Related: Why Academic Papers Are A Terrible Discussion Forum, Four Layers of Intellectual Conversation
During a recent discussion about (in part) academic peer review, some people defended peer review as necessary in academia, despite its flaws, for time management. Without it, they said, researchers would be overwhelmed by "cranks and incompetents and time-card-punchers" and "semi-serious people post ideas that have already been addressed or refuted in papers already". I replied that on online discussion forums, "it doesn't take a lot of effort to detect cranks and previously addressed ideas". I was prompted by Michael Arc and Stuart Armstrong to elaborate. Here's what I wrote in response:
My experience is with systems like LW. If an article is in my own specialty then I can judge it easily and make comments if it’s interesting, otherwise I look at its votes and other people’s comments to figure out whether it’s something I should pay more attention to. One advantage over peer review is that each specialist can see all the unfiltered work in their own field, and it only takes one person from all the specialists in a field to recognize that a work may be promising, then comment on it and draw others’ attentions. Another advantage is that nobody can make ill-considered comments without suffering personal consequences since everything is public. This seem like an obvious improvement over standard pre-publication peer review, for the purpose of filtering out bad work and focusing attention on promising work, and in practice works reasonably well on LW.
Apparently some people in academia have come to similar conclusions about how peer review is currently done and are trying to reform it in various ways, including switching to post-publication peer review (which seems very similar to what we do on forums like LW). However it's troubling (in a "civilizational inadequacy" sense) that academia is moving so slowly in that direction, despite the necessary enabling technology having been invented a decade or more ago.
It's probably not as simple as that. Part of why online discussions work as well as they do is probably that there's no money riding on them. If funders start making grant decisions based on mailing list discussions, we might start seeing mailing lists becoming politicized to an uncomfortable and unproductive degree. I think for now the "fix" is just for people to monitor efforts to reform peer review in academia and adopt the ones that work well into the AI safety field, and also maintain a number of AI safety research institutions with diverse cultures instead of e.g. demanding that everyone publish in academic venues as a condition for funding.