Masch’s Works

Vladimir A. Masch has various articles that tackle aspects of uncertainty. He has some interesting ideas on economics, but here I comment on his long-standing interest, practice and academic contributions to ‘planning under uncertainty’.

I look for any such work to be informative for any strategizing under ‘radical uncertainty’, as it is. But (as with even the work  of ‘the masters’) I also look for snags and raise some quibbles. It is up to you as to how significant they are for the problems you have in mind. Most of them are fairly generic and may, perhaps, be explained by the need to fit within the ‘best practice’ of such papers. You may think that my quibbles are only pointing out the obvious, but in my experience what is obvious to us may not be so obvious to everyone else.

Extensions of stochastic multiscenario models for long-range planning under uncertainty

This presents an explicit method, risk-constrained optimization (RCO) for ‘long-range’ planning under uncertainty.

  • Inevitably the method has explicit stages that are considered in a definite order, which ends with the injunction to ‘start the process anew’ if things aren’t working out.
    • Actually as I went through the stages I would have concerns that I would look out to be dealt with at the later stages. Ideally, I would review these concerns before committing to action, and ‘start the process anew’ but with renewed insight to see if I could do better.
    • I would typically think ahead through the stages without doing them in detail, and if I could foresee any particular issues that were likely to be raised I would do some preparatory work before formally starting the process. Alternatively, if the case seemed simple I might take a simplified approach.
  • The approach relies on identifying ‘factors that contribute to uncertainty’ (2.1.1).
    • This is a key insight, and – in my view – an improvement on simply considering ‘worst case’ scenarios.
    • In theory, and my own experience, one can hardly rely on one’s initial findings. In many cases – such as when considering potential economic crises – it seems to me that there is not even a theoretical possibility of being able to come up with a definitive list. Rather, one has an initial list of what seem to be the important factors, and develops that list as one goes along. At any time one might have new insights and want to go back and add or amend a scenario. I agree with Keynes that one can never be sure that one has all the relevant factors for a particular situation, and so this relies on a rare combination of imagination and judgement based on soundly comprehended experience, which often requires engaging with a broad range of contributors.
  • Each scenario is scoped such that the associated uncertainty is relatively ‘tame’: the radical uncertainty is essentially dealt with by considering  broad of scenarios.
    • This seems ‘state of the art’, in that I don’t see any reasonable alternative. It is certainly better than ignoring radical uncertainty.
    • In practice, though, in working through had seemed to be one scenario one may come to recognize another source of radical uncertainty and hence need to split the scenario to keep them relatively tame. If this does not lead to a mass of scenarios that would overwhelm any analytic capability then maybe you aren’t drawing on people with enough imagination or experience. So it seems to me that here one need some developed practice at planning under uncertainty, or else doing it for real under real pressures seems to me risky, almost doomed to some sort of failure.
  • ‘The real goal is to reveal and remove dangerous outcomes of any risk type in any scenario’.
    • Again, a key insight.
    • Base on my own experience, I think that analysts who had a sound appreciation of radical uncertainty, based on there own experience and practice, would find Masch’s approach insightful, more so than some of the more familiar texts.
    • But I doubt that analysts who lack such a sound appreciation would actually achieve the kind of results that Masch would seem to favour.
    • Perhaps ‘war gaming’ some issues, so that analysts can either demonstrate or develop their  appreciation of some of the involved in facing up to and struggling with radical uncertainty, will be necessary.

Overall, I think that no methodology is enough. RCO alone cannot be the whole story, and we cannot really judge RCO except in a broad enough context. I think that both culture and underpinning theory matter. Maybe I should delve deeper into Masch’s underpinning work.

“Shifting the Paradigm” in Superintelligence

This notes that ‘ordinary’ artificial intelligence(AI)  has largely been developed without consideration of radical uncertainty, and argues that in many of the applications for which variations of AI (including ‘big data’ and ‘deep learning’) are being touted, decision making under radical uncertainty will be required. It has an interesting discussion of some of the issues:

  • Some of the discussion reminds me of Taleb’s notion that people who do not have ‘skin in the game’ will not be motivate to recognize let alone tackle uncertainty.
  • More hopefully, though, Masch’s link to issues of sustainability and human survival suggest to me that in fact we all have ‘skin in the game’, if only we knew it.

It seems to me that a critical limitation of Masch’s approach, as of many that one might compare, is that in the end all it seems to promise is that it will make the most intelligent use of the decision makers’ insights.

  • This is an advance on the current situation, where many ‘experts’, analysts and academics lack recognition of any uncertainty beyond the tamest ‘probabilities’, and constrain themselves to reason ‘pragmatically’ as if probabilistic reason were adequate.
  • It needs to be complemented by steps to ensure that the main parties have an adequate grasp of radical uncertainty and its implications. It is not that Masch does not adequately address these concerns in some technical sense, but I do not judge his account to be sufficiently compelling.
  • Currently, ‘populists’ are – I think rightly – critical of the motivations of experts and elites, whereas more mainstream folk are, I think rightly, sceptical about the ability of populists to develop and promulgate the necessary understanding without going through some painful ‘learning experiences’. Thus, it seems to me, the issues are much wider than ‘strategic  planning’.

Overall comment

Masch’s work should be of interest to practitioners and theorists, and deserves wider comment, hence this critique. At the least any alternative practice could benefit from review in the light of these ideas – unless you know of something better, in which case please let me know!

Dave Marsay

P.S. Masch’s bio is also worth a look.

 

 

 

Advertisements
%d bloggers like this: