- I have awarded $15,000 of prizes for the second round of GoodAI’s General AI Challenge, Solving the AI Race.
- The six top solutions out of 59 were awarded prize money.
- This was just the first stage where the goal was to get as diverse pool of ideas as possible.
- Prize winners have been invited to the Human-Level AI Conference in Prague this August to discuss their submissions.
- We intend to continue working on solutions to the AI Race. In the future we will focus on concrete KPI’s and testing for robustness of solutions.
I am very excited to announce the prize winners of the second round of the General AI Challenge, Solving the AI Race. The round was the first qualitative round of the challenge and followed on from the Gradual Learning round last year. In January we challenged the public to come up with solutions to the pitfalls of an AI Race, where:
- Key stakeholders, including the developers, may ignore or underestimate safety procedures, or agreements, in favor of faster utilization
- The fruits of the technology won’t be shared by the majority of people to benefit humanity, but only by a selected few
I was amazed with the response, in three months 194 people registered (from 41 different countries) and we had 59 submissions. The team at GoodAI read all of the submissions and created an anonymized shortlist for the judging panel, which I was part of. On the panel I was joined by:
- Virginia Dignum, Associate professor at TU Delft
- Danit Gal, Project Assistant Professor, Cyber Civilizations Research Center, Keio University
- Rodolfo Rosini, Partner at Zeroth.ai
- Roman V. Yampolskiy, Associate Professor at University of Louisville
It was not an easy task to decide how the prize money of $15,000 would be split. We judged the solutions on five criteria:
- Impact: the potential the solution shows to maximize the chances of a positive future for humanity
- Feasibility: how practical it will be to implement / apply
- Acceptance: how likely it is that actors involved will accept the idea. E.g. in case of an actionable strategy, what is the chance actors would publicly pledge to it? In case of a framework, how easily could it be adopted?
- Integrity: how ethical the solution is (ideally solutions should not disadvantage any actors and take into account diversity of values)
- Novelty of ideas: has it been suggested before?
All of the judges then responded independently (as to not influence each others judgement) and we then came together to count our scores and determine the prize winners. We split the prizes between the three “top scoring solutions” and three “runners-up”, you can see the winners and read their solutions below:
Top scoring solutions ($3,000 each)
- Kesavan Athimoolam, Solving the Artificial Intelligence Race: Mitigating the problems associated with the AI Race
- Alexey Turchin and & David Denkenberger, Classification of Global Solutions for the AI Safety Problem
- Ehrik L. Aldana, A Theory of International AI Coordination: Strategic implications of perceived benefits, harms,capacities, and distribution in AI development
Runners-up ($2,000 each)
- David Klimek, Framework for managing risks related to emergence of AI/AGI
- Gordon Worley, Avoiding AGI Races Through Self-Regulation
- Morris Stuttard & Anastasia Slabukho, The AI Engineers’ Guild: proposal for an AI risk mitigation strategy
This was just the first stage where the goal was to get as diverse pool of ideas as possible and we intend to continue working on solutions to the AI Race. We are glad we got such a diverse range of ideas and opened up many points of discussion. The judges concluded that none of the submissions provided a complete robust solution to all of the possible pitfalls of a race to AI. Therefore, the next round of the Challenge is likely to focus more on concrete KPI’s and test for robustness of solutions.
The prize winners have been invited to the Human-Level AI Conference in Prague this August, which is being co-organized by GoodAI, where they will be able to discuss their ideas further in an AI Race and Societal Impacts panel discussion on Friday 24 August. Everyone interested in the General AI Challenge can also benefit from a 10% discount to the Human-Level AI Conference using the discount code: Challenge-HLAI18.
As the submissions really got the judges thinking, they have also been invited, by jury member Roman V. Yampolskiy, to submit their papers to a Special Issue of peer reviewed journal Big Data and Cognitive Computing journal — “Artificial Superintelligence: Coordination & Strategy”.
A big thank you to everyone who took part in the Challenge and to all of the judges, it has given us some serious food-for-thought. We will continue to work on creating a robust solution and urge the community to take part as well! If you would like to collaborate in any way please get in contact with us.
Thanks for reading,
For more news:
Space Engineers: www.SpaceEngineersGame.com
Medieval Engineers: www.MedievalEngineers.com
General AI Challenge: www.General-AI-Challenge.org
AI Roadmap Institute: www.RoadmapInstitute.org
Keen Software House: www.keenswh.com
Marek Rosa is the CEO and CTO of GoodAI, a general artificial intelligence R&D company, and the CEO and founder of Keen Software House, an independent game development studio best known for their best-seller Space Engineers (2.5mil+ copies sold). Both companies are based in Prague, Czech Republic.
Marek has been interested in artificial intelligence since childhood. Marek started his career as a programmer but later transitioned to a leadership role. After the success of the Keen Software House titles, Marek was able to personally fund GoodAI, his new general AI research company building human-level artificial intelligence.
GoodAI started in January 2014 and has grown to an international team of 20 researchers.
At this time, Marek is developing both Space Engineers and Medieval Engineers as well as daily research and development on recursive self-improvement based general AI architecture.