Monday, August 22, 2016

Questions about Future Jobs and Foresight

Questions about Future Jobs and Foresight

I was stunned a few weeks ago when a major new international study (see reported that there will only be enough jobs for one-third OR LESS of the work force in just a few decades, in all the many scenarios they considered plausible, based on inputs from many different informed points of view from all over the world!! That’s not a small issue.

While US politicians have worked hard to convince the voters that they all belong either in jail or in the insane asylum (along with all the major party officials responsible for the situation), the voters themselves clearly worry more and more about the future job situation and about terrorism. “The voice of the people is the voice of God”? Actually, the electorate is a major part of the noosphere, and it is a shame that political operatives underestimate and misunderstand what this means, not only in the US but worldwide.

People generally agree that the world is heading for disaster on the jobs front. Lately I often remember a classic anthology by Feierabend and Feierabend, when they talk about the “J curve”, about thwarted expectations leading to political upheavals.  I also remember the French revolution, which started with a “revolution from above” as the aristocracy played legal games oppressing the people (and the noosphere itself!), resulting in their losing their heads. What COULD be done to avoid disaster? How could we come up with a new social contract or covenant, building on the excellent spirit of the original US and German constitutions, but keeping that spirit alive in a new era?

In the new era, I would argue that a realistic “new deal” or “new covenant” requires a bit of basic situational awareness of the (nonzerosum) game we are playing here. As I see it, the emerging new world will be a balance of three really paramount information forces – the humans (still important but not the monopoly they imagine), the noosphere and the information technology (IT) especially as in the new all-embracing Internet of Things (IOT).

Is the IOT really THAT important? Yes, folks, it really is, and will be more and more. That is why a new, more conscious IOT platform will be more and more essential to avoiding disaster here. For example, AUTOMATION is the number one factor causing this jobs issue, and that’s part of the IT sector. And the cyberblitzkrieg concerns I have mentioned are urgent life-or-death, and just a first step in aiming for sustainable IT.


At another level, there are really serious intelligent groups of people looking at the future jobs issue, and the role of IT, which will be important EITHER in a “traditional” employment scenario or in a “new age everyone self-employed” scenario or in a freedom-based mix of the two.

I have noted before, this requires:

(1) A really urgent effort to prevent premature ossification of IT, by folks who are pushing very hard for an old-style top-down IOT control system, which could wipe out market-based internet companies in much the way that IBM’s old Wylbur system once crushed its more user-friendly competition years ago. (When PCs changed that, I had friends who said “Beware, the Empire will strike back.” And it already has, more than you know.)
(See for colorful slides on both the Empire’s plans and on a better way for IT.)

(2) Serious technical efforts to improve the core IT used in electric power systems like Independent System Operators (ISOs) and extend it as a paradigm for other sectors, not only in security but in market design.

But in truth, more thinking and research is needed for (2).

This past week, people designing IT to organize the job markets of the future (with some input to some candidates) have asked my views of new ideas for “employment dating services”. My response:



Your remind me of things I should have asked about that I haven't gotten around to analyzing seriously yet.

We are really entering a new world here, and I have been thinking about foundational issues -- and haven't even gotten around to simpler questions involving markets simpler in nature than employment/service markets. Even for simple electricity markets, there is a crucial foresight function (will I need more of this n the future? how do I prepare now for future needs) which somehow needs to be apportioned between humans thinking about the future and partially intelligent systems which can help do the same, How? I need to look up the recent work on collaboration systems for prediction which work better than the betting system markets which were fashionable a few years ago.

For humans... many of the systems designed recently are based on the simple type I error versus type II error metrics used for other purposes. But additional metrics may be important. It's not just a matter of average classification error, but of systemic effects which can come from different types of systemic biases. I suppose that people designing dating services are far down the line in experiencing such things, but how much is their experience available to learn from?

Some systemic effects to pay attention to in the design process are obvious. Even if you don't think about racial or ethnic effects on day one, others will, for you. But even so, there are unanswered tricky questions involved.

But there are organization culture aspects to consider. For example, I long ago had a long conversation with a guy in DOD who had studied the use of lie detectors in CIA. "For technical reasons," he said, "related to the issue of calibration questions, they automatically exclude that very small percentage of people who always tell the truth about everything. Since it's a small percentage, they don't think it matters much." Ah, but some work on corporate culture says that the adaptability and integrity of large institutions may be highly dependent on that minority, most of which gets integrated into the organization in a crucial and constructive way. The role of whistle-blower personalities is another tricky issue, from the viewpoint of society as a whole.

In many organizations... the role of the minority which pays attention to reality, and meets a certain level of sanity, is another systemic issue to consider. As is the question of experiences in the organization which may foster growth in those areas.

In things like dating services... some people might say: "The most important thing is not to miss the really best opportunities." Back when I was looking for good proposals at NSF, it was especially important not to miss the high-potential (even if risky) options. But with things like dating... I tried such a service long ago, when I was...a freshman in college... and if the first two are a total miss, one might not even evaluate later ones which might be better. In the NSF model, failure of a risky investment is not much of a problem (as universities give backup security and opportunities to people who do not get grants or who fail in risky projects), but people really want stability in their lives; in my very first job, at DOE, I remember it created huge problems when people at ORNL expected stable close relations and the folks at DOE leading a grant were not prepared to provide that (they just wanted the near-term product). When does management on the receiving end (or society in general) really ask itself "what is the best use I can make of these specific people, long-term? How can I do the best for these people?" (Is there an imbalance between the demand side of this market and the supply side?"

In fact... between these elevated questions and the simpler questions I was asking about electric power, I told myself a day or two ago that I need to think more about monopsony effects and how they can very seriously interfere with the quality of these kinds of markets, unless we are very careful to understand what we are doing and where it may lead. If society jumps ahead and implements things before we know what we are doing, it could be very hard to fix it later.

(Of course, I should also have mentioned the need to consider how people might game such systems. And also what we can learn from the great innovation of test-based employment going back to the Song dynasty of China.)


No comments:

Post a Comment