[foxyv]

There are two kinds of knowledge. There is explicit knowledge which can be codified easily in markdown files or a wiki. Then there is tacit knowledge which is mostly encoded in the experience of an organization's individuals. Explicit knowledge is like the tip of a giant institutional knowledge iceberg.

[HPsquared]

Maybe they could use a distillation process. Have the AI prompt the senior engineers repeatedly (don't do this). Like squeezing the oil from olives!

[bobson381]

One of my favorite pieces of writing on this topic

https://books.worksinprogress.co/book/maintenance-of-everyth...

[foxyv]

I think that this is doable. Similar to having a new employee shadow a more experienced one and observe, you could implement a sort of program where AI shadows experienced employees and asks questions when they do something it doesn't understand.

But this is difficult to implement since AI doesn't have a body to follow someone around and it would take immense amounts of compute to do so using telemetry and cameras. You would literally be spying on employees 24x7 for weeks at a time with the express goal of replacing them someday.

[khuey]

> You would literally be spying on employees 24x7 for weeks at a time with the express goal of replacing them someday.

Isn't this almost exactly what Zuck is trying to do at Meta?

https://www.reuters.com/sustainability/boards-policy-regulat...

[foxyv]

If so, that's pretty on-brand.

[thewebguyd]

And that tacit knowledge doesn't have easily quantifiable value, it doesn't show up on the P&L so most execs don't consider it. I've seen it time and time again over my career, someone leaves or layoffs happen without considering this and then the company is scrambling to figure out processes that someone was quietly running or maintaining for years that no one else even thought of.