By Eliyahu Solomon
I am almost certain that 99% of those reading this are familiar with Artificial Intelligence (AI), specifically ChatGPT. Whether it is used in the classroom, office, or for recreational use, AI promises to be the next frontier of productivity and innovation. It writes essays in seconds on the most convoluted topics and can even make broad connections within two almost completely unrelated themes. This is just the beginning of what AI can do
Of course, a lot of my recent conversations have been centered around AI, and one common theme has emerged: AI will replace our jobs, no matter what we do. Such a topic came up during chavrusa time in shiur (after we had finished all our learning, of course). The goal of our conversation was to find potential jobs that would not be stripped away due to AI, and some positions that came to mind were that of teacher/Rabbi, social workers such as therapists and psychologists, and a salesman in a marketplace like the Mahne Yehuda Shuk. While I am sure we could have thought of more if we were not engrossed in our learning, there seems to be a common thread between all these jobs: they are all interpersonal and require human contact.
This got me thinking. Are all other jobs aside from these really devoid of human interaction and intelligence? Can all our professions be taken over by a bot? While I am willing to believe that many manual labor jobs can be eventually replaced by “burger-flipping robots,” I am far more hesitant to abide by the claim that many other jobs can be replaced by AI.
The real question posed by AI is, what is so unique about human intelligence that a computer or robot cannot replace it? If the answer is “not much”, then we may see ourselves in some serious trouble. Of course, human intelligence is hard to quantify, it involves the most complex machine known to the universe- the human brain, and even after centuries of study, it largely remains a mystery. An individual’s brain, from the time of birth, is always forming and changing, responding to outside stimuli and adapting to certain circumstances.
The goal of AI is to replicate human intelligence in order to automate certain tasks, interact and converse with humans, as well as guide us to see or think of connections or ideas we may not have put together. In chapter one of The Book of Why: The New Science of Cause and Effect by computer scientist Judea Pearl and writer Dana Mackenzie, the argument is made that AI has failed to live up to these lofty expectations:
What we have gotten from deep learning instead is machines with abilities—truly impressive abilities—but no intelligence. The difference is profound and lies in the absence of a model of reality. Just as they did 30 years ago, machine-learning programs (including those with deep neural networks) operate almost entirely in an associational mode. They are driven by a stream of observations in which they attempt to fit a function, in much the same way that a statistician tries to fit a line to a collection of points. Deep neural networks have added many more layers to the complexity of the fitted function but still, what drives the fitting process is raw data. They continue to improve in accuracy as more data are fitted, but they do not benefit from the “super-evolutionary speedup” that we encountered above. They end up with a brittle, special-purpose system that is inscrutable even to its programmers.
Simply put, AI is good at receiving what it already has, far better than humans in most cases. AI bots in games such as chess have far outpaced the best human players on Earth and even solved complex jigsaw puzzles. Whatever the point is, AI seems to be doing pretty well when it is given all information, although there are instances where AI has put out wrong or misleading information. Even if AI would be perfect, can we really pass the torch off to AI? Humans, as previously stated, are creatures that evolve. We innovate, learn, and interact with one another in ways that AI cannot. So while AI can write music, draft essays, and generate art, it is only within the given parameters of human invention. AI cannot create a new art form, cannot break the boundaries of conventional writing, and certainly cannot have the same interactions that humans have.
In the business sphere, will AI be a replacement for those who interact with the public on a daily basis? Realistically, it could replace a basic customer service rep, but for an expertise such as marketing and advertising, which is an area heavily connected to human psychology, can we trust AI? After all, the amount of ads that have received heavy backlash proves that we humans have not fully mastered this area yet. Are we ready to hand it off to AI?
ChatGPT, is good… for an AI. The number of times I have asked AI to write me a creative writing piece or a cover letter only to be utterly disappointed by its generic output has been too many to count. Supporters of ChatGPT would tell me that I am not implementing the right commands or that I should fine-tune it more, but none of this has worked. AI works well within its conventions, but to break those, to truly write a groundbreaking novel or write the script of an Oscar-winning movie, is just not within the bounds of AI.
This is not to suggest that AI is useless and will pass as a fad. I truly believe that AI will be implemented in many jobs and professions. This piece by Ian Bogost puts it best:
“OpenAI assumes that its work is fated to evolve into an artificial general intelligence—a machine that can do anything. Instead, we should adopt a less ambitious but more likely goal for ChatGPT and its successors: They offer an interface into the textual infinity of digitized life, an otherwise impenetrable space that few humans can use effectively in the present.”
AI can dig up information and make unknown topics to us relatively accessible. Yet the ingenuity of humans and our analytical prowess, the ability to take a concept and broaden our knowledge of it, or to take numbers and not just understand them but to apply them to a broader context, is something that only true intelligence can do.