AI is undeniably the hottest thing in tech right now. It's reached the point that I can't go more than a few hours without hearing about it, and chances are, you can't either. All the big tech firms seem desperate to shoehorn some AI advancement into new and old products, and AI fatigue is already beginning to show. Microsoft might already be causing Copilot burnout with its frantic pace to incorporate into systems before it's fully developed, but that's just how capitalism is played. You want to be first, and you want to remain in first place forever.

But we've never played this game with a technology so capable of disruption in so many areas of human life. AI's potential is seemingly limitless, and nothing seems to be able to slow the investment bubble that, if continued, will surely eclipse almost any other tech bubble we've witnessed. It only makes sense that major tech firms like Microsoft, Google, and Meta are leading the way with partnerships and in-house development; they have a lot of money to spend and are bound to benefit greatly from the technology.

I'm not arguing that AI shouldn't exist or its development shouldn't continue. On the contrary, it's one of the most exciting technologies I've ever seen, and it really is part of the utopic future that many of us have envisioned. However, watching AI's power become centralized by giant corporations doesn't exactly bode well for the general public's future. Where money can be made, ethics and morals often get pushed aside.

Generative AI didn't appear overnight

Though it feels like it did

Generative AI is quickly becoming mainstream, and it might seem like it came out of nowhere. It didn't. The application of Generative AI, at least with its current capabilities, was even hinted at by Canadian philosopher Marshall McLuhan in the 1960s. McLuhan, who focused his career on media theory, foresaw a time when individuals could request wide swaths of information and promptly deliver it in a neat package. In an interview with Robert Fulford on This Hour Has Seven Days, McLuhan said, "Products are increasingly becoming services" in an attempt to explain where we were headed all those years ago. Of course, McLuhan didn't foresee the internet or its ability to deliver this information in such a timely manner, and he certainly did not predict the massive stores of data with which AI could be trained. He lived in a time of libraries and Xerox machines.

Generative AI in the form of popular systems like GPT, DALL-E, and Bard didn't pop up overnight. Not only did the required hardware for such heavy computing not yet exist so many years ago, but the massive hoards of data on which Generative AI has now been trained also didn't exist. You know, the data we didn't know could be so powerful and profitable until it was too late to take back any power from the corporations collecting it.

Geoffrey Hinton, a computer scientist and cognitive psychologist, was a leading force in developing deep learning techniques in the 1980s as they are used today. To explain it simply, he helped shape the idea that instead of teaching a computer what to think, we could show a computer enough data to let it form its own conclusions, largely from recognizing patterns. He later joined Google in 2013 to help build Google Brain, but exited after 10 years to begin warning the public of the dangers of AI.

Not only would AI begin to replace humans, argued Hinton, but it would also be a perfect tool to manipulate humans without their knowing. Sounds familiar, doesn't it? And while massive controlling firms have pledged to stick to self-imposed ethics and morality surrounding AI, the bottom line in a capitalist system is remaining competitive and demonstrating growth year after year. When ethics and morals get in the way of boosting profits, history shows a less-than-flawless record.

The next step in AI development

AGI is somewhere in our future

Generative AI is a stop along the way to true artificial intelligence, otherwise known as Artificial General Intelligence (AGI). AGI differs from regular AI primarily in its ability to reason. While the likes of ChatGPT (which powers Microsoft's Copilot) and Google Bard rely on recognizing patterns in their supplied datasets, AGI would be able to understand complex concepts and, essentially, answer questions and apply reasoning on its own. My colleague Adam Conway wrote a great article explaining OpenAI's Q* algorithm and its ability to "threaten humanity," a quote from OpenAI's former chief scientist Ilya Sutskever.

This is the version of AI that a lot of people have always envisioned: something so smart that we can only tell it apart from human intelligence due to it being so far beyond our organic capabilities. It's unclear what actually is going on behind closed doors at OpenAI. Still, there's at least some evidence — like OpenAI's CEO Sam Altman being ousted only to quickly return — that the company is beginning to tap into something we have no idea how to control. I understand that development has to happen somewhere, but I don't know if it's currently in the right place.

Being at the mercy of massive tech firms when it comes to our data alone has already resulted in harm to society. I don't think it's a big secret that our data is being used against us to farm engagement (anger and outrage are also so hot right now), sway political and societal decisions, and sell us things we don't really need. Even more frightening is that it's being used to predict how we'll act on the individual and societal scale; how this information is often being applied leads down a path of authoritarianism and predictive programming despite being framed as necessary for a better future.

Our data created simply by using the internet has now been used to create arguably the most powerful tool of the 21st century, set to disrupt our world on a scale similar to the Industrial Revolution when it really kicked off. Massive industries employing countless people fear being completely replaced by AI, educational institutions are almost powerless against rampant misuse, and we haven't even begun to scratch the surface of ethical dilemmas like deepfakes and scams. Just look at the current hubbub surrounding explicit Taylor Swift AI fakes and the reactionary rush to regulate these tools somehow. These problems are not new, but they're being reframed in a way we've never faced before.

What does AI look like in 10 years?

There's a lot to think about

We're at the point where a lot of money is being made simply by putting "AI" in the name of a new product, but that will quickly dry up as AI fatigue settles into consumer trends. It can be argued that Microsoft still has no idea what it's doing with Copilot, and it's only getting started with relatively basic tools (at least compared to what's coming in the future).

We're only dipping our toes in the massive ocean of AI potential, and already, we're confronted with subscription costs, paywalls, and competitive uncertainty. In many industries, AI-based tools are already too good not to use if you want to be competitive, and that will no doubt worsen the wealth and knowledge gap already yawning in 2024. AI PCs — those with a built-in Neural Processing Unit (NPU) to handle AI-related tasks — are in their infancy, but they're expected to raise system requirements (and thus costs) as they progress.

Looking for AI answers

Not as easy as a simple prompt

I don't know who or what should be in charge of AI or if that's even possible when artificial intelligence reaches a certain point. Even more frightening is the belief that no person or institution (including world governments) has any idea what will happen tomorrow, next week, or next year, never mind having any idea how to protect the rest of us from unforeseen dilemmas. Playing catch-up and only reacting to problems with something that moves as quickly as AI isn't going to work out well.

We're off to a misguided start, but not an unsurprising start. Continuing with the current momentum, I foresee a market where those with the most powerful AI (now only available at a very high cost) are years or decades ahead of those without AI assistance, further centralizing power into the hands of a very few unelected and largely unsupervised groups. This structure seems like a capitalist's dream, but it's a sleepless night for the rest of us.

AI is here to stay. I hope we don't get pushed aside as it grows.