I AM not comfortable with AI. It’s the name, for one thing. The very term “artificial intelligence” conjures a terrifying image of something beyond our control; an inscrutable, soulless consciousness that regards humans as little more than variables in a code, and maybe problems to be resolved. Fear of the unfamiliar, however, can be as dangerous as the thing that elicits the fear, so with that in mind, in order to prepare this commentary I conducted a little experiment. I asked a common AI program – which one is not important, as I have no interest in discussing the comparative merits of competing versions of the same product – to “create a bullet-point summary of the main ‘pros and cons’ of AI.” Its response is shown below. Two things stood out to me about this result. For one thing, the anthropomorphism, by which I mean the programming that makes the tool simulate speaking like a human, is a little unnerving. Second, while the “pros” of AI as described by an AI program are fairly familiar, the “cons” are noticeably selective. All of the ones listed are indeed commonly cited risks of AI, but over the past six months or so we have seen an extraordinary and expanding public backlash against AI on a few different fronts as well: Risks posed to energy and water security from enormous data centers, risks to intellectual property, the seemingly unworkable economics of AI, and discomfort with the apparent concentration of control of AI in the hands of a small number of hyper-wealthy tech barons.
In my opinion, what makes these observations significant is that even in this relatively simple requested task, AI seems to be acting beyond its design. It is, after all, nothing more than an advanced algorithm, designed to scrape data that already exists on the internet – or has been scraped from the internet already and internalized as retrievable information in its own code – sort that which is relevant to the query presented, and present that result in a templated output. Granted, each of these steps is highly detailed, but taken together, they only represent “intelligence” and “thinking” in the sense that the process mimics what a human without access to an AI tool would do, albeit much slower, to answer the same question.
And yet this non-human, mathematically designed and therefore predictable piece of programming did a quite unpredictably human thing in response to my simple instruction: It made a choice of what information to provide me as an answer, rather than provide me all the available or the most current information. One can see this easily from doing a non-AI supported web search of the same question I asked the program, in which the top results based on recent news stories are precisely about those things I noted above – growing public backlash (for example, anti-AI speeches or reactions to speeches at recent commencement exercises at schools such as Harvard University and the University of Arizona); increasing community resistance to the building of AI data centers; and a growing cost crisis for companies that is causing them to scale back their AI usage. As far as I’m concerned, selective honesty is also selective dishonesty. That’s a shockingly human response, and not one of the good ones. And it’s enough to make me rethink my already extremely limited use of AI.
Coming to terms with AI
For all my personal trepidation about AI, a feeling that is clearly shared by a vast swath of the general population, I have softened my initial reaction to its explosive growth over the past couple of years, which was, “AI is evil and must be destroyed!” Not because I’ve grown to appreciate AI, but because I am compelled to acknowledge logical reality. AI is, after all, the natural evolution of the algorithm that has been driving most of digital work and life for decades; AI is inevitable. And it has actually been around a lot longer than we may realize. Apple’s Siri and its Amazon cousin Alexa are examples of familiar forms of AI. The banking and insurance industries have been using logical relationship database programs for years – these were AI but just weren’t called AI – for things like credit scoring and claims adjusting. The difference now is that in the past two years or so, we have seen the advent of much bigger AI programs with broad capabilities, whereas those older versions were limited to comparatively specific applications. Again, this is evolution; we do not have to be happy about it, but because it is not in human nature to willfully reject technological advancement, we have to accept it. Calls such as comedian and TV host Ronny Chieng’s exhortation to recent graduates of Harvard University that “your mission is to kill AI” are, no matter how much our emotional human instincts may tell us otherwise, as nonsensical as saying something like, “We ought to reject the use of mathematics.” AI is not going to go away, although I believe there are strong reasons to anticipate that it will evolve further into a very different form than the frenzy-created monster it is today, which I will discuss in the next section. And, from reading the various articles written by our esteemed contributors for this edition of The Manila Times 500 (And the next 500), I take great solace in the evident fact that among Philippine businesses and institutions, a practical open-mindedness tempered with healthy skepticism toward AI is the common perspective. They see promise where some see peril.
The bubble will pop
We may not really begin to appreciate the usefulness and value of AI as a tool until its current boom goes bust, and there are ever-increasing signs that will happen soon. I realize this assertion is quite debatable, but allow me to share a bit of personal insight on why I believe this is so. Quite by chance, during my old career with a major European automaker, I was assigned to oversee a project in San Francisco during most of 1998-1999, a time period those with a bit of knowledge of economic history will recognize as the peak of the “dot-com boom.” That was the time when the potential capabilities of the internet fired the imagination of investors and entrepreneurs, and there was runaway growth in the sector. For people like me, whose business was to cater to a market of consumers who suddenly had more money than they knew what to do with, life and work was surreally fantastic. My friends have heard the stories; picture the lifestyle depicted in the movie “The Wolf of Wall Street” but without the underlying criminal activity, and you can have a pretty good idea of what it was like. But as those of us close to it but not quite a part of it knew then, and everyone knows now, it was not sustainable. There were mind-boggling amounts of money circulating throughout a closed-loop economic system, which allowed the creation of products and services without any regard to their utility or desirability to a paying market outside that system. That eventually collapsed, and what remained after the bubble burst were the companies who had thought ahead, and developed products and services that people and businesses would actually use and pay for. More to the point, in the context of our magazine here, the enterprises who were not part of that ecosystem but were able to successfully take advantage of its growth and eventual collapse were those who could discern which parts of it were valuable and value-adding, and which parts were pure speculative fluff, avoiding the latter and patiently waiting for the former to emerge from the chaos of the bursting of the dot-com bubble. As I see it, the AI industry is now at the point the dot-com sector was in the latter part of that heady period, say the second half of 1999, when the first signs of trouble for dot-com start-ups starting appearing. In recent weeks, there have been a growing number of reports of companies scaling back their spending on AI services, because of a combination of high costs and uncertain ROI from using them. AI companies themselves are beginning to hit the wall in terms of expansion; not only are communities becoming increasingly resistant to the launch of new data centers, energy and supply chain constraints are becoming serious obstacles. And then of course there is still the significant public backlash against AI. One example of a manifestation of that, as I am writing this (during the first week of June 2026), the search engine provider DuckDuckGo has reported that users of its “AI-free” version have shot up by some 30 percent in a matter of days, after the announcement by search-engine giant Google that it would embed Gemini, its own AI tool in its search engine. Whether users ask for it or not. What may be left after when the AI bubble pops is difficult to guess at, but experience suggests that the collapse, when it comes, will be less disruptive than feared, and that potent and potentially world-changing AI products will be the ones left standing. The collapse of the “dot-com” bubble, after all, is what gave us much of what we take for granted today, and in fact, have changed the shape of modern society – social media; e-commerce as epitomized by the Amazon model; streaming services for music, film, and TV entertainment; instant messaging applications; and ultimately, the “virtual meeting” tools that helped the world keep working during the Covid-19 pandemic, and have changed the shape of work ever since. Whatever emerges on the other side of the great reset that is coming for the AI sector is likely to have no less profound an effect on our world, and the way that we do business and interact with one another.
A principled approach
If we are to make the most of AI and the transformation it must inevitably undergo, we need to approach it with heads up and eyes open. As I said, the perspectives of Philippine business and institutional leaders toward the perils and promise of AI are encouraging, because almost everyone seems to recognize that AI will work best when it is regarded for what it is – a tool of potentially immense power when applied properly – and what it is not, which is a solution for everything that can replace human intelligence and ingenuity. What constitutes the “principles” may very well differ depending on the industry involved and the circumstances at hand; the concerns of my industry with respect to AI are very different from those of, say, manufacturing or banking. And overlying all of it, no matter what industry or sector, are the broader social risks of environmental impact, economic displacement, unethical and criminal activity, and psychological impacts on people. The best approaches to managing AI, through laws and regulations, best practices in organizations, and even personal behavior are likely to debated forever; they still are for the internet itself, and it’s been with us as a core component of modern civilization for 30 years. But I know that leaving it at that sounds like a cop-out, so let me close with this thought. From my perspective, as an author, as a media practitioner, as a single dad doing my best to manage my weird little household, as a citizen of a dying planet, my use of AI will always be guided by two non-negotiable principles: First, AI must increase and not diminish the value of human ingenuity and imagination. Second, the output from AI-generated (generative AI) or AI-assisted (agentic AI) use must justify the cost of using it, in time, resources, and risk exposure. That’s it. If AI can meet those conditions, I am happy to make use of it. If not, well...I’ve been alive since dial telephones were commonplace and most newspapers formatted their pages with linotype machines, I think I will thrive well enough without it.
Here’s a clear, concise summary of the main pros and cons of AI — the kind you can use for reports, presentations, or quick briefing notes.
Pros of AI • Efficiency and automation — AI handles repetitive, timeconsuming tasks quickly and with fewer errors, boosting productivity across industries.
• Data-driven decision-making — It can analyze massive datasets, detect patterns, and generate insights that humans might miss.
• Improved accuracy — In fields like healthcare, finance, and engineering, AI systems can outperform humans in specific, welldefined tasks.
• 24/7 availability — AI systems don’t get tired, making them ideal for customer support, monitoring, and continuous operations.
• Innovation and new capabilities — AI enables breakthroughs in drug discovery, climate modeling, robotics, and personalized services.
Cons of AI • Job displacement — Automation can replace certain roles, especially routine or manual jobs, creating economic and workforce challenges.
• Bias and fairness issues — AI can inherit or amplify biases present in training data, leading to unfair or discriminatory outcomes.
• Privacy and surveillance risks — Large-scale data collection raises concerns about misuse, tracking, and loss of personal privacy.
• Dependence and skill erosion — Overreliance on AI may reduce human expertise or critical thinking in some domains.
• Security threats — AI can be used maliciously (deepfakes, automated cyberattacks) and can also introduce vulnerabilities if poorly designed.
If you want, I can also turn this into a one-paragraph summary, a bullet list for slides, or a more nuanced editorial-style analysis depending on your needs.