Predictive Mediocrity

[Just a pretty picture of Valencia, as a Calatrava … follower but not groupie]

With all the talk about Predictive Analysis lately, it is time we separate, once again, the hype-humbug from the real(ity). Which is that the Predictive part is only partially true, and the Analysis part by definition is the contrary of predictive.

1. Analysis can only look at what’s here and now, or behind us. One can only analyse history. And, as Hegel pointed out, “Only at dusk does Minerva’s owl takes flight”: Only when the sun sets on some development, can we start (sic) to assess its importance. Analysis is devout of Meaning. It’s mechanical, it may abstract but not create Knowledge, Information at best but usually just more meta metadata. Again, interpretation, the error-prone and nearly-completely ununderstood brain process, is needed to make analysis worthwhile. And the result looks back, not forward.

2. Predictive… If we know anything of the future, it is that it’s uncertainty defined. One can predict some trends to continue — but may be disappointed.
Anything near the precision and the precise predictions that Predictive Analysis is pictured to deliver, will be certain to not come true.
This, as the dragnet of the analysis can never be complete, by the theoretical framework of needing a model of the universe which then would need the model itself to be anything distantly approaching completeness. And by practical considerations; analysts will have to work with the hammer they’re handed and view any world as nails only.
Also, trends may continue but ‘never’ (unless by extreme exception) linearly so, and non-linear modeling is still in its infancy (and may prove impossible by lack of suitable data). Hence, you’ll miss the mark by just extending lines.
Oh, it’s not about the detail, you say? Why don’t you stick to qualitative predictions, then?

3. Anything that can be analysed today, is tomorrow’s mediocrity as it was/is already known. The spurious results may be interesting, but we know whether they’re spurious or early indicators only afterwards, too late.

4. “Sh.t happens!” Why can’t we accept that in the medium-sized world we live in, quantum jumps do happen as well, with similar levels of chance calculus? They may not be binary quantum events, but appear to be. And make life both dangerous and fun! No surprises is boring par excellence. And will fail.

5. Go ahead and hype the mundane. But don’t oversell. You’re not a magician, or the magician is a fraud …

Leave a Reply

Maverisk / Étoiles du Nord