How often do many of us now look at a picture and wonder if it's real or the output of an Artificial Intelligence (AI) algorithm? What about an article or a blog? What about a song? What about a person? Yikes! You've taken this too far, Darren (well, maybe for now).
I've wanted to take on the topic of technological evolution for over a year. I just had to find the right time to put pen to paper (so to speak). Well, here we go.
If I'm like most of you, it is becoming increasingly difficult to keep up with all the latest apps and systems that can create, let's call it, "things," which, until the last decade (or two), was the sole domain of humans. Whether it's an AI tool or a 3D printer, the advancements have been fast and furious.
AI has been around in terminology or discussion for over 50 years. Many credit Alan Turing with the initial concept and beginnings of the field. While mathematics and computer science form much of the backbone of AI, they pull concepts and applications from many fields. In Physics, advancements in electronics, thermodynamics, optics, and quantum mechanics all helped to migrate AI from fiction to fact. I'd be remiss if I didn't include the contributions in linguistics, psychology, neuroscience, and others. AI is quite the potpourri of sciences, both concepts and applications. Its complexity is likely approaching (but not yet at) the level of humans. Doesn't this help explain the work and effort (and time) that went into the development of AI?
Despite the long history and work done over the decades, I want to focus this blog on the last few years and, more so, the rest of the century. I can wait another day to evaluate all the books, movies, and articles since the 1970s and 80s to see which ones were more accurate in predicting AI's advancement.
Who hasn't used ChatGPT at this point? Or, at a minimum, Google search (which now feels extremely dated)? I have a whole series of bookmarks for different tools (songs—Udio, admin work—Spinach, presentations—Gamma, personal assistant—Pi), and so on.
What are the results so far? I'm very impressed with the tools out there currently, and to think, they really are the 1st generation. If we consider any technology, such as printing presses, automobiles, airplanes, and computers, how much did any of those advance in 100 years? Substantially, if not exponentially. If the pattern holds, we are at the beginning of this chapter of AI. The question is, really, how much do we (humans) want it to advance? There is no doubt it will.
I've spoken to many people about where they see it wanting to go. It's amusing; most agree with the simple "things." Providing insightful information, even some new content (generative AI), and most will agree that simple tasks are all "OK" for machines to do (smart homes, roomba, basic productivity gains). However, I also hear a lot of "it can't replace humans," "it can't mimic emotions," "it can't teach," "it can't operate," and "it can't replace my job xxxx."
I say, "Nonsense." It will eventually do all that and then some. Not surprisingly, I think the answers are quite different from what humanity would prefer AI and technology to do compared to what it will be able to do. Scientists (or, more likely, business people) will see to that. Regular people, including ethicists, will have more of a back seat. Change can be uncomfortable, but that doesn't mean it shouldn't occur. A fantastic article in the Stanford Medical Magazine discusses where the future might go. Can AI ever best human brain’s intellectual capability?
Returning to all the reservations and limitations people see enforceable on technology. The questions I have for those people are really two-fold. 1) Why do we think our skills, roles, and, yes, even personalities, are unique and not replicable? 2) How will we place some type of control on the evolution and abilities of AI? I'm thinking Ultron (a Marvel Universe reference, for those unsure).
Regarding the first question, I guess I shouldn't be surprised. The word 'protectionism' has been around for a long time. It is typically used to represent how a country/province would protect its economy against that of others. In this case, the economic entity could be a product (typically a raw material), a type of worker (again, often associated with raw material), or sometimes an entire industry (think Canada and the automotive sector).
Now, why do I often hear from people that AI won't replace their specific jobs? Are they worried about losing their job (it's still decades away for most)? Is it that AI won't do the job as well (the truth is, in most cases, they will do it better - faster and higher quality)? Do people not want to interact with a machine (how many felt the same way with ATMs and cell phones)? It's human protectionism (based on Maslow's hierarchy of needs) at its finest.
I'm sure the answer is all of the above. I don't know what the future holds. Still, I realize that AI, once developed to its ~fifth generation, will be able to teach, operate, manufacture, and service as well, if not consistently better, than humanity. We will teach them (yes, I'm sure AI machines will have names), and they will self-learn emotion through tone and facial expressions; heck, they might be able to see brain thoughts happening through some advanced technology. Being able to do physical processes better than humans will be a no-brainer in my opinion, it is the emotional ones that I agree, will be towards the latter phase of AI development.
I'll return to the 1st rebuttal, namely, "I'll lose my job." Do people wonder what is on the other side of that? I wonder about the possibilities for humanity if we are less constrained by "work." I know I'm jumping the gun, but what does life look like? Do all the basic necessities become part of life (a commodity) that doesn't require money? Is there some type of Eden awaiting us? If the future AI can drive us, shelter us, feed us, how much money do we need? I know, I know, I'm well into the future, Blade Runner meets The Jetsons.
But it is getting closer.
I'll wrap up this topic in my next blog when I discuss question 2: How will we (or should we) control AI?
Comments