Alex WellersteinVerified account

If the argument is, tacit knowledge will stop proliferation (of whatever) — very dubious. In some cases it matters a lot (but can be overcome with effort), in some cases it matters not at all (because knowledge required is "easily" acquired — e.g., in totally dual-use fields).

No, tacit knowledge isn't going to stop proliferation, but I have my doubts about how easy such knowledge is to acquire outside weapons research. Still, I agree on a general level, partic. about techs that embody previously tacit knowledge.

It depends on the field/knowledge in question. It used to be that "how to operate a nuclear reactor safely" was classified and tacit nonpro knowledge. Today you can get an undergraduate degree in that, globally.

In some cases, specific knowledge is going to be tightly held within said complexes ("how to safely handle plutonium" is not something you can easily learn outside of them — or inside of them, sometimes!). Others are not.

But the overall, cumulative effect is that tacit knowledge requirements can drop in general in a field, narrowing the kind of expertise. If it maps very closely onto a fully open field, then it ceases to be a very effective barrier to proliferation.

("narrowing the kind of expertise *required*," it should say.)

Which is just to say: the tacit knowledge "fetishists" tend to act like tacit knowledge is an effective barrier. I think history and sociology of science has shown us that, at best, it slows things down a bit. At worst, it can vanish overnight (e.g., with new instruments/tools).

(And again, to highlight what I mean by the effect of instruments/tools: I can teach undergrads with no coding experience how to make something like the NUKEMAP in about 4 mos. That is because tools+APIs+browsers+etc. have changed. Would have been impossible to do 20 years ago.)