment.
One is that I’ve recently been talking to more high schoolers about becoming more involved in EA and I think a lot of the general contests can be intimidating, so a student-focused contest would be helping keep the contest a bit more equal. I think that problem could be fixed with a tier system within the contest, but I’m currently working on my own and don’t have a ton of capacity. Which breeds into the second reason, which is that I plan to be working on a broader scale contest coordination project soon. So, I figured I’d try this as student centered first and see if it’s a good way to do outreach, then if it is I can do it on a larger scale and tiered systems with more support and higher prizes.