“Skin up, f**k off, Dave’s a fat b******d.” These were the piping melodic sounds that greeted us slumbering students as we awoke in the house I once shared with my brother and his budgerigar, Blue.

He had, of course, trained the tiny creature to utter inane profanities from dawn to dusk.

We got used to little Blue, even learned to enjoy his little Tourettic enthusiasm, and of course, never did we blame the bird for the views it expressed. It was just mimicking the sounds it heard. It wasn’t bad behaviour, but simply an ethological urge to fit in.

This is what artificial intelligence is like right now. Someone, somewhere, has asked a machine to pretend to get along with us, and so it has listened very hard to what we’ve all been saying. Now it is chirping nonsensical sounds, and squirting words all over whomever is foolhardy enough to trust it with writing a homework assignment.

It is just as well, that there are so few of these broad language-based learning models out there. You still have to queue for the free version of ChatGPT, last I checked. But imagine when these things become ubiquitous, everywhere, computers will be squawking and squealing anything that their rudimentary minds tell them people may wish to hear. The internet, and by extension the world, will be full of the idiot babbling of the silicone equivalent of a billion budgerigars.

Mind you, we may not notice the difference.

^{The million dollar shopping trolly}

There is such a thing as a Turing Test. To pass this, a machine must fool a human into thinking it might be another human. That sounds well and good, but for the sneaking suspicion I have that most humans would fail this same test. Lately, the overseers of our banks, our police, our health service, all seem to have a slavish, nonsensical, robotic devotion to policy and procedure, without showing any common sense, or much sign of thinking for themselves.

Artificial intelligence seems to be on the rise, but it isn’t. Humanity is just in decline.

Soon, they’ll be a reverse Turing Test, when a human has to pretend to be a machine in order for an AI to take the input seriously. Google will show us a grid of pictures and ask us to prove we are a robot, not the other way around.

A spokesperson for the Church last week suggested that we need not fear AI because it will never be able to talk to God. I pray he’s right. We pretend we’re worthy enough for Him to listen and take an interest. Whether AI can do this or not is beside the point – self-delusion is our job, not theirs.

Budgerigar photo by Jaqueline Smith on Pexels.com. Other Images (C) David J Harrison.