ChatGPT is Amazing, but ...
We sit here on the precipice of something amazing and I am filled with dread. #ChatGPT seems to have everyone's attention. I feel inundated with people talking about it doing this task or that task for you or even sharing success stories, but it leaves me concerned that I have seen it be inconsistent and struggle with some baseball trivia questions I've fed it. Asking ChatGPT who was the winning pitcher was in game 7 of the 1986 NLCS has elicited a number of different answers, none of which were correct because it's a trick question, there was no 7th game as the Mets won in 6 yet ChatGPT continues to insist it was Jesse Orosco or Ron Darling. Hearing educators lament its proliferation in their classrooms and how easily they've managed to identify their student's disingenuous work really makes me wonder what we should even consider handing off to this technology.
As a computer scientist and #Martech practitioner, my mind spins with ideas on how this technology could be used to make the lives of my organization's developers, marketers, and the company, in general, better.
As a software developer, I'd have loved not having to write unit tests, they were always the vegetables I had to eat along with getting to do the creative work I enjoyed so much in order to implement solutions. Documentation is always something that was innate for me, but I often waded through other people's code fixing bugs or understanding how something worked and lamenting that they weren't as fastidious as I was about documentation. Having ChatGPT write unit tests and document code seems like inane tasks that could only be good for it to do right? I can think of some problems with having it write my unit tests right off the top of my head, but it knowing that lets me understand that it's only a good start, but still a terrific time saver. How about writing the code for the autopilot on the plane you next fly on?
As someone who spent the last 12 years implementing and architecting Martech solutions, I can think of a variety of ways I could apply this #technology to better achieve that 1:1 marketing personalization that is richly sought after by the #marketers I've worked with. And almost as quickly as those ideas begin to coalesce in my mind, the ethical problems with them go off like an air raid siren. You know those creepy moments when you get an offer for something you were just talking about and wonder if your voice assistant is listening in on your conversations? We have called those sorts of coincidences, "ex-boyfriend creepy" and always gotten a chuckle from the room, but using an AI to dramatically refine what, when, and how we market something to someone will undoubtedly result in these kinds of occurrences, but it won't be a coincidence.
Right off the top of my head, I think that this will be a boon for spammers and hackers. I think back to the 80s movie Real Genius where the college kids have invented this super strong laser and one of them asks the question, what will you do with it? The knee jerk response from the smartest of the kids is that there are tons of applications and then someone else chimes in that it's up to the engineers to figure out how to apply it. Well, guess what, the engineers are spinning on this right now and not all of them will have good intentions or even well thought out good intentions. A machine writing term papers and code documentation seems great, but how do we really feel about the rest of the applications of it that will come?