A few days ago, I was having a conversation with some coworkers and the subject of A.I. came up. Currently, I work in academia and my colleagues were expressing concern about the use of A.I. in schoolwork, specifically around how they have already come across papers and assignments done by A.I. systems. Conversation continued over the concerns of this emergent technology (that has actually been around for awhile) until eventually someone mentioned how they are worried about how A.I. will pretty much takeover all industries and creative outputs because of an article on NPR that interviewed a former Google employee who was growing concerned that A.I. was becoming sentient and was soon going to reach an inflection point of no return. My colleague was, understandably, frightened by this information.
Now, they are, admittedly, an older individual, but still highly educated with seemingly active critical thinking skills, and even they seemed to have no real concept of what A.I. actually is or does. This should be concerning. Not the use or proliferation of the technology, but how poorly it is understood in its use, conception, and actual output. It is that ignorance and misinformation that will allow the rich and powerful who are trying to sell the idea and future of A.I. to be somewhat successful in their machinations. That is why NPR and Wired and The Times and so many other publications have had so many articles about A.I. It is why Musk and the Google guy and other tech bros are now stating concerns over A.I. while still having deep financial interests in the technology and companies pushing it. It is why the AMPTP is pushing for the use of A.I. across the many creative fields in Hollywood to be in contracts and negotiations moving forward. Because for all the pomp and circumstance and promise, all A.I. is the current popular tool to circumvent workers and audience by exploiting other workers and audience to make more money for the richest of the rich while taking away more from the dwindling middle, worker, and poor classes. And it is not even remotely doing what it promises; much less doing it well.
One of the major misunderstandings is that A.I. is in its current form actually sentient. It is not and will never actually be. Sentience would require original thought. A.I. is incapable of that. In its current forms, it is basically a search engine on steroids with those steroids mostly being poorly paid and exploited workers in the global South being asked to sit at a computer for hours a day and process the requests being made to the A.I. systems. These systems just have a more specific database that are being used compared to the entirety of the internet like more popular search engines use. The constant data entry from the exploited workers also accounts for the speed of the systems developments. This still does not result in actual sentience. It is just a superpowered search engine with some extra writing processor tools.
In essence, imagine training (by utter theft by the way because that is how A.I. models really work) a current A.I. system on nothing but Stephen King’s fiction. Not everything he has ever written but just his works of fiction. This computer system now has every work of fiction the man has ever written. Can it produce a work equal to Stephen King? No. This isn’t me trying to be glib or anything. The answer is literally no. For one, the A.I. system can only respond, so for it to produce anything, someone has to type an input to get anything out of it. And even then it would be an approximation of King’s craft. At most it could provide an imitation of a story because the input would be some douchenozzle typing into a line with like a prompt that states “write a story about a cat” or something. It would give you a mish mash of King’s works relating to cats and style and text without any soul or substance. However, it is still just a response machine. On the other hand, Stephen King is constantly writing and making deeper connections from his observations and experiences to create new works and ideas that an A.I. could never replicate or perform.
Basically, an A.I. system would look at a blank page forever and ever without producing anything until told to while J. R. R. Tolkien looked at a blank page and created an entire new mythos. That is the difference between human writers and A.I. Creation versus regurgitation. And that is what is at stake: a future of interesting new stories from various minds and voices or a continuous conveyor belt of the same shit over and over again consistently diminishing in quality and intrigue. Hopefully, the choice is obvious.