A lost computer

We used to have this amazing thing that we called a computer. A computer is a device that allows one to gather, process, store, and transfer information from point A to point B.

(Just an AI-generated space-filling image)

The main advantage of a computer is that the information processed by it is secure, processed quickly, and always precise. The information we transfer from computer to computer never changes, and that allowed us to reach a new level of development across various fields.

Our financial records no longer have to be a constant mess. It’s very easy to match a credit card transaction to your balance sheet because we have computers and everything is done electronically. Many companies now use electronic document processing, and every time you need a new contract with someone — or something like that — you just sign it on #DocuSign. No need to fax or mail anything anymore.

Computers are cool because they are precise, fast, can process enormous amounts of data in a very short time, and can transfer information from point A to point B at nearly the speed of light — and do it with precision.

Now, Hail AI

AI is our future. AI is a computer system that, despite being a computer system, lacks one of the most fundamental principles of one: it cannot communicate a predictable answer. The output from any AI system requires verification and correction.

I remember how, in the early 2000s, my boss would tell me, “I don’t trust this data — check it with a computer.” Now, we’ve turned that upside down. We say, “I don’t trust this data. It must have been generated by #AI.”

Our Basic Purpose

So it seems that in this amazing age of quote-unquote “advancement”, we humans haven’t changed our basic purpose. We are still here to say the one thing no computer system can ever truly say: “This is correct.”

A computer system can’t say “This is correct”. It has no sense of responsibility. If you don’t believe me, try telling ChatGPT it failed a task and ask it to take the blame. The only quote-unquote “responsibility” you’ll get is: “Oh, I totally understand. I’m so sorry. Let me try again.”

It’s one thing to be sorry when you generated an image of a kitten instead of a dog — who cares? But it’s a totally different thing to be sorry when you’re responsible for building a plane, and that plane is supposed to fly. What if someone forgets to double-check a measurement and now the plane is crashing to the ground? “Oops, I’m sorry” isn’t going to bring anyone back from the dead.

There needs to be someone who steps up to resolve that situation. Someone who takes responsibility and says: “This is correct. We have verified this, and we are absolutely sure this plane is not going to crash. We are ready to put our reputations — or our lives — on the line for it.” That’s how certainty works.

So yes, people are talking about AI taking jobs or automating tasks. But I don’t think that’s the biggest problem. The real danger is this: AI is eroding our ability to be #responsible for things.