builds on only three components: 1) prompt to design searches, 2) search, 3) use the data for grounding and forecast
the name is due to they wanted to name it “Nate Gold” as “Nate Silver ++” but they didn’t get permission so instead they 538++
“what’s the probability for a budding photographer to make a living in california as a stock photographer starting this year, and has to survive at least 10 years”
It said 30%
now, after you’ve played with it a bit, a question for a data scientist: how would you make a better design for similar tool?
better as in more accurate, as well as able to answer quite precisely for wider array of questions