I write a column for an Italian magazine called Automazione Oggi. The column is called News from America. The editor and I met in Shanghai at a computer supplier user conference when I was editor at Automation World. We traded columns. I left AW behind 10 years ago. She’s still there. I’m still writing.
This month as I looked for what was happening around the country, I couldn’t escape the conversations around AI and ChatGPT. Most of these thoughts are from the column for AO.
The Deep Question I pose for you today—Have you played with ChatGPT, yet? Do you even know what it is? Have you tried puzzling out all the conflicting conjectures from mainstream media that confuse what AI is (versus machine learning) versus what Artificial General Intelligence (AGI) is? This AI is not even close to the evil overlords found in The Terminator.
Today’s podcast subject from Seth Godin concerns ChatGPT and AI, it’s written by AI, and it’s read by AI. Something to ponder.
But we wonder:
Will anyone human ever write anything anymore? What will become of education? Will humans become slaves? Will this replace all our jobs? I think it is more than an American strategy of news writers and speakers to try to instill as much fear of the future as possible in their audience.
There is little new in these musings that couldn’t have been found in the Terminator movie series beginning in 1984. The theme of humans rebelling against the technological overlords is older than that.
I’ve covered technology advances for industrial and manufacturing applications for 25 years. I’ve observed, learned, used, contemplated technology for much longer than that. We have all seen this theme repeated to ad nauseam. The new technologies never turn out to be as apocryphal as worriers feared. Humans learned how to use each new technology for our benefit. Some have replaced jobs. Most have added other jobs. Yes, some technologies have been used against us (think the aggressive misinformation on social media). On the whole, technology has been our friend.
We have been living with a form of these chat machine learning devices for about 10 years. I’m at my sister-in-law’s winter house in Florida where they ask Alexa such important queries as “what is the temperature outside” or “what is the humidity” or “turn on the living room lights”. Apple brought us Siri.
The difference is that Alexa and Siri will perform certain control tasks. ChatGPT uses statistics. You cannot ask it questions of logic. It is not a decision-making device. You ask it a search question and it is just as likely to return as much wrong information as correct. Google, Bing, and the like return a list of links. You can click each link and decide if that is the information you were looking for. ChatGPT will return a narrative it has gleaned from its machine learning. You don’t know the source. You don’t know if it is correct. If you ask it to write a memo or essay, it will look at a word, then using statistics it will predict the most likely next word, and so on.
Only yesterday media was filled with stories about how we will live in the “metaverse” using virtual reality headsets and augmented reality glasses. I have worn headsets and glasses at many trade fairs including Hanover Messe in Germany for many years. We still have not devised useful applications for them. We have had ML embedded in our software and neural networks in our motion control for many years. Those are forms of AI. Can we use a chat function in industrial applications? I think we are many years away from an answer.
Go make a difference.