Richard’s review of Rise of the Robots: Technology and the Threat of a Jobless Future > Likes and Comments

Comments Showing 1-21 of 21 (21 new)    post a comment »
dateDown arrow    newest »

message 1: by David (new)

David You often refer to reviews of books you're interested in. A review of this book is supposed to be in the 3/29/2014 Economist.
On my to-read list.

message 2: by Richard (last edited Jul 26, 2015 11:27PM) (new)

Richard I found what you're referring to — but it isn't a review about this book, and surprisingly doesn't even refer to Martin Ford, despite his earlier book, The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future .

It is one of the Economist's Special Reports that uses the same title in their "leader" (their term for an editorial essay that also discusses something that is covered inside an issue in more depth). If you want to check it out, that leader is here: New roles for technology: Rise of the robots , and a PDF of the special report is here: Immigrants From the Future.


Note: I originally had a long list of online reviews and interviews holding the place for my eventual review. I'm moving them here, so folks can still peruse them:

Note: I haven't read all of these, or listened to the podcasts, so some of the following might be misguided, amateurish, or tangential:

Good overview video at PBS Newshour (8+ minutes):
Related article about three employment fields susceptible to replacement:

Text on Wired:

Text at the Wall Street Journal:

Text at the New York Times (along with Shadow Work):

(The author of the above discusses her review on the NY Times Book Review podcast, discussed here, but the only way I can see to download it is via iTunes, here on May 14th, 34 minutes.)

Text at the Daily Beast:

Text at the Financial Times:

Text at Business Insider:

Audio on NPR 's All Tech Considered (30 minutes):

Audio at NPR's Fresh Air (48 minutes): http://freshairnpr.npr.libsynfusion.c...

Audio at the Commonwealth Club (59 minutes):

Audio on The Week (4 1/2 minutes):

Audio on Review the Future (55 minutes): (transcript:

Many more are linked to in the author's Twitter stream:

message 3: by Caroline (last edited Jun 09, 2015 09:54PM) (new)

Caroline The FT and the Wall Street Journal ask you to subscribe before viewing. Am going to explore some of these other links... Thank you so much for providing - as always - such a wealth of resources :O)

message 4: by Richard (new)

Richard FT asked me to create an account for a few free views, but didn't force me to subscribe. But they also gave me the opportunity to answer some trivial questions, which seemed painless. But while the two-page or so review is pretty good, there wasn't anything in it that I'm sure isn't covered by the others.

message 5: by Caroline (new)

Caroline Okay, thank you. I picked up quite a lot at the PSB site...

message 6: by Caroline (last edited Jun 16, 2015 05:23PM) (new)

Caroline I sent my brother links to some of your recommendations and got this email by return...

"I think we are with robots where we were with PCs in the early 70s.
Sigh humans just get in the way."

I think I need to recommend the other book on AI that you have just read!

message 7: by David (new)

David Regarding "humanlike" AI / AGI's: Defining humanlike is an issue. How many standard deviations from average can the IQ of a humam or machine be and be "humanlike"? Is the fictional character Superman "humanlike"? If we could transfer Richard's mind into an android body, would that entity be "humanlike"?

Most important, as far as I'm concerned, is that an AI / AGI have a "conscience" which would at very least motivate it to treat us as members of "intelligent beings" which have certain rights and protections. I'm not clear to what extent a non-evolved entity such as an AI needs a selfish element in order to be an "individual" rather than just a tool for humans. At least without that knowledge, I tend to think I'd put an AI's selfishness at a very low level - which might make it "not humanlike".

message 8: by Richard (last edited Jan 11, 2016 11:33AM) (new)

Richard David, the fact that defining human is difficult is the problem.

All of the functional things that we want AI to do are pretty easy to define. It's even possible to envision an AI project intended to serve as a robotic personal care assistant, such as the one portrayed in the movie Robot & Frank, that would still be "merely" functional programming, with no need to program a sense of its own identity, its own wants and desires, etc.

The idea of an AI "consciousness" really belongs over at that other book: Superintelligence. This book, Rise of the Robots should have focused almost completely on the functional AI issues.

But I'm preempting the review I'm supposed to sit down and write, so I'm gonna probably ignore any more Q&A until I've done that :-)

message 9: by David (new)

David Thanks for the extensive discussion and links.

One of the proposals for individuals wanting to avoid occupations which will be automated is to find a job where you must "think outside the box". But the books and articles discussed here don't seem to "think outside the box" in approaching solutions. They propose individual or institutional actions constrained by the current socio-economic framework. They don't tend to say, "This isn't just a potential crisis, but has a potential to mold a new era in human history." Instead, there are ideas on how humans can fight over the few remaining good jobs, and how to avoid mass starvation (ignoring the existence of this immense potential). Automation offers the potential for producing our goods and services more cheaply, with less risk to health of human employees, with the possibility for more leisure time, the possibility for more specialized "workers" than there are humans with those capabilities, etc. We could be asking, "How can society provide the greatest good for the greatest number while taking wise advantage of this potential? Would it be better to democratically choose and organize how automation reshapes society, or to have it decided plutocratically by businesses solely based on individual profit for a tiny minority?" (I don't believe in depending on an "invisible hand" rather than using intelligence and knowledge to find solutions.) At very least, it's worth including in the discussion as part of the process of clarifying what is well-founded and which are based on mere assumptions.

message 11: by Caroline (new)

Caroline Very interesting. It is impressive the number of variables it can encompass, eg how much someone has had to drink within an hour.

I would however feebly (hopefully) speak up in favour of cocktail bartenders. Surely it is the whole razzmatazz of their performance that adds to the flavour of cocktails - also the amazing range of drinks they are able to produce. Would I rather go to a bar manned by a bartender maestro, or would I rather be served by a black box?

message 12: by Richard (new)

Richard Like many other of the service professions, it really depends on what part of the market one is in — which is actually a bit backwards.

Let me explain. At the low end of service professions, service can easily suffer; after all, you get what you pay for. And if an employer is offering minimum wage for a lousy job, only the desperate (who probably have few skills) will apply.

At the high end, payments are higher and the skills will be broader. For example, a prior girlfriend of mine is now a private flight attendant. She only works corporate jets, but gets paid very nicely, only has to deal with business people on their best behavior, etc. But she has a nice degree and has had several careers and is a great conversationalist, which is why she's able to get a nicer job than on regular airlines.

Right now, the skillset at the low end would be easier to replace, such as in a joint where the bartender is mostly handing out bottles of beer. But they don't get paid as much, so there isn't much incentive to swap them out for experimental robotic equipment. Once the cost of that equipment comes down far enough *and* more and more of the clientele have gotten used to interacting with machines, that will change.

At the high end, the cost of labor is much higher, but the skill set is much, much tougher for robotics and AI to replicate. At the very high end, very skilled humans will probably always have an edge.

The "foxtender" gizmo is a foray in a niche between those two. Obviously, it won't be adequate for a real bar. But it might be great in a back room of a bar that the waitstaff doesn't go into very often. If it actually succeeds (I doubt it will — I think it's about three or five years ahead of the curve), it might roll out as something that is rented out to parties, for example.

In the coming years, as more and more of us checkout our own groceries, skip the check-in lines at the airport or car rental agency, and maybe even start using those high-end espresso vending machines that already exist elsewhere in the world, these "foxtender" type devices will start to gradually replace some bartenders.

Not all — I suspect they'll replace the bar itself, leaving the waitstaff as humans who use a tablet when taking the orders, which tells the robot what drinks to prepare. So many interactions will still be with a person. But the bar, itself, will eventually have a fancy robot at its core, able to produce any drink ever invented, and then some.

But, as you point out, at the very top end, where humans are considered part of the decor, very talented bartenders will still be hired — although I'm suspicious they'll increasingly be like the models hired to essentially be furniture at extravagantly decadent parties that (probably) only exist in Hollywood's imagination.

message 13: by Caroline (last edited Aug 13, 2015 01:36AM) (new)

Caroline Ahh, all of what you say makes complete sense. Thank you for responding so thoroughly.

Your recent readings/reviews about robots and AI have made me extremely grateful to have passed retirement age.

message 14: by Richard (new)

Richard When you add climate change into the mix, and the (in my opinion) slide of U.S. political culture into endless dysfunction — what's that horror story tag line? "The Living Shall Envy the Dead!"

Maybe not quite that, but the Chinese curse, "May you live in interesting times" definitely applies.

message 15: by Caroline (last edited Aug 13, 2015 02:17AM) (new)

Caroline I agree totally. The only difference is I am subject to the English rather than the American political system, and given the corrupt and power-crazed political situation in many, many countries, I must admit that for the most part I feel fortunate.

I watch American politics with a large degree of ignorance, but from our vantage point some of it does seem rather strange.

message 16: by Jim (new)


Not quite on topic but interesting

message 17: by David (new)

David My wife just read me an article by a (former) public school teacher who, with sadness, switched to working at a rich private school. She gave up on the falling pay & benefits, increasing out-of-pocket expenses, time wasted on standardized tests (some of which were never acted on), etc. Like your ex-girlfriend, she was skilled enough for the posh job.

So, I wondered, "How long before we have robotic public school teachers?" The US is clearly more interested in small public school budgets than highest quality teachers, so the robots only have to be so-so. "Teaching" in a format for students to regurgitate answers on standardized tests sounds like a job description for a robot. Those troublesome public school advocates will say that as automation shrinks human job opportunities we need better student education. But sensible business people will know it just means there's no point wasting tax money on human teachers. Besides, it will be good business for the private schools with human staff. I think I'll go ask that robot bartender for a stiff drink.

message 18: by Richard (new)

Richard Robotic school teachers will take a long, long time, but augmentation will happen much faster — there are already several startups that are trying to find ways to improve teaching.

One, for example, has computers monitoring students facial expressions during class, trying to predict which ones are "getting it" versus being distracted, looking confused, etc., and providing that to the teacher as feedback (eventually as instant feedback, but not at first). That can get better over time, too, as the software correlates different prior scoring with what it perceives and how scoring is altered.

The book also talks about how MOOCs have tried and failed to "disrupt" education (mostly higher ed).

message 19: by Caroline (last edited Oct 19, 2015 02:16AM) (new)

Caroline I have now re-read this for about the third time - and again with much interest. I watched the 15 minute video (re the horses who hoped to be re-assigned to different tasks once cars, lorries and tractors came into being...)

This time I was particularly interested to read what you had to say about the difference between AI, and AGI. I don't know what that stands for, but I'm pretty sure you are referring to robots or artificial intelligence which functions in lots of different respects, rather than just working on one specific job. One that fills our sci fi fantasy of a classic robot. These latter creations are mind boggling in the extreme. I find it very hard to believe they can really exist - but that video showed us that they do.

message 20: by Richard (new)

Richard Right, but a little more extreme. AGI (Artificial General Intelligence) is something of an approximation of human intelligence.

(This is related to the idea of "general" intelligence that psychometrics claims to be seeking in human IQ tests, called the "g factor".)

Most AI is purely functional: designed to solve a specific problem. Software designed to, for example, drive a car, wouldn't be capable of even seeing a chess board, much less playing a game of chess. Similarly, an AI program designed to play chess won't "know" anything about driving a car.

That's not to say that a computer inside of a car couldn't run the two programs simultaneously, but that is quite different — having two distinct programs running on a single computer doesn't somehow let those two programs combine anymore than having two people in the car will merge them into a single being.

Very distinct from functional intelligence is general intelligence. Artificial General Intelligence would be intended to solve many or most of the problems humans are able to solve. The "would be" is important: nothing like this exists, and the research labs working on it aren't really even close. They're still trying to figure out many of the basics.

Keep in mind, though, that advances in functional intelligence might make parts of the job of AGI easier, such as vision systems, or voice recognition.

Yet another level of intelligence would be consciousness. It should be possible to design an AGI that is complex enough that it does a great job mimicking humans, but still not trying to mimic a mind — with no ego, no personal ambitions or desires, no "soul". But beyond that, there is the concept of actually creating a fully conscious "person" that would have those attributes.

It is conceivable that AGI that mimics a human level of intelligence wouldn't be possible without consciousness, or something so close to consciousness that the difference might not matter (although this is quite controversial). Of course, this debate is complicated by the simple fact that no one really has come up with a strong explanation of what human consciousness is, yet.

An important point I tried to make is that all this latter stuff doesn't matter in the world of replacing human jobs. Functional intelligence is rocketing ahead, regardless of progress in AGI or consciousness-level intelligence, and functional intelligence is all that will be required to replace most human jobs.

message 21: by Caroline (last edited Oct 19, 2015 04:27AM) (new)

Caroline Thank you for your thorough explanation - and yes, all of what you say makes complete sense. I totally understand that AI is what is going to replace human beings at work... as it has done in so many different ways already. I also now understand a bit the massive scope of the challenge in creating AGI - even if it doesn't come near the scope of human consciousness - it is still a massive challenge.

By the way, after watching the video about humans losing jobs to AI, I saw another video on the sidebar which I found very interesting. The guy giving the speech looks about 15. Here is his TEDx profile:

And here is his talk about the possibility of giving people free wages in the future.. I found it most interesting. Given that so many people are likely to lose their jobs in the near future, it seems that perhaps we have to look at this as a possible way forward.

back to top