re: Ex-Facebook Exec Warns Of ‘Revolution’ Caused By Job Automation

quote:

---

Someone will have to be in the truck incase of computer failure or mechanical problems.

---

I'm pretty sure that's not the plan.

quote:

---

I'm pretty sure that's not the plan

---

What are they going to do the first time their software fails or the truck needs a mechanic specialized in that trucks automation?

If it's going from Mississippi to California they'll tell it to pull over and then have a guy drive all the way out there.

Because we are only just understanding the degree to which AI is capable of replacing human labor. It really is in its infancy.

And no one is claiming that rampant automation is replacing jobs overnight right now (to my knowledge). We're saying that we're on the cusp of groundbreaking technological advances which will potentially prove more cost effective than human labor for a majority of jobs.

The transition won't be immediate, but a sufficient number of experts are warning us of a coming transition that it needs to at least be considered now. If we wait until it occurs and we don't have a plan in place, it will be too late to develop one.

I'm not in the 5-10 years away crowd. But within 2 generations, there is sufficient reason to believe that the majority of jobs which exist today will become obsolete. New jobs will always rise up, but the rate of new job creation will need to grow substantially in order to replace all of the jobs at risk of being replaced by automation.




And not to be too much of a snood about it, but I've read significantly more than 'a couple of blogs and articles'. That's a woefully inadequate description of my personal familiarity with the topic. I won't speak for anyone else though.

quote:

---

Deep learning algorithm does as well as dermatologists in identifying skin cancer:

---

So?

Dermatologists do a lot more than look at moles.

Someone still has to remove the cancer.

There are lots of different types of skin cancer that have different treatments.

Someone has to oversee the robot. Planes fly themselves, but there's still a dude in the cockpit.

The role of the doctor will change. AND there may be more business for dermatologists bc everyone with an iPhone will know they need their lesion removed.

As have I.

It's incredibly arrogant to think you can just hand wave a bunch of jobs away. The same people that say that stuff like to assume that something like public accounting will be automated away. People are just totally ignorant about the depth and dynamism of the service economy we have.

Spoken like a true doc. I'm withya brother. Take the lying out of medicine and then tech can take over.

quote:

---

So?

Dermatologists do a lot more than look at moles.

Someone still has to remove the cancer.

There are lots of different types of skin cancer that have different treatments.

Someone has to oversee the robot. Planes fly themselves, but there's still a dude in the cockpit.

The role of the doctor will change. AND there may be more business for dermatologists bc everyone with an iPhone will know they need their lesion removed.

---

The highly skilled, highly specialized doctor will still be needed. The topic is more relevant for the support staff that perform functions that could be automated or their activity becomes redundant. Whether those are administrative functions, nurses, etc.

Administrative - Patient records and account keeping is an area ripe for automation that only has regulatory hurdles in its way at the moment.

Nurses - Newer, more efficient medical treatments could be introduced that reduce patient care time (which reduces required nursing hours) or entirely new procedures may make a nurses job redundant. Maybe you can just directly see the doctor and don't have to go through intensive prepping and other tasks that nurses perform.

These are just a couple examples.

quote:

---

Its going to create new jobs and push people to new areas of industry.

---

This is probably the ideal situation, but then I find it odd that so much focus is on manufacturing jobs and bringing them, when technology has supplanted more manufacturing jobs than anything else.

In other words, I think you're right, but our priorities are misplaced to maximize this.

quote:

---

Many times the best course of action is uncertain, and decisIons are made based on probability, cost, risk, and the wills of the patient. That will be tough to automate.

---

I actually think most of those will be fairly easy to automate, or at least the initial aspect, and could eliminate cognitive biases that even highly intelligent physicians are not immune to.

As an example, there was a study that is presented in the book Thinking, Fast and Slow by Daniel Kahneman.

Physicians were presented a treatment and were randomly assigned to two groups. The only difference is that the treatment was presented with a 10% mortality rate or a 90% survival rate, which means the same thing.

HOWEVER, despite meaning the same thing, physicians presented with the 90% survival rate were like 2 or 3 times more likely to recommend that treatment than the physicians presented with the 10% mortality rate.

A machine that makes decisions would not be prone to the framing effects that impair our decision-making like in the example above.