Technology advancement has rapidly increased during the twenty-first century. With the advent of AI, medical patents, and other research, it’s time to ask the question of how unfettered advancement benefits–or harms–humanity.
It had been months since I visited a bookstore, and it was my first time in a foreign one. I had expected to find a range of unique books unheard of that were more intellectual than what I was used to seeing back in India. However, with a mere glimpse of the science fiction section and the junk it contained, my hopes were immediately dashed. There were even two mothers happily flaunting how their kids had read some of the fat fantasy books that masqueraded as sci-fi. Only one thing reassured me: the general US population did not seem to read any more competent material than Indians.
Disappointed, I had turned to leave when my father called out to me. He asked if I wanted to buy a new book on artificial intelligence by Anil Ananthaswamy. He called it “How Machines Learn.” I corrected my father. Why Machines Learn was the title, not how. That sounded like a good idea. However, the store had not convinced me it could house such an esteemed book. Nevertheless, there was no harm in trying.
It turned out that the book was available. Thirty dollars later, I walked out of the shop with a hardcover clutched to my chest which bubbled with anticipation. I had read two of the author’s previous books on physics. Each of those instances accounted for two of the three total times in my five and a half years of medical school that I had regretted not getting into the engineering field.
However, I wasn’t as excited to read the book as I had been before starting the previous ones. Perhaps it was because I was walking through barely populated restaurants and streets, an environment grossly different from what I was used to in India. But no, it wasn’t loneliness dampening the pleasure of expecting pleasure. My uneasiness led back to what my father had said: “How machines learn.” Why had I corrected him? Other than it being the incorrect title, what was the difference between “how machines learn” and “why machines learn”?
The “how” versus the “why”
One week later, 667 miles away from the heart of Silicon Valley that basked in the warm rays of capitalism, my father and I were rushing past the cold suburbs of Portland, Oregon. We sat in our train’s dining car with various exotic food items on our table. Well, exotic to me.
As we sat, silently crunching on crackers and bread dipped in cream cheese, staring out at the blackishness that seemed to be a river, my father recalled a discussion he’d had with someone from the Bay Area. The person had commented on education being the key to fixing the world. My father had been annoyed. If the answer was so obvious to that man, why wasn’t he in the pursuit of trying to make a difference instead of just complaining?
I felt disinhibited by the non-judgemental environment, a stark contrast to what I had been experiencing in California, so I went a step deeper. What would education change? Current education involves an emotionally detached perspective to the sciences with no philosophical debates regarding its social or even environmental impact. Students are merely exposed to the mechanisms of the sciences. They don’t understand how to empathize. Any good they unexpectedly leave behind is a serendipitous byproduct of a selfish pursuit. We are never exposed to discussions about what good and beneficial science is or, more importantly, the dark history that most scientific advancements are invariably based on.
I sometimes wonder if my nihilism regarding the tech world runs deeper than my father’s.
And then the book clicked into my mind. The reason why I approached it with trepidation was because I couldn’t sense a difference between the questions of how machines learn and why machines learn. And that disturbed me greatly. Would I have realized my thoughts in the environment of manufactured nature and antisociality? I don’t think this would have struck me back in Palo Alto.
I began contemplating the differences between the questions. How does a dog eat? Using its mouth controlled by muscles directed by the brain. Why does a dog eat? Because it feels hunger, a drive originating in the hypothalamus. Seemed straightforward. However, the question grew complicated when I began to consider non-living objects. How does a train move? With motors and wheels. Why does a train move? It seems to have the same answer as the how.
But what about entities where the concept of life was a grey zone? Was artificial intelligence living or nonliving? How do machines learn? I would probably find out that it’s a mixture of computer science and math once I felt comfortable enough to read the book. Why do machines learn? Like before, I came up with the same answer as the how.
Perhaps it was just a problem with the English. To me, how and why led to the same answers regarding non-living objects, but when considering living beings it had significant differences. In the latter case, the how seemed to imply a mechanism while the why explored the emotional and philosophical standpoint behind it.
Weak justifications for technology advancement are everywhere
By extrapolating my earlier statements, this might imply that AI was non-living. However, this was something that neither I nor the world could convince ourselves of with surety, thus opening up a new question. If AI was somewhat living, why did they learn?
AI learned because they were programmed to learn. The how was justifying the why. The very existence of the strange concept of AI was justifying its potential functions. If artificial intelligence did have some degree of intelligence, we didn’t know its reason for performing tasks beyond the fact that it was capable of doing so. The implications of this thought were jarring to me.
Giving AI the benefit of the doubt, let’s say it doesn’t know why it functions. Perhaps because it’s an infant. But as its creator, humans can ask the question: why should AI function?
Because of the controversy surrounding the livingness of AI, the debate of whether or not it’s good for humanity is widely discussed. But what about all the other technologies and scientific advancements which have either had a direct tremendous impact on humanity or have grown on the base of existing ones? How often is the “why should” question asked there?
Before judging him, I once asked a tech-born finance worker about the possibility of travelling to Mars in the future. With the utmost surety I had seen second only to my sure hate for fantasy books, he replied that he would go to Mars the second he got a chance. On asking why, he responded, “Because we’d be able to go to Mars!” I then labelled him a Musk-grunt.
This incident is a small manifestation of something more profound. People justify the use and existence of science simply by reiterating its theoretical possibility or physical reality. Same with AI. The first thought regarding science is never about all the destruction or suffering in its roots, nor is it about carefully considering whether it will benefit the world. But that’s understandable because no one exposes us to think along these lines. Science and technological advancements are based on identifying what can be created and sold, not what should be sold. If it leads to a benefit somewhere along the line, that’s all well and good, but it wasn’t the primary intention. If a technology potentially useful isn’t likely to sell, it has a high chance of never being invested in. Ironically, this is true even in the field of healthcare.
Controversial medicine remains the norm
Two weeks ago, I was rotating in the cancer clinics of one of the best hospitals in the world. I had just talked to a patient, noted his declining health despite being treated with the drug called brentuximab-vedotin, and sympathetically discussed his worsening whole-body PET scan with him. From my understanding of the National Comprehensive Cancer Network (NCCN) guidelines for diffuse large B-cell lymphoma (DLBCL), we had no other alternatives to offer the patient. However, there were always available clinical trials that the Doctors knew about.
I asked my attending doctor about it, and he informed me there was a clinical trial. Curious about his distinct lack of enthusiasm, I asked him about it. Brentuximab-vedotin combines a monoclonal antibody called Brentuximab that attaches to cancer cells and delivers the payload drug vedotin inside the cells. He explained that a linker molecule connects the antibody and the drug, and the pharmaceutical company’s patent for the linker molecule was expiring. The ongoing drug trial contained exactly the same antibody and payload drug but with a newly synthesized linker molecule. The study sought to evaluate whether this new version of brentuximab-vedotin had any progression-free survival or overall survival benefit in DLBCL patients.
Even with an expression that was difficult to read I could tell that my attending had a pessimistic outlook. I stated my understanding of the situation: the drug company had synthesized a new linker molecule and was trying to show a slight benefit through the study. This would effectively give them a new drug with a fresh patent and allow them to set the price. My attending didn’t have high hopes for the new drug because the antibody and the drug that would kill the cancer cells were the same as the old version. He didn’t think there would be a significant benefit due to the new linker molecule. I agreed with him.
Has healthcare benefited humanity? Apparently. However, would the research have been so heavily invested in if it wasn’t to return a profit?
Back in medical school, I worked on a research project about Neglected Tropical Diseases (NTDs). One of the questions I had at the time was regarding the term “neglect.” Who exactly was neglecting it? NTDs like snake bites and dengue were of great concern in India. We weren’t neglecting them. But the Western countries were, thus, birthing the term. The West wasn’t concerned about the NTDs, as their countries were free of these diseases.There was no more money to make off them, hence they had changed their focus to diseases that could be more profitable. If their primary motivation for research was to help humanity, some people should still have been working on it.
Science is based on what can sell, not what should sell. A benefit for humanity, if any, is an illusion used to mask the primary doctrine. I also feel that one of the greatest myths is believing that it is science that improves the world. There is no proof that the earth is overall happier, healthier, more sustainable or morally advanced now compared to before the Industrial Revolution.
This discussion is not to claim that the world would have been better off without science. It is to highlight that we have never once asked the question of “did the world really need some particular sciences?”
The “why should” question is glaringly absent
Education is mechanical. Everything taught answers the questions of “how” and “why”, but the “why should” question is seldom asked, forget discussed. How does a motor work? Everyone knows. But why should a motor even exist? The question seems insignificant. But more than that, we avoid thinking about it because it’s messy. There are a plethora of dirty intricacies: where the materials came from, who mined them, and how they are assembled and eventually used, to name a few. But without this constant introspection, we gradually forget to ask the question altogether, even when things become abundantly controversial. How do oil rigs, the economy, fighter planes and social media platforms work? We learn the how. But why should they work? Did we really need them? Who was consulted each time something came about? Has it benefited humanity or the earth?
The history of science is plagued by tragedies at an individual level and at massive scales. The atom bomb is a blatantly obvious example. But what about the Tuskegee study, where hundreds of Black men were experimented on without consent to understand syphilis; or the doctor Sims, who experimented with some of the first gynaecological surgeries on enslaved women without anaesthesia? The modern world thrives off the invisible sufferings of the past. There is no final verdict of right or wrong; it is merely a discussion that visionaries need to have instead of sweeping it under the rug.
It’s easy to think of science and scientists as pure. It sounds clean, ethical and beautiful. The truth is quite the opposite. Until education and popular books continue to divide the how and why with the why should, people who study science cannot be expected to make a positive contribution to the world because they have no understanding of how their knowledge has affected it previously. Without a background in this, it would be impossible for them to predict how their creations will affect the future.
