Questions
- What is the transformer of artificial life?
- Why isn’t there something that is infinitely scalable?
- Can you really use human notions of creativity and
novelty to search for cool spaces?
- Is this inherently bounded
- Is the model’s idea of what we find interesting actually the seed to new life?
- Do we need the arbitrary cutoff between agents/environment? You don’t have to evolve them separately if other agents directly lead to the problem a single agent is trying to optimize.