Yup, twitter monitoring is a thing that I have seen implemented. We did not allow it to page us, however. As you say, some of the barriers around that are low o

Yup, twitter monitoring is a thing that I have seen implemented. We did not allow it to page us, however. As you say, some of the barriers around that are low or gone as of late. I wonder if someone has already secured seed funding for social media monitoring as a service. The feature set you can build on a LLM is orders of magnitude better than what was practical before. Looking at my post up-thread, I wish I had emphasized the time aspect more - of course all of these problems are solvable but it takes both time and money. They have the money now but two months ago the parts of this incident were in place but the scale was so small that it never actually leaked data. Or maybe a handful of early adopters saw some weird shit but we’re all well-trained to just hit refresh these days. Hiring even one operator and getting them spun up takes calendar time that simply has not existed yet. I assume someone over there is panicking about this and trying to get someone hired to make sure they look better prepared next time, because there will be a next time, and if they’re even half as successful as the early hype leads me to believe, I expect they are going to have a lot more incidents as they scale. One in a million is eight and a half times per day at 100 rps.