Sunday, September 05, 2010

Thoughts from Wells and Reading

1. I write this on Sunday evening from Alex's flat in Reading: I've just driven up from Wells, ready for another hard week on my current contract. My inbox has more than 400 unread emails; I have 24 voice-mails yet unlistened to and all this stuff has to be transformed into orders to be entered into a slow and brittle computer system. So it's another week of real early starts.

Where is Alex I wonder? I thought he'd be here. He's been wandering the continent the last two weeks from Paris to Amsterdam to Berlin to Heidelberg and then we think to Herculaneum, Pompeii and Vesuvius. Mostly by rail.

Anyway, he's got work in London tomorrow so I assume he'll be showing up here at the last minute.

2. Adrian must also have an adventure to tell as he's snowboarding in NZ South Island based near Christchurch. That would be the epicentre of the recent earthquake then. Still, I guess there are more important things to think about than emailing home.

3. Clare's sister Mary and husband Gerry were staying with Clare down at Wells last week. I caught up with them Friday night and we had a pleasant Saturday, dining at a recommended pub The George Inn, Croscombe. The food was every bit as good as promised. Yesterday evening I got some much-needed exercise by walking up to the top of the Mendips, Clare snapped the picture.

The author north of Wells

4. You may have seen those (American) vehicles being driven around simulated towns by computer - it was a Department of Defense challenge. The automated chauffeur is a promised spin-off "soon".

Not so.

A useful test is this. If you can do a task so automatically that your mind can drift off to a quite different topic while you're doing it, this means that the task has been automated in the subconscious neural machinery. Athletes talk about being in the zone or in the flow; they're not consciously thinking about what they're doing.

I think these kinds of accomplishment are amenable to automation because they're closed condition-action solutions.

Driving can be like that a lot of the time - we sometimes go off on a train of thought and several miles have passed on a familiar route. Yet over a longer period inevitably your attention is brought back to some road problem you have to think consciously about. Someone has done something strange, or there's decision to make which plainly depends on what you assess other drivers are likely to do. Then all of a sudden the task requires your full attention.

That's what will trip-up the proposed automation. Just watch progress - you'll see success after success in semi-realistic, controlled environments. But in the real world it will all fall apart once in a while - and that's enough to rule it out.

The real answer is to re-engineer the roads themselves so that they're electronically readable and semi-smart. Then the open nature of today's driving experience can be engineered to be closed. More like the railways which are surely more amenable to driverless solutions.

5. My take on this is that the jarring, isolated, international jet-set life of a theoretical physicist can be profoundly incompatible with the balanced life evolution equipped our bodies for. I wish her the best with her reassessment and sympathise with the existential dilemma she faces.