In Bayesian statistics there’s the concept of the Uniform Prior - in the absence of evidence, assume all options are equally likely to be true. I often find myself using an alternative strategy, which I’ll dub “the masochistic prior” - in the absence of evidence, whichever option is most painful...
When people talk about AI futures, the picture is usually centralized. Either a single aligned superintelligence replaces society with something utopian and post-scarcity, or an unaligned one destroys us, or maybe a malicious human actor uses a powerful system to cause world-ending harm. Those futures might be possible. However there’s...
Summary: The collection of a massive first-person video dataset for use in large model pretraining is technologically & economically feasible with modest resources. Such a dataset might enable high-fidelity imitation of human behavior, with positive implications for the capabilities and safety of trained models. Such models may also enable a...