I’m hearing people sing the praises of “AI” services like ChatCGP and Grok2. Tech companies, including my own employer, are pushing this idea that their products contain “artificial ingelligence”. It’s the new marketing buzzword.
I’m a systems engineer and I help manage hundreds of servers. They’re nothing inherently intelligent about any of them. They’re just the same, dumb, binary-speaking machines we’ve always had. What makes them seem intelligent is their ability to understand human language and quickly reference a vast database of knowledge. They can do this now because processors have gotten fast enough and storage capacity has become large enough to support complicated algorithms that answer our questions. To be fair, we’ve had this capability since the search engine AskJeeves hit the web around 2000. The engineers behind it have had plenty of time to improve the algorithms.
The knowledge database is built by humans. The algorithms are programmed by humans. What’s an algorithm? “If a = x, do y; else do z”. That’s a very simple one, but it illustrates what’s happening behind the scenes. The computer parses what you type, executes logic as it tries to correlate your input with its database, and then it creates output, giving you an answer to your question. I use Grok2 and it’s pretty awesome but it’s not magic and it’s not alive.
There’s nothing intelligent happening here. It’s just computers being computers. Machine learning is something a little different, and it’s pretty cool actually. A machine is programmed to analyze and store new data. That’s no easy task, but it’s not about to take over the world unless a human programs it to do so. Computers will never be sentient until man learns how to create new life. We’re not Gods, but I suppose that’s not going to stop anyone from trying to be.