Archive for October, 2014

Tom Forth and John Popham Speak

Sunday, October 19th, 2014


The second Bettakultcha podcast is available to listen to.

Behave predictably

Monday, October 13th, 2014

Having just jumped through a variety of security hoops to move some money from an old online bank account to a new account which pays better interest, I’m told by the ’system’ that my transfer is being reviewed. I can see the sense of this – any unusual activity could be seen as being potentially fraudulent and needs to be checked out. On the other hand, I could have just woken up from my inertia and realised that I was losing money by leaving this money in the dormant account.

The problem is this: to get a decent service from ‘the system’ it pays to behave in a predictable manner. If you’re impulsive, then the system will put obstacles in your way for ‘your’ benefit and security.

What this means is that we are being slowly conditioned into behaving in a predictable manner. And authorities just love populations that are predictable and compliant.

Of course, when it comes to the banks themselves behaving in an unpredictable and unlawful manner, there are few, if any, security checks.

Why algorithms are so scary

Friday, October 3rd, 2014

Crocodile trainers have an algorithm for the machines that they work with. The machines are the crocodiles and they display highly predictable behaviour. The trainers know that if they do such-and-such within such-and-such distance from the crocodile’s mouth, the crocodile will do such-and-such. In this way they can ‘control’ the crocodile.

Humans are machines too but our behaviour is not quite so predictable when compared with reptiles due to the sophistication of our brains – probably the most complex thing in the entire universe. Our behaviour is predictable though to a large extent, hence all the psychological research into shopping patterns and the insidious harvesting of online data by all the large corporations.

Eventually, our highly sophisticated machine brains will be able to produce highly sophisticated algorithms that will be able to predict what will become ‘viral’ in our social set ups.

Some are arguing that this is nothing to be afraid of, these algorithms are merely more tools that we have invented that will improve our … what?

Here is the crux of the problem; if our brilliant tools mean that humans have no work to do, what is left for humans to do? Do we all become artists? And what if algorithms can produce decent art?

And of course, this argument fails to extrapolate from what is current in our society – out of control secret service agencies, profit motives that don’t consider human welfare, power-mad politicians creating division etc.

Image an algorithm that could accurately predict human behaviour in nearly all circumstances. Sure, you could argue that such knowledge could be beneficial to the human race but all the historical evidence suggests that it would become the most powerful weapon ever devised in human history. Who would you want to have control of that?