By Ian Duckles
As an educator and a philosopher with an interest in sci-fi and futurism, I follow debates and discussions around the emergence of generative AI with great interest.
I have attended many workshops and seminars on the subject, and it is with increasing dismay that I grow weary of the forced optimism I am presented with. I am told how AI will simplify my workload and remove the drudgery and repetition from my work; how it will automate all those troublesome tasks like reading and responding to email; how it can do my grading for me and produce incredible analytics to track me and my students’ performance; how it can summarize articles and meetings for me and save everything in an easily searchable format so all the information I need will be ready at hand on demand.
And I’m sick of it.
While I recognize the incredible power and benefits that AI can provide, people really need to take a more realistic attitude towards the impacts it will have on our lives and our professions. Typically, the workshops I’ve attended do have a slide about potential negatives, but in an hour-long session cheerleading for AI, this discussion of negatives may take up five minutes.
As a corrective, I offer the following concerns that we, as a society, need to grapple with and take far more seriously than we do, and I would like to see more attention devoted to these topics. This list is by no means exhaustive, but I think it is a good starting point.
First, AI is likely the end of truth in contemporary discourse. In the wake of the Trump presidency, many have commented on our “post-truth” environment, but AI will only accelerate this to an unprecedented degree.
Essentially, we are reaching (or perhaps have already reached) a point where one can no longer trust any image, recording, or video they are presented with. While image, recording, and video manipulation have existed as long as these media, it typically took time and expertise to successfully generate a convincing fake. No more. Now anybody with access to the Internet can generate deep fakes that are essentially indistinguishable from authentic ones.
In addition, in this post-truth environment, anyone who is caught on a hot-mic or who is photographed or videoed in a compromising position can assert that the evidence is AI generated, and there is no way for anyone to adjudicate the truth of these claims.
I dread the amount of shit that will flood the political environment in the upcoming election as political players domestic and foreign seek to use AI to influence the outcome of the upcoming election (something we have already seen with the Joe Biden deepfake robocall in the New Hampshire primary).
Similarly, in my classrooms, if I don’t see students physically writing on a piece of paper with a pen or pencil, I can’t be confident that the work they are submitting is their own. Again, this was true before, but it took a great deal of time and energy (and probably money) to generate acceptable writing. Now ChatGPT will do it for free in seconds. And generative AI is only going to get faster, more powerful, and more customizable.
Finally (for now, but by no means the end of this list of concerns), I think we need to seriously consider how AI will be used to take away our jobs.
In a recent workshop I attended, one of the presenters was extolling how AI could help one learn Spanish, or even obviate the need to learn a new language because the AI could (in the near future) provide real time translation. This was presented as a positive, but all I could think was “whelp, I guess we don’t need language instructors anymore.”
Another area where I see significant issues is in writing instruction. I think a very real sentiment that we are starting to see is people asking why we need to teach writing when AI can do it for us. All we humans need to do is just come up with the right prompt and the AI will do the rest for us.
This aspect in particular is of significant concern because I believe the act of writing is an important element in critical thinking as it forces one to organize their thoughts and consider the most effective way to present and evaluate ideas. If writing instruction devolves to teaching students how to write AI prompts, our students will lose out on acquiring these important skills that are necessary for leading a fulfilling and enriching life.
AI is an incredible tool for generating massive amounts of bullshit, and we are going to use it to eliminate teaching our students how to recognize and wade through all the bullshit they are going to encounter.
AI is likely the most significant technological advancement in human history since the invention of the Internet (and possibly more significant even than that). It will radically reshape all aspects of human existence. Some of these changes may be beneficial (though the fact that AI is developed and promoted primarily by capitalists makes me skeptical), but many of these changes will be massively disruptive and extremely negative.
As we think about AI, I would like to see a healthy dose of skepticism and perhaps cynicism as a counter to the cheerleading of the techno-optimists.
***
Ian Duckles teaches philosophy at San Diego Mesa College and is an active member of the union there (AFT Guild 1931) that represents most of the workers.