In a short span of three years, AI (Artificial Intelligence) has moved from the confines of cutting-edge laboratories to an integral part of global technological infrastructure. It represents one of the swiftest, most far-reaching, and unprecedented technological proliferations in human history. With the promise of improving the overall quality of life for consumers worldwide, AI nevertheless carries significant risks.
In his new book, "The Coming Wave", Microsoft AI CEO, Mustafa Suleyman, warns that the rapid spread of AI will bear unpredictable consequences.
My father recommended The Coming Wave, and I owe him thanks, since I nearly dismissed it as another tech guru memoir. I've long been wary of such books and self-help literature. They tend to peddle lofty ideals while leaving readers no practical path to reach them, which is counterproductive, even destructive.
An example that comes to mind is "Rich Dad, Poor Dad" by Robert Kiyosaki. It provides little in the way of reasonable advice to the reader, instead acts as an advert of sorts for Kiyosaki's other endeavours like financial education courses that primarily benefit Kiyosaki himself.
Judging this book by its cover, I thought it was a newer age version of self-aggrandisement. I was wrong.
Premise
The Coming Wave is Mustafa Suleyman's attempt of providing guidance in an era of uncertainty due to the proliferation of AI. It serves not as a guide, but rather a checkpoint that ensures that we are prepared before we inevitably reach an increasingly AI integrated world.
Mustafa Suleyman opens the book by drawing upon the dual nature of "the wave." He recalls the flood myths of various civilizations and emphasizes that these tales signify not only destruction but also renewal. Floods, he elucidates, are transformative events that punctuate our history.
According to Suleyman, the "dual drivers" of the coming wave of technology shall be synthetic biology and AI. However, he prioritizes AI, and considers synthetic biology as one of the byproducts of AI's proliferation.
He defines the main challenge that will arrive with the coming wave - the containment problem. It refers to "technology's predisposition to diffuse in waves and have unforeseeable consequences"; an issue of great significance and one that lacks concrete solutions.
He implies that we are in a technological revolution. These come in "waves", says Suleyman. He defines them as bundles of inventions stemming from general-purpose technologies that enable step-change advancements. Historically, stonework and fire have served as proto-general-purpose technologies that enabled a slew of advancements and inventions as they rippled out.
Suleyman also highlights the irony in how they become invisible and are taken for granted after a certain point, as their proliferation happens like the spread of wildfire. People of the younger generation have come to associate this irony with the classic "you people have it easy" talk we receive from our elders. What Suleyman does is highlight that it comes naturally with the passage of time and improvement of technology.
The containment problem becomes apparent when we realise that as technologies become more widespread, they proliferate, become more affordable, and improve rapidly.
The Dilemma
The Microsoft CEO presents the grim reality that we cannot stop the AI machine from moving. He provides several examples from history and traces the history of various technologies and their proliferation, such as the Gutenberg press and the space shuttle. He explains how, despite the initial Luddite response, proliferation still occurred. The reason for this is because they simply served some central utility, such that other inventions could also come up thanks to them.
Save for a few rare exceptions, such as nuclear weaponry, proliferation seems to be the general trend for all technology. The severity of our current dilemma skyrockets when we come to the realization that AI has rapidly improved in the past decade. He narrates the story of how AI has shaped up to be the most transformative technology in human history, specifically focusing on its ability to learn, showcasing early models such as AlphaGo and AlexNet, before briefly discussing LLMs such as ChatGPT.
He notes how the current wave of generative AI is yet another example of technology drastically improving with demand. He ends this chapter by remarking that we are rapidly approaching more and more general AI models that display more holistic capabilities, which he refers to as "ACIs" or Artificially Capable Intelligences.
Ultimately, if push comes to shove, the four horsemen of the coming AI apocalypse will be its four main capabilities. These include the ability to cause global redistributions of power, a never-ending cycle of improvement for AI models, a state of "omni-use" as it becomes more entrenched in our lives and finally, autonomy.
The philosophical idea that if AI manages to fully automate itself, it may escape human grasp. From here, the AI could hypothetically bring forth an apocalypse of sorts. At the very least, it could bring a drastic change to the world order.
To Suleyman, these four cornerstones of the coming wave are warnings that are not to be taken lightly. Otherwise, we may very well risk our control over the happenings of the planet itself.
One of the most bone-chilling passages in the book is Suleyman's quotation of Stuart Russell's "gorilla problem." The analogy suggests that while gorillas are physically stronger than humans, they remain confined because humans are more intelligent. The same dynamic, Suleyman warns, could one day apply to humanity in relation to AI.
Even so, the plot thickens. The improvement in AI and the improvement in military technology in the past decade is no coincidence. Suleyman describes it as a "fragility amplifier", something that improves with AI. Other examples of amplifiers include cyber-attacks and deep fakes.
We are already seeing the repercussions of these technologies. Imagine what might occur in a decade. The growth and proliferation of AI is fueled by rivalries, amongst which the Sino-American rivalry takes center stage while other nations also play their roles in the current AI landscape.
The Microsoft AI CEO further explains how tech companies could easily overpower governments and believes that governments must modernise to remain relevant in the face of a potential era of "techno-feudalism" that rewrites our social order.
He also believes that the coming wave is "rocket fuel" for authoritarians who could combine existing data with emerging technologies to create a perfect surveillance system. Early prototypes of such technologies exist, Suleyman writes, AI might be the final missing component for these ideas to be actualized.
Taking all of this into consideration, the future does certainly seem bleak. However, for the world to have any hopes of not becoming a dystopia, containment becomes essential.
The solution
Masterfully describing all the threats AI may pose, Suleyman also gives us some hope of achieving containment.
Just as the proliferation of technology will greatly empower organisations, it will also empower the individual, which could have myriad complex outcomes globally. Thus, the coming wave shall be filled with contradictory ideas, similar to the state of affairs during the current social media era, where any person "armed" with the correct technology has had the same influence, following and power as the leader of a small nation, literally!
He again highlights how we are at the centre of a crossroads between technology's positive and foreboding consequences, and that it is up to us to navigate through this dilemma.
Rejecting stagnation as a solution, the author reminds that innovation is an imperative to ensure sustainability of our current economy, threatened by challenges like dwindling world population and rising demand for materials for cleantech.
Towards the end of the book, Suleyman offers a set of ten guiding principles towards containment. The genius of these steps lies in the fact that they embrace AI in moderation. Suleyman's commandments are arranged in concentric circles, each commandment builds on the values of the last.
The innermost value is that of "feeding" AI the right information so that it may not have "hallucinations" (When AI comes up with fictitious claims and confidently claims them to be real. I have seen one too many of my peers get tricked) and perpetuate misinformation.
Other principles emphasize rigorous monitoring of emerging technologies and global cooperation across researchers, corporations, governments, and international institutions.
Ultimately, Suleyman concludes his work by reminding us that each one of us has a say in deciding how the world we wake up in tomorrow will shape up to be. The responsibility lies within us to innovate and bring hope. It serves a positive ending to an otherwise bleak book filled with warning after warning.
In conclusion, The Coming Wave stands as one of the most important works in recent memory and is unquestionably a worthwhile read.
Have you liked the news article?