[jgtrosh]

I wish there was an equivalent of CI and peer-reviewed pull-requested contribution to their system, to ease collaborative improvement to the methods. You'd really want many sets of eyes able to look at all aspects of the drugs to vastly reduce risk.

[yosito]

I saw one of Michael's talks. They do share the data about their hardware and processes openly (though I don't know where/how, since it's not the kind of info I understand). But one of the things he mentioned needing more open information about is the chemistry. There are databases of chemical reactions that can be used to train AIs to find new, easier and cheaper ways to synthesize medications, but those databases are controlled by large for-profit corporations who refuse to release the data. Presumably for "safety" reasons, but it's not hard to imagine that "safety" isn't the whole story.

[Footkerchief]

Some of said databases: http://www.organicworldwide.net/content/reaction-databases

[femiagbabiaka]

that sounds like a magnificent startup idea that I completely lack the expertise to execute on. hopefully someone else tackles it.