« Sandbox - 30jan18 | Main | Cirque du SOTUS (updated 3feb18) »

31 January 2018

Comments

Scenes

You can argue that one of the interesting side effects is to find out how many human occupations take little in the way of higher brain function. The tendency is to think in terms of moonshots like self driving cars, sexbots, and fully automated production facilities (or robot Presidents and radio newsmen).

Just to take a mundane example...grocery cashiers. At this point, their main value is that they provide a cheap robot arm to take items from a moving belt and run them over a scanner. That really is the hardest part of their job to automate.

I just mention this because people have this notion that the work will roll up from the very bottom. Instead, you'll see chunks of work bitten out of everyday jobs. Computers can do better jobs of diagnosing medical patients, I can see where pharmacies could be totally automated, etc.

It's all a continuum really. It isn't like jobs haven't been eliminated by pre-recorded music, travel agencies haven't died, income tax preparation is feeling the hurt. I'm afraid that underemployment is in our future.

As a side note, an interesting bit of machine training that is going on is the mapping of faces on video. Check it out.

George Rebane

Scenes 432pm - Points well made. It is though worth noting that RR readership is not all that much concerned about the near-term fate of Man as they are about flinging mudballs at each other for their take on the Nunes memo and Trump's SOTUS. Go for what you know.

Account Deleted

"Let’s take a break from the political goings on in Washington and Sacramento,..."
Can't. Washington to some extent and Sac-O-Tomatoes to a large extent are hastening the arrival of automation replacing jobs. In California, they are happy to brag about the Golden State being a hot bed of high tech and AI advances. And they are also happy to artificially drive up the cost of hiring humans via the govt gun. This hastens the bottom line answer to the cost of change over vs simply increasing the worker's pay question. Oh noes! What to do? The Dems have the ready answer. Tax the companies an extra amount that 'replace' humans with automation. Besides being another reason for a company to re-locate to another state, the money confiscated will supposedly be used to re-train the newly out of work workers. Re-trained to do what is the un-answered question.
Another important driver in the race to produce human like automatons is the porn industry. Remember the role porn played in the Betamax vs VHS war and the rapid adoption of CD ROMS with the resultant huge drop in price of that tech when the pornos saw the advantages of that medium. I was astonished when visiting the computer shows at Cal-Expo in the 90's to see that a lot of the larger dealers of 'software' were simply selling porn. Porn addicts are abundant and are quick to open their wallets as well as their zippers for the latest to satisfy their urges. The big money right now in developing human-like robots is in sex-bots. Make no mistake. A lot of what the porn industry is developing right now will accelerate the state of the art of AI humanoids in ways that will cross over into mainstream production and acceptance of human like automation in daily commerce. We already accept robo-calls and play the game of wading through automated choices when using the phone. Why will it be much different when confronted with a robot receptionist at the front counter?
Now - there is also the rising spectre of who will decide the sex, appearance, tone of voice and attitude of all of these robots. But that is a discussion for another day.

George Rebane

ScottO 611pm - Good points, and much discussed (almost). The point least discussed has been retraining workers. And as you correctly ask, 'Retrained to do what is the unanswered question.' It's as if when machines take over one area, another one manned by humans will suddenly open up and demand the addition of thousands of newly retrained humans. That scenario is not plausible, and most certainly nothing on which to bet tomorrow's workforce.

Russ

Here is a list of the top 50 jobs in 2018.

http://time.com/money/5114734/the-50-best-jobs-in-america-and-how-much-they-pay/

As you look through this list ask yourself could an intelligent machine do this job? What in the jobs listed requires the unique abilities that only humans have? Would you go to school, invest $60-100K, to learn a profession that in ten years would be done better by an intelligent machine? Students in school today are going to face some tough choices as they compete with the machines.

The comments to this entry are closed.

Blog powered by Typepad