Surgeon family doc and internist go duck hunting. Family doc is up first and sees ducks turns to the surgeon and internist and asks their opinion. Ducks are gone by the time he decides to shoot. Internist next sees the ducks and starts talking about what kind of ducks and describes their flight etc. Too late ducks gone. Surgeon up next sees a duck shoots immediately. Duck falls from the sky. Says I wonder what kind of duck that was?

---

They constantly try to escape from the darkness outside and within
Dreaming of systems so perfect that no one will need to be good T.S. Eliot

Got a few minutes to chime in. Well you're probably more qualified than I am to discuss technological viability and trajectory. When it comes to economic viability and affect I often start with this quote from William Gibson:



And then I often look at Gartner's Hype Cycle



IMO, autonomous vehicles are one of the more impactful narrow Ai's on the horizon but in the last year it's become apparent that we're probably 5+ years from any sort of large-scale rollout. It will be more like the tide coming in than a tsunami and this has less to do with the technology (e.g. Google/Waymo is very close to a viable product now) and more to do with the reality of dealing with regulation, consumer perception, and actually building the things and getting them out into the world. If you consider that the average LDV in the U.S. lasts ~11.5 years, and let's just say we're five years from autonomous vehicles shipping in volume (e.g. Tesla actually having a Level 5 vehicle, Ford having a dedicated Level 5 vehicle, etc.), and let's be optimistic and say that, initially, those Level 5 vehicles make up half of all vehicle sales

*throws pencil at napkin*

We're looking at ~15 years before a meaningful amount of the U.S. LDV fleet is autonomous and 20 years before the conversion is complete. Sound right? That's a lot of time for truckers etc. to see the writing on the wall.

[Picks up napkin. Looks at Rorschach pencil smudge...]

That seems like a reasonable timeline to me. WRT autonomous vehicles, I agree that the biggest hurdles will likely be political/regulatory rather than technological.

But over that same 15-20 year period, I think you'd be hard pressed to find an industry that *won't* undergo a similar shift to automation. China has an AI that has passed their medical boards. Musk has one that can beat humans in a battle simulation game (Imagine that tech in a drone...). Japan is developing robo-assistants for the elderly. Facebook has an AI bot that developed its own language. IBM has one that can beat humans in chess; Google has one that can best humans in Go. And Google also has has an AI that designs AI's better than humans can. (Yah, this one is particularly frightening. Fortunately, it's a limited-scope narrow AI in image recognition. But, still...)

I suppose the non-technological issues may affect the rise of AI in other industries besides AV too -- Will they face political pushback? Pay taxes? Sabotage? Anti-tech revolutions? A new arms race?

30 years ago, I thought that blue collar factory jobs would be the first to be lost en masse due to manufacturing robots. But, now I think that white collar desk jockeys will be the first to *really* be hit hard. In finance, medicine, manufacturing, etc., if your job involves manipulating/processing/interpreting some kind of data on a screen, your days are numbered. Radiology is the poster child here. In 10 years or less, there will only be a tiny group of Radiologist MD's left in someplace like India whose job is merely to rubber stamp the findings of the AI Radio-bots that will read all of the images.

I think that the jobs that are most resilient against AI displacement are those that involve manipulating physical objects in complex environments -- i.e., the trades. Stuff may soon be made entirely in automated factories, but installing, servicing, and maintaining that stuff out in the real world are challenging problems for AI/robotics for the foreseeable future.

Cheers,

-Mike

---

---

"100% of the people who confuse correlation and causation end up dying."

With the caveat being regulatory and industry group pushback (I'll use the example of anesthesiologists pushing back on 3M's automated anesthesia machine), I completely agree with you.

Waiting for an explanation/answer to this rhetorical question, what does nominal GDP have to do with this?

Um no. Actually it is the Liberal Arts major who actually markets and sells the product.

---

2017 Cervelo P2
2017 Cervelo S2
itraininla.com
#itraininla