Outsourcing Our Thinking to AI

It’s an undeniable fact that we are currently experiencing a very rapid adoption of AI in the workplace. Seventy-two percent of companies are already incorporating AI in one business capacity at the very least [1]. With companies such as OpenAI, Google, and Microsoft allowing for the widespread proliferation of AI, the only thing stopping someone from using a new AI system is an account – and if one wants a more advanced AI system, a little money for a subscription fee. With this rush to adapt, many future issues are being significantly downplayed or ignored by those implementing or creating AI systems. It’s clear that AI will have devastating social, environmental, and political impacts but we must also look at how AI will affect the individual. More specifically, how AI will affect our ability to think critically. Critical thinking can be defined as: our ability to recall information, comprehend information, apply concepts, synthesize, analyze, and evaluate ideas [2]. Currently, proponents of AI are preaching that AI will increase our creativity [3] and deliver us into a ‘Knowledge Age’ [4] without realizing how absurd these claims are in the face of trends that are rapidly presenting themselves.

A Brief History of Decline & Dependence

Before continuing, we must first discuss what self-propagating systems are. These are groups comprised of individuals that promote their own growth and survival. These can vary drastically in size (both a nuclear family and a modern government count as self-propagating systems) but for the purposes of this article, we are more concerned with companies and governments. One aspect shared by every self-propagating system is that they face constant competition from other self-propagating systems and are locked in a constant struggle where only the most ruthless and efficient can succeed in staying ahead of their competition [5]. This means that they must use every technique and resource they can get their hands on.

Since the start of the First Industrial Revolution, there has been an accelerating trend of mechanization that has made itself present in almost every economic niche. In the mid-18th century, garments were made by hand but by the mid-19th century, almost everyone wore something made in a large textile mill [6]. By the start of the 20th century, timber was cut by axes and saws but by the end of the century, large machines that did as much work as a dozen loggers, were roaming the forests [7]. With this mechanization, we replaced our physical dexterity and strength for machines not out of some lazy inclination but to maximize efficiency. This has given those who embraced new forms of mechanization the ability to stay in the market or dominate more of it. And governments that adopted new forms of mechanization, were able to garner more power for themselves. However, in the process, the effort that workers have had to put into their work has shrunk tremendously. Creating clothes is no longer a task in which great care and dexterity are needed, instead, it has become simply following instructions and operating a couple of machines that do all the work instead. Logging used to be synonymous with hard manual labor, requiring those who were in good physical shape. Now, all that’s needed is the ability to sit in the cab of a machine that cuts down the trees and the only physical effort needed is the press of a button. While these are only two out of a countless number of economic niches, they represent an almost ubiquitous trend. We have seen also what happens when the need for an active lifestyle is removed. Rates of obesity have increased drastically [8] along with a general decline in overall fitness and far fewer people are able to make things with their hands. And this is without even mentioning how such trivialization greatly harms our psychological well-being [9].

From Physical to Cognitive Dependence

The mechanization we have been able to observe taking place in factories and forests is also starting to occur with white-collar jobs as AI has found itself in a position analogous to the power looms and the feller buncher. AI systems will undoubtedly increase efficiency and due to the competition between self-propagating systems, companies wanting to expand/keep their market share or governments wanting to hold onto/expand their power will be obligated to rapidly adopt this more efficient technology. Why spend hours drafting a report when you can have an AI do it for you? Just type in a prompt, copy and paste, and then touch it up to make it passable. Why read the full report when you can just have it summarized in easy-to-digest bullet points? And why spend your time thinking something through when you can have an AI speed up the process for you instead? But all of this comes with a cost to our critical thinking. To understand why we must understand how our brains work: our brains are in a constant state of creating new connections between synapses with priority given to those used the most frequently and in this particular way, our brain functions like a muscle [10]. And if we do not exercise our muscles they will suffer from atrophy and become weaker. A recent study conducted by researchers at Carnegie Mellon and Microsoft identified that 62% of respondents in a survey self-reported engaging with less critical thinking while using AI, primarily in low-risk scenarios such as when writing reports and emails. Drawing from the idea that our brain functions like a muscle, the researchers point out that if we do not routinely use critical thinking in low-risk scenarios, our ability to think critically will deteriorate [11]. Indeed, you can’t just get up from a couch and run a 5k: you need to exercise and adopt a healthy diet beforehand. If our ability to think critically is impaired due to prolonged periods of mental inaction, we could become dependent on AI even more which would create a negative feedback cycle that would leave us in a position of intellectual impotence without AI.

Our work occupies a great deal of our routines and many simply prefer to relax during their time outside of work instead of engaging in rigorous efforts – mental or physical [12]. This means that our work determines the physical and mental condition of the average person. There are also very real medical consequences for not exercising one’s cognitive capabilities [13]. It could greatly increase the risk of developing dementia or Alzheimer’s which kills more people than diabetes in the United States [14].

An argument given by proponents for AI when confronted with the problems of cognitive decline is that it’s simply a matter of how we use this new technology and that it’s our own responsibility to use it in a manner that aids our thinking rather than supplementing it. What they fail to realize is how companies and governments will want to use AI as a way to increase how efficient each person can be. It’s much more cost-effective to have one coder doing the work of a dozen coders and that can only be done by heavily utilizing AI to meet workloads. While those who are already proficient in their field of work may be able to get by without using AI for the time being, this will not be the case for long. An increasingly large amount of work will be dolled out and people will be expected to get up to speed with colleagues quicker until it gets to a point where people cannot keep up without AI. Thus, as the demand for efficiency increases and the capabilities of AI increase, anyone wanting to keep their job or advance their career will be forced to heavily use AI to automate tasks and in doing so, the need for critical thinking will be removed from the equation.

Regulation: A Fool’s Errand

Some may see regulation or reform as an acceptable solution but they fail to remember that self-propagating systems are always striving for more efficient means to out-compete others. The desire for more efficient means would overpower any safeguards that may be devised. Over time, large corporations and governments would be able to alter public opinion or legislation in a manner that suits further AI development and adoption of AI. With figures such as Sam Altman testifying in the Senate that more AI regulation should be enacted [15], they are simply throwing a bone - whether they mean to or not - at those who are rightfully worried about the continued development of what could very well become the most dangerous element of the technological system. Instead of buying into these platitudes, we must recognize the sobering facts that we will not be able to reform the technological system and that a revolution with the sole goal of dismantling the technological system is of foremost importance. Whether we can bring ourselves from the verge of technological ruin or plunge headfirst into it will be the defining question of the century.