Sister blog of Physicists of the Caribbean in which I babble about non-astronomy stuff, because everyone needs a hobby

Monday 23 April 2018

The march of progress is not inevitable

Well, I dunno. Certainly the hype about AI being able to do everything for us by next Tuesday or in five years or whatever is a massive exaggeration. But I think this particular article goes too far in the other direction.

At the same time, we cannot predict the numbers of new jobs/careers that new technology will create. [So let's immediately give a prediction then] One study from Gartner Research states that while 1.8 million jobs will be lost by 2020, 2.3 million new ones will be created. Even today, there are a huge number of technology jobs that did not exist ten years ago: State-of-the-art programming, data science, web security, marketing and sales. There is no reason to believe that the need for humans to create and manage new technology will decrease.

Literally all of those jobs existed very much more than twenty years ago, with the single exception of web security. Technology didn't create any of those jobs, but it has certainly made them easier, e.g. programming in Python versus punch cards. However, while it would be nice to think that essentially anyone is broadly willing and able to do those jobs, I'm far from convinced that is the case. "I liked astronomy but can't / don't want to do the maths" is something I hear all the time. And you've got to have people who are both willing and able to do the jobs replaced by robots/AI (see https://plus.google.com/u/0/+RhysTaylorRhysy/posts/MUhGwJT3HqR) for more complex definitions of AI), otherwise you're just replacing one dystopia with another.

Robots and AI will certainly replace jobs – boring, dangerous, and dirty ones mostly. Consider coal mining for example. How many people still want to go down into a mineshaft and dig out coal? How many want to subject themselves to black lung disease and a host of other health problems from that job? This is why coal mining towns are dying out. Young people in these towns are moving on to the brighter job prospects. And technology is taking over what’s left of the mining industry. Green energy is taking over, and with it, a host of new, clean jobs and careers. It’s the march of civilization that will never cease.

OK, that's ludicrously happy-clappy naivety right there. The author needs to swallow a bunch of history books before being allowed back on the internet. Also, young people today are earning less than those of their parents generation of the same age (http://www.bbc.com/capital/story/20171003-millennials-are-the-generation-thats-fun-to-hate). Granted, that doesn't mean technology is causing a problem, but for that very reason it's crazy to infer an inevitable march of progress based on technological advancement. Politics, in its broadest possible sense of how humans relate to each other, is already the reason we're not living in a golden age; in the material senses we've largely never had it so good. In previous ages a lack of knowledge (e.g. of the technological advancements needed to cure diseases) may have been the limiting factor, but today, by and large, it is political concerns which prevent discoveries and abundances from being applied correctly and distributed fairly.

Those entering the workforce today will have to be adaptable. What robots can do is make humans more productive than ever before. Workers will need to develop technical skills and keep those skills updated as technology moves forward. Those who do not want to deal with technology need to pursue careers where it is not as much a factor or where demand for human skills and talents remains high.

Yeah, what careers ? Are people actually going to be willing and able to do them ? The author is guilty of the very sin he accuses the technophiles of : unjustified assumptions with so little context that a debate becomes very difficult.

https://www.forbes.com/sites/andrewarnold/2018/04/20/why-robots-will-not-take-over-human-jobs/#155a23a747cd

5 comments:

  1. What we now functionally have are decision-tree systems, no free initiative, some learning by example or experiment, but no practical intelligence, and i do not see that or even the holy grail of sentience even as a glimpse on the horizon. Even those very 'sophisticated' machines, beating humans in some thinking sports, are just doing very dumb shit, just very fast, no flicker of insight, just wading trough all options and selecting the best one.
    It is not even possible to not work down a (for a human) clear loosing proposition and dismiss it without havening to run the whole scenario down.
    Thats not intelligence, so i stand against the doom-sayers that predict all human jobs will be lost, as it will be a long time down the road before that happens.
    Live and the way we live it will evolve, we survived the steam era and the following industrialization, even than they predicted the end, horsewhip makers, buggy builders, horse breeders and many more would starve jobless.
    New jobs will evolve, we just need to be ready and not let our fears rule us.

    ReplyDelete
  2. There are several issues here :
    1) The fraction of jobs that the current trend in automation may replace human workers.
    2) The rate at which new jobs could replace them.
    3) Whether either 1 or 2 are significantly different from previous increases in automation.
    4) If jobs are replaced (with other jobs or something else), whether that makes people feel more or less fulfilled than previously.

    I don't see a shred of evidence of any true higher reasoning being developed by AI yet. But I don't think that this means a more drastic change than in previous shifts isn't happening - the methods used may not be true intelligence, but that doesn't matter if it can produce the same results. So I think it's entirely credible to suggest for point 3 that this time we might be replacing labour in a way that's fundamentally different from previous generations : we could end up replacing jobs faster than new ones are created.

    This may or may not be a good thing though, hence point 4. Personally I think it would be no bad thing if an awful lot of people worked quite a bit less and had no threat of living in squalor as a result. Not doing any work at all I'm less sure about.

    I don't think there's any threat of a technological singularity, but a very steep cliff does seem like a realistic possibility to me.

    ReplyDelete
  3. Yes, we do live in interesting times, unfortunately.

    ReplyDelete
  4. For the life of me, as someone who's been doing industrial automation for three decades now, I cannot fathom the torrent of bullshit surrounding AI, automation and job creation.

    Since the invention of the first effective steam engines, they've been "taking jobs" - first of all those donkeys down the mines in the Black Country, then the carters were replaced with railways. Child labour ( of children under 9 ) was outlawed in 1833. With the improvement in metallurgy, we got the internal combustion engine and more disruption of old job descriptions and the creation of new ones. Iterate the problem every decade ( more often these days ), mankind's "jobs" are increasingly defined as managing the machines. Thus we see both fewer meaningful jobs and lower wages paid for those jobs.

    That's capitalism and it's pointless to whine about it. Capitalism is a mighty engine, capable of pulling entire societies to a modicum of prosperity, if we have the good sense to harness it to the service of a better society and not the creation of great glaciers of frozen capital.

    The problem is our definition of the Job. Jobs as we understand them didn't exist before the Industrial Revolution. And as mechanisation proceeds apace, the old "Job" paradigm of managing the machines will be reduced to handling the few exceptions and not the tedious rules, which AI can reduce to a policy network of rules, returning Verdicts. I write this sort of code all day long. My cameras use it. So does my rice cooker. And my washing machine.

    No age is a Golden Age to those living in it. As a species, we've got choices to make. Our machines are evolving faster than we are, as a species. Neither optimism nor pessimism is entirely warranted but one thing seems clear: we've come to the end of this particular phase of the Industrial Revolution. The Job is dying.

    Out there on the South Downs, in Sussex, all those wide, windblown open spaces, sheep here and there - those were once all beech forests. Neolithic man cut them down. Here and there a few pitiful remnants remain: Arden, Epping Forest. And now, those forests are likely to make a comeback. I have a bit more faith in humanity than some. We're going to make it.

    There will always be Jobs of some sort, designing things, selling them, handling exceptions. But they'll be rare. Mankind will evolve into a post-job world. The Job was always a rotten deal, anyway. Marx told us so, early on. Don't love your Job. It won't love you back. I believe we're going to see something emerging which looks rather like democratic socialism. Won't be called that, though.

    ReplyDelete
  5. Dan Weese Well, you've answered your own question. For surely no-one ever said, "it's only a paradigm shift, so it's nothing to worry about". ;)

    ReplyDelete

Due to a small but consistent influx of spam, comments will now be checked before publishing. Only egregious spam/illegal/racist crap will be disapproved, everything else will be published.

It's okay to like vinyl

Here's a nice if somewhat over-lengthy piece about why people prefer antiquated technologies like vinyl records instead of digital medi...