What is Obelisk's research agenda?
Obelisk is a research lab focused on brainlike AGI safety and the development of brainlike AGI more generally. They think that there is a good chance that the first AGIs will be neuromorphic. Aligning brainlike AI comes with a different set of opportunities and challenges than other development paths, and Obelisk is studying computational neuroscience and human goal formation to prepare and explore the options.
One threat model they think about is misaligned model-based RL A machine learning method in which the machine gets rewards based on its actions, and is adjusted to be more likely to take actions that lead to high reward.
The main crux for their agenda is whether transformative AI
An AI that is capable of transforming society, as drastically as the industrial revolution or even more so.
Interested in helping out?
Making progress on Obelisk's agenda requires advanced knowledge of neuroscience and related disciplines. A strong grounding in machine learning
An approach to AI in which, instead of designing an algorithm directly, we have the system search through possible algorithms based on how well they do on some training data.