Hiya, I work at OpenAI. I think the Grover paper is a good place to read about some of this:https://arxiv.org/abs/1905.12616 We're likely publishing more on detecting fine-tuned outputs in the future, also.

Many thanks! Looking forward to reading the OpenAI research when it comes out as well.

This does not have anything to do with OpenAI (though it's a cool initiative.) Source: I work at OpenAI.

Hi there, I work at OpenAI. This isn't related to OpenAI in any official sense, though there are some people involved who either have had links to us in the past, or currently have links to us.

Correct - we saw a similar phenomenon of rapid capability gain via self-play in our Dota 2 work: https://blog.openai.com/more-on-dota-2/

(Disclosure: I write Import AI). Link: http://jack-clark.net/import-ai

Hi there, the weights and model are here: https://github.com/openai/generating-reviews-discovering-sen...

This is awesome, thanks! My apologies I must have missed it somewhere.

I write a weekly AI newsletter called Import AI which is also cross-published to this WP blog. I try to cover a mixture of fundamental research papers and applied stuff. It also includes some OpenAI updates: https://jack-clark.net/

Hiya, (I'm Jack Clark) - this is a good point. The child probably gets about 50 to several hundred distinct 'frames' of the chicken. Still, a remarkably small number of examples.

It would also be a terrible drawing. An adult who had seen many other types of birds would be able to do much better, as they would have learned more abstractions - for example the experience required to condense a complex visual pattern to, say, "mottled".

You can create environments - it's coming! We'll be releasing many components over next few months.

There was an interesting academic research paper that showed you could train in GTA and transfer over to the KITTI dataset and do ok: https://arxiv.org/abs/1610.01983