The basic calculus is that if you decrease the number of minutes of doctor time enough, it doesn't really matter if they are 100% correct (0 minutes of 100% correct = 0 advice).
Patients are substituting 15 minutes of 100% correct information for 2 hours of 80%, or whatever percentage the AIs are.
Additionally, they're pretty much free at the moment and available 24/7 with no travel required.
My one and only experience of dealing with the police in the US was when I was visiting NYC. A tourist was being attacked on the subway because he was taking pictures and since we were still at the platform I jumped out and told 2 officers further down the platform what was going on. I expected them to sprint into action, but they could not have cared less and casually strolled along towards the carriage!
In a similar vain I was the first on the scene of a car crash in the UK, where the driver had exited the vehicle through the window (no seat belt) and was bleeding in the road. When the police turned up they casually and slowly walked up the road towards the scene.
It made me wonder if there was a good reason for this, like to control adrenaline, make better decisions, have time to assess the situation. Or if they were just jaded from seeing it a lot.
When working for LUL (London Underground limited) I was told to never run towards an emergency because you risk tripping and falling and then you’re another person that needs help instead of being able to provide the help. So maybe that’s why? I’d walk with urgency though, not casually stroll.
I'm less interested in what "grade" the AI gave and much more interested in what therapy or remedy it would have suggested. That's curiously lacking here.
Useful perspective. I think viable second sources can also lower costs. They might not increase QUALYS but would probably lower the cost of existing QUALYS so that money could be spent on other needs.
That raises an interesting question: if 10 people in a room request ubers without confirming the ride-hail, does the price go up for successive requests?
Hm... I think I would predict that it would depend on speed because querying prices reveals information, but not accepting also reveals information, so I would expect that it would go up initially due to an expected shortage, but then when it becomes clear that no one is accepting, it reverts to normal or lower, because that suggests low urgency or willingness-to-pay compared to what the models had forecast for that exact location & time, so they have to be offered lower prices to accept at all.
(This is the flipside of personalized pricing: it's also lower prices, as noted in some of the anecdotes, although he mostly emphasizes the potential for higher prices.)
> Medical AI: In 2016, Geoffrey Hinton—the “father of modern AI”—predicted that radiologists would be replaced within five years. We are now a decade past that prediction, and radiologists are as essential as ever.
This one is instructive. State of the art radiology image classification models match or exceed the performance of expert human radiologists.
But radiologists haven't been replaced... why?
The simplest answer is that classifying images is not all a radiologist does, and therefore cannot be replaced completely by a computer.
Radiologists serve as the human-in-the-loop to blame if something goes wrong. Also, there are myriad regulations in healthcare that slow or prevent the adoption of tools like this: FDA regs on medical devices, insurance stuff, professional associations, etc.
Like much of the tech world, the problem often isn't the tech, it's the context.
> ALL UNMANNED ACFT ARE PROHIBITED FROM FLYING WITHIN A STAND-OFF DISTANCE OF 3000FT ... LATERALLY AND 1000FT ABOVE ...
> TO: DEPARTMENT OF DEFENSE (DOD), DEPARTMENT OF ENERGY (DOE), AND DEPARTMENT OF HOMELAND SECURITY (DHS) FACILITIES AND MOBILE ASSETS, INCLUDING VESSELS AND GROUND VEHICLE CONVOYS AND THEIR ASSOCIATED ESCORTS, SUCH AS UNITED STATES COAST GUARD (USCG) OPERATED VESSELS
I came across this guide (dated 2025) a couple years ago and thought it was interesting. Not a quant or even in finance though, so I don’t know how accurate it is:
I use markitdown[0] religiously. You’ll lose fidelity for anything complex (math equations, images), but it does a great job 95% of the time in my experience.
I’m assuming the attack surface is reduced. I invoke it through a docker container. But this might be a misplaced sense of safety.
A one-way link (data diode) transmits it to a box with simplified hardware (eg RISC architecture). The box has a dedicated monitor and keyboard. Once you're finished, you sell the box on Craiglist. Then, buy a new, sealed replacement from Best Buy.
Pay per view was an expensive, business model for cable. For PDF's, it's even more expensive.
Note: It's more convenient than full, per-app, physical security.
Don't think it's a guide like "Hey here's how YOU can BECOME A QUANT!" bs, more about the stories/history of 25 different quants. So more of a personal story/history doc. I think it's an interesting read if you are a finance nerd
https://www.roadandtrack.com/news/a39481699/what-happens-if-...
reply