The trade you describe at the end, a human catching what the models misread, assumes the misreading is idiosyncratic. My worry runs the other way: when the first pass is produced the same way everywhere, interpretations stop disagreeing with each other. I see a version of this already in sell-side target revisions, which cluster more tightly than they used to without carrying more information about what the stock does next. Homogeneous reading shows up as names inside a theme moving as one block, and the dispersion that used to mark genuine disagreement quietly thins out. Rarer opportunities, larger when they come, which is an awkward payoff to size.
Nowadays we don’t remember the phone numbers but have used that headspace for something new and hopefully more productive .
In the 60s they used paper spreadsheets. It would take rooms filled with paper, many men and hundreds of hours to come up with a number. If they have to change one number, it would take them another few hours and more paper.
Now I could run 100,000 simulations on 5 variables with a click of a button and get way more insight that those men in the 60s.
hi, can the webinar recording be made available, please. thanks
The trade you describe at the end, a human catching what the models misread, assumes the misreading is idiosyncratic. My worry runs the other way: when the first pass is produced the same way everywhere, interpretations stop disagreeing with each other. I see a version of this already in sell-side target revisions, which cluster more tightly than they used to without carrying more information about what the stock does next. Homogeneous reading shows up as names inside a theme moving as one block, and the dispersion that used to mark genuine disagreement quietly thins out. Rarer opportunities, larger when they come, which is an awkward payoff to size.
Nowadays we don’t remember the phone numbers but have used that headspace for something new and hopefully more productive .
In the 60s they used paper spreadsheets. It would take rooms filled with paper, many men and hundreds of hours to come up with a number. If they have to change one number, it would take them another few hours and more paper.
Now I could run 100,000 simulations on 5 variables with a click of a button and get way more insight that those men in the 60s.