Running Commentary

Some years ago, I returned a purchase to Future Shop (the Canadian subsidiary of Best Buy back before they all just became Best Buys). Instead of refunding the purchase, they accidentally charged me a second time for it, which I didn't notice until I checked my bank statement later. So I went back to Future Shop to get now double my money back. I had my receipt, and my bank statement. The clerk tried to find the transaction in his system, but says he couldn't find the return, even though I had the receipt in my hand, singed by the previous clerk. He argued that the transaction never occurred, and said “Computers don't lie”.

I looked him dead in the eye and said “I have a degree in computer science. Don't tell me computers don't lie.”

1s and 0s. On or off. True or false. No ambiguity. Which at the most basic level is absolutely true.

We've been sold the myth that “computers don't lie”. That a system built on unambiguous logic is incapable of lying. But we've built a complex system of symbology on top of that basic binary construct. And the more complex a system, the more ambiguity can creep in.

Computers don't lie. They don't tell the truth either. What they do is mindlessly follow instructions. Garbage in, garbage out.

It's trivial to write a computer program that tells you that you're a cabbage.

print(“You are a cabbage.”)

Yet the myth persists. It's tempting to blame pop culture for perpetuating it, and it's not entirely blameless. But service providers love it, because they can manipulate the algorithm to tell you what they want you to hear. And who are you to argue? Computers don't lie.

Every time your favorite Youtuber gets a video demonetized, age restricted, or taken down, and appeals, only to be told that a human reviewer checked and agreed, they are asking us to believe the myth that “computers don't lie”. Arguably, YouTube want to keep more ad revenue for themselves, and not pay their creators. So they let the algorithm decide. And the algorithm is never wrong, because computers don't lie.

And now we have LLMs. ChatGPT, Grok, Copilot, and the like. We've built a system plagiarizes creative work, sucks up a town's water supply, spews CO₂, and cranks up our power bills just to lie to us. Every customer service front-ended by a chat bot, every call center that answers with an AI agent, are asking us to believe the myth that “computers don't lie”.

Now to be sure, machine learning is a mature field. There are valid uses of it, usually in certain research fields that have to analyze large datasets. And the researchers who use them are theoretically versed enough in their subject matter that they can spot an error when they see one. But these glorified chatbots that they're trying to push as AI assistants and digital agents and cramming into every product whether we want them or not, and generally being a solution in search of a problem, needs to fuck right off.

If it's going to be inescapable anywhere, it'll be in research and software development. Hell, even Linux is using it. And the recent PHP vulnerabilities prove that they will need it, because the hackers are using it to crack things open so fast, that they'll need tools that can close those cracks quickly.

Part of me is concerned about loosing my job to AI, or that my customers will adopt AI to a point that they move on to another product that I don't support. My skillset is highly specialized, and many of the skills I used to have in abundance are stagnated. But part of me dreads the alternative of having to support this crap. I never specialized in it, and I don't care to start. But I may have to. Or when the bubble bursts, and it all comes crashing down, and I have to pick up the pieces and put it all back together.

I hate what tech has become. If I'd known 30 years ago that this is where we'd be… Software as a service, attacks on ownership, and crypto scams. An AI driven subscription service that's driving us to become dependent on the corporations who supply it, and use it to manipulate our opinions and our politics. And the normies who blindly use it, not knowing what it's doing to them and to society. And the politicians who go along with it whether because they're complicit, or don't know any better. And other tech users who are compelled to go along with it, either for the sake of their livelihoods, or because they've bought into the hype. We see the slide into techno-feudalism. It's frustrating to see so many raising the alarm, and nobody doing anything to stop it.

You will own nothing, and they'll tell you that you are happy, and you won't know any different because computers don't lie.

AI lies to you. You don't own AI, their corporations do. And corporations lie to you. They own the AI, and they will manipulate its output to suit their interests at the expense of yours. Anything you put on their cloud or write in their hosted editors belongs to them. They will scan your documents, steal your work, use it to feed their slop machines, and feed it back to you, even if they tell you they won't, and they'll charge you an annual subscription fee to do it.

Computers don't lie. Corporations do.