Dave Snowden

robotic systems require robots

RSS Feed

A frantic few days, getting ready to fly out to New Zealand tomorrow. I never get used to the date line; you either loose a whole day, or you have to work the same day twice! This morning I gave a keynote at the UK’s main KM conference and choose to take a controversial stance on KM practice and the need for radical change. A few years ago challenging the orthodoxy of the standard KM approach (create a taxonomy, set up some communities, run an SNA and try and engineer a knowledge sharing culture) would have resulted in derision rather than approbation. However now experience has caught up (finally) with theory and people are ready to change. I’ll aim to pick up on some of the ideas in future posts.

From there to a meeting with one of a growing number of IT professionals who believe that their role in life is to reduce complexity by making it simple. In practice they are talking about complicated systems, not complex ones and being simplistic, not simple. If its complex you dare not simplify, you need to manage the ecology of the system not engineer an ideal solution. Faced with one of those idealised engineering drawings in which disparate systems are integrated by a centralised control system and repository I asked a simple question: how are you going to get humans to work that system? It’s predicated on ideal behaviour from all participants. His response was to argue for the benefits of all employees being run through a two day NLP indoctrination training session. I gather one Dutch utility has inflicted this particular torture on all its employees. It probably works as well, anyone intelligent would leave. OK a robotic system needs mirroring robots!, but that is sub-human and sub-optimal.

When I suggested (well I was stronger than that) that NLP was a pseudo science the counter argument was a testimony straight from the mercy seat to the personal transformation he had undergone. Self-reporting is the last way to authenticate a method, if that works then all religious cults are validated by the testimony of their adherents. The trouble is that self-reporting is the only validation of pseudo-sciences like NLP are able to offer. Indeed some of the academics at Surrey involved with this nonsense are arguing that only phenomenological research is valid in the field. What they really mean is that there is no objective validation, so they fall back to anecdotal testimony. I remember a good friend and intelligent thinker going off for the three week NLP programme and coming back with his intellect and critical facilities reduced significantly by the indoctrination of the cult. You expect teenagers to go through this, I’ll confess by own lapse at 16 in a religious retreat, but not adults.