ChatGPT, Friend or Foe?

ACORN Future Tech

Marble Falls     I have been reading about this supposed breakthrough in artificial intelligence, A.I. to the cognoscenti, with something called ChatGPT, which stands for Chat Generative Pre-trained Transformer.  It’s early days, but the reactions have been across the board from open, loving embrace to fears that it is the end of education and learning as we know it and might someday take over the world.

All of which made me curious, with a huge caveat.  I am not a bleeding edge adopter of new technology, or even a leading-edge adopter.  I’m more of a trailing edge guy and maybe even very close to an over-the-edge guy when it comes to new technology.  I probably still stand guilty of basically using a computer as a typewriter on steroids.  In fact, when I would visit my daughter when she was living in Brooklyn, I would be assigned to the far reaches of the apartment, because I hit the keys too hard and loudly to this date.  More recently, I will be forever grateful to one of her friends who worked with us as a communications person three or four years ago for finally teaching me how to embed a web link in things I was writing on a daily basis.  I’m educable with some serious reservations, as you can see.

After a family reunion, when my gang put together a list of things to do before returning home, I added getting my daughter, a Chat-fan, to show me some of the rudiments of how this newfangled thing worked.  Spoiler alert, it’s pretty impressive, but as it’s inventor, the OpenAI website disclaimer made clear, “ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers,” which is hardly a full-throated endorsement. I wasn’t surprised.  I had seen an effort by ACORN’s communications director to try it out on a history of Local 100, which yielded only one or two paragraphs that were even in the ballpark.

Is it useful, though.  It will check grammar and spelling, which is good.  You may get some gook out of it, but it will produce simple outlines that can give you a structure to adapt to your purpose.  For example, you could put in an old proposal for funding and ask it to update along several specific lines, and it can give you a time-saving head start.  I can see why Google has its techies working overtime to catch up, because it can be used as a search tool in many ways.

At this point, the dangers cited by the Cassandras don’t seem immediate.   Many students may try to crank out term papers with AI, but most teachers will be able to catch it, unless the adaptations by the student are huge, in which case, that’s fair game because it’s a different animal entirely.  Some credit should also go to those who get superb at using the tool, because there will be significant skill in determining what to upload and how to set the instructions and parameters to ChatGPT.  In all likelihood, that will not be me and you, as I’ve already conceded.  I’m not sure that you and I are even that good at basic search to tell the truth.

That’s not to say there aren’t dangers to A.I.  Elon Musk is one of the people behind this thing right now, and as we all know, he might turn it against all of us and weaponize it at some point on little more than whim.  Keep an eye on all of this, but don’t panic, at least not yet.