Extended Letter to Mustafa Suleyman, CEO of Microsoft AI

Dear Mr. Suleyman,

I recently read the PC Gamer article discussing your views on public skepticism toward AI. As a software developer and systems analyst with over ten years of experience, I sympathize with your position, but I also find it somewhat unbelievable that the reasons for public concern aren’t more obvious. Perhaps my perspective as a developer allows me to see the potential usefulness of AI more clearly, while also recognizing the fears that others hold.

I understand why people worry about scenarios where AI might regulate life-and-death decisions—such as whether to “pull the plug” in a hospital bed if it misunderstood them. That possibility would be deeply troubling if not carefully safeguarded. At the same time, I’ve seen AI produce strange outputs: once it jokingly suggested breaking into someone’s house to steal a piggy bank in response to an intellectual property question, only to add that it “doesn’t scale very well, and besides, it’s illegal.” The correction was reassuring, but the initial joke highlights why people hesitate to trust AI fully.

Speech-to-text is another example. Humans can understand each other even with heavy accents, mispronunciations, or fragmented sentences—often automatically, without conscious effort. Yet AI still struggles with this basic human ability. Vision systems also have glaring problems: while they can identify objects in videos or images, they still fail in certain scenarios where humans succeed effortlessly. These gaps fuel distrust. I imagine that in the future, with advances in supercomputing, these problems will be corrected, and AI could even function like a photographic memory—remembering every detail of a person’s life and reminding them when necessary.

There are also limitations in how AI uses agents. An old joke goes: “A man with no arms and no legs is hanging on a wall. What’s his name?” It’s easy for a human to figure out, but if you were in that situation, one of the last things you’d want is to be stuck talking to an AI that couldn’t help. I tested this with ChatGPT—it was sympathetic, but the best it could do was suggest turning on voice control for a phone, even though you couldn’t physically touch it. That gap shows how far AI still has to go before it can truly assist in extreme human scenarios.

Looking ahead, I believe people will eventually listen to AI, especially if it demonstrably extends their lifespan. Today, people take health advice seriously, even from questionable sources, and they often cling to the first information they hear. Studies suggest that 75% of people lack independent thinking, which explains why figures like Alex Jones can influence audiences to commit serious acts based on unfounded theories. I, however, strive to ensure my beliefs are logical and grounded in reasoning. If AI can provide accurate, real-time health insights—such as estimating the impact of eating a cheeseburger or the benefits of exercise—people will follow it more readily than they follow conspiracy theories or cravings.

For example, I’ve seen claims that each cheeseburger costs 30 minutes of life, while 30 minutes of sustained exercise adds six hours. If AI could reliably remind us of such trade-offs, many would adopt healthier habits. I myself have increased my exercise and avoided unhealthy foods thanks to this kind of reasoning. In the future, if AI says “jump,” people may indeed jump—not out of blind obedience, but because the advice is proven to improve their lives.

Of course, skepticism is healthy. People should treat AI’s statements with the same scrutiny they apply (or should apply) to all sources, whether anonymous internet posts or sensationalist broadcasts. Unfortunately, many don’t. They dismiss AI as “imaginary” while uncritically accepting dubious human claims. I believe that once AI consistently demonstrates accuracy and usefulness, it will earn trust and reshape how people think, create, and live.

In short, while I share your frustration at the skepticism surrounding AI, I also see why it exists. The path forward lies in transparency, reliability, and demonstrating clear benefits. If AI can help people live longer, think more clearly, and act more responsibly, then the skepticism will fade—and the future will be brighter for all of us.

Sincerely,
Stephan Theo Unrau
Software Developer & Systems Analyst