This world is full of technology that helps and protects us. Trouble is, in a lot cases this equates to "pisses us off".

Take my bank's fraud protection system, for instance. A number of times over the last few years I've had debit card transactions declined because the bank's system has decided they might be fraudulent - mainly because I've been overseas at the time.

Then consider the engine management system of my Renault. It developed a misfire the other day, and ran as rough as hell - and when I say "rough", I mean it - we're talking about a V6 firing on five cylinders (which ought to be pretty bearable) sounding rougher than my old Allegro firing on two-and-a-bit.

Then there's Sky's broadband support line. My friend had a problem with her ADSL, and called the support people. The problem was "obviously" a fault on the line, and it got escalated to third-line support. I went to her house, took one look, proclaimed the router dead (the line was fine) and called Sky support - only to be told that because the problem was with their third-line support people, the first-line people couldn't send us a new router (as they would if it was a new call). Instead, we'd have to wait for 3-5 working days for the third-line support people to call us back.

All these incidents suffer from what I've christened "binary syndrome": some kind of rudimentary test is done on the input data, and the result is either one thing or the other. There's no middle ground.

Take the bank example. Their excuse, whenever I phone them to ask them to unblock my card yet again, is that their fraud prevention system is there for my protection. No: it's a bloody pointless excuse for them to pretend they're TRYING to protect me. Why do I think this? Well, in some cases, only one of my cards (my joint account debit card) seems to get blocked. I can often simply present a different debit card (same bank) and it goes through fine. Also, it's usually true that when my joint account card gets blocked, my partner's card for the same account is fine. Hint to fraudsters: don't bother skimming individual cards - just mug people for their whole wallet, cos you'll eventually find a card in there that the bank doesn't block. Oh, and this wonderful fraud prevention system is also stupid enough to block the same transaction on an annual basis - so the last twice I've tried to renew my MSDN membership, the card has been blocked even though there have been equivalent transactions to the same merchant (Microsoft) at almost precisely 12-month intervals.

Now the car example. The AA guy informed me that my car runs rough because there's a computer in there which, when it detects a misfire, simply cuts off the fuel supply to that cylinder. So whereas in the old days you'd have the occasional blip because of the minor electrical fault that had developed, you instead get that cylinder missing on EVERY turn of the crankshaft. So a minor fault turns into a major inconvenience - by design!

Then the Sky example. The process has been written such that if the first- and second-line people misdiagnose the fault, and it gets to the third-line guys, there's no way to back out. Eventually, through sheer persistence (i.e. hanging up and redialling until I got someone helpful) I persuaded the first-line people to pretend the call never existed, and they agreed that the router was broken and that they'd send a new one out. Took a fortnight.

Why do systems have to be so damned black-and-white? Answer: they don't - it's just cheaper to do it like that. Because to introduce grey areas, you need either (a) to introduce real people with brains who can consider things on their merit; or (b) to develop heuristic software that makes a half-sensible decision based on some intricate ruleset. These are both expensive, of course - the result being that we're stuck with the nonsense that is today's "yes/no" automated decision process.