Statement

The World Already Changed. Most People Just Haven't Checked.

This is not a prediction. Predictions are about things that have not happened yet. What I am going to describe has already happened, is happening now, this week, in the industry you work in, in the company you run, in the profession you built your life around. The only question left is whether you understand that, and whether you understand it in time to do something about it.

Artificial intelligence did not arrive on a schedule anyone published. It arrived the way most genuinely disruptive things arrive: incrementally, then suddenly, and by the time most people registered the change, the ground they were standing on had already shifted under their feet. The world that existed before it is not coming back. It is not going to stabilize into something familiar once the hype settles down. The hype is not the story. The hype is the noise around a structural transformation that has no reverse gear.

I have been building with this technology every single day for the past three years. I have been working in and around the systems it is replacing for forty. What I am telling you here is not commentary. It is observation from inside the process.

What concerns me most is not the technology. What concerns me is the people. Business owners I know. Professionals with real expertise and twenty years of experience. Colleagues who are genuinely good at what they do. Every week I watch them make the same quiet decision: that this particular development does not require their urgent attention. That they can engage with it later. That the people raising the alarm are the same people who said the same thing about the internet, about smartphones, about the cloud, and things turned out fine.

Things are not going to turn out fine this time in the same way. The internet gave people new tools. Smartphones gave people new access. What artificial intelligence is doing is replacing the cognitive labor that was the economic foundation of entire professional classes, and it is doing it faster, across more domains simultaneously, than any previous technological shift in recorded history. The comparison to prior disruptions is not wrong because it is pessimistic. It is wrong because it is not pessimistic enough.

The legal assistant who spent her career managing correspondence, preparing filings, and handling the routine work that kept a law firm running does not need to be replaced with a plan. She needs to be replaced with a paragraph. The translator who built a practice over a decade is competing with systems that operate in hundreds of languages simultaneously, at a cost per word that rounds to zero. The medical receptionist, the junior analyst, the proofreader, the bookkeeper handling accounts that do not require judgment: these are not professions that will adapt around artificial intelligence. They are professions that artificial intelligence is absorbing, and the pace of that absorption is accelerating.

New categories of work will emerge from this transition. They always have, after every major disruption, and there is no serious reason to believe this time will be different in the long run. But the generation currently in the workforce, the one that spent the last three years deciding this did not require urgent attention, will largely not be the one that fills those new roles. The people who fill them will be the ones who started paying attention in 2024 and 2025, who learned the tools, who restructured how they worked, who understood that the question was never whether to engage but how fast. The people who waited will be waiting for jobs that no longer exist in the form they remember.

I am not a political person, and this is not a political statement. It is a structural observation.

When unemployment reaches a scale that existing social safety nets cannot absorb, governments intervene. They have no other option. The social contract that holds democratic societies together requires that people have some means of participation in the economy, and when the market stops providing that at sufficient scale, the state fills the gap or faces consequences that are worse than filling it. That is not ideology. It is history, repeated in every industrial transition that came before this one.

What that intervention looks like in the age of AI is the conversation everyone who matters is already having. A universal basic income, funded by the productivity surplus generated by AI powered systems, is not a fringe proposal. It is the conclusion that serious thinkers across the political spectrum, Elon Musk and others among them, keep arriving at when they follow the logic honestly. I have arrived at it too.

The question of how to fund it is where the politics get complicated and where I will say one specific thing plainly. In Germany, the instinct will be to tax the productive more heavily to pay for the transition. That instinct will produce the opposite of its intended result. Every person with serious assets and any degree of mobility will do what an increasing number of people I know personally are already doing: they will move to Cyprus, to Malta, to Portugal, somewhere with a tax regime that reflects a government interested in keeping productive people present rather than extracting from them until they leave. The capital goes. The expertise goes. The tax base shrinks. Germany does not become more competitive as a result.

I watch this happening in my own circle. Wealthy friends who spent their careers here, who built things here, who paid taxes here for decades, are making plans, and some of them have already made the move. This is not a threat. It is a description of human behavior under predictable incentive structures. Any policy framework that ignores those structures will produce the same outcome regardless of its intentions.

This is the part that people in large companies and sophisticated industries hear and assume applies to someone else. It does not.

The bakery with three employees and a point of sale system from 2018. The tax firm that still runs on desktop software and processes everything manually. The law office that has been meaning to update its intake workflow since before the pandemic. The family medical practice that considers technology a distraction from patient care. The online shop built on an architecture that was outdated five years ago, offering no personalization, no intelligent search, no conversational interface, no service layer that a well built AI system could provide in an afternoon. Every one of these businesses is exposed. Every one of them is losing ground to competitors who are moving, and the gap compounds every month they do not.

The window in which this is still a strategic choice, rather than an emergency, is not permanently open. The businesses that restructure now, that find someone who actually understands what they are doing rather than someone who read three articles about AI and bought a subscription to a chatbot, those businesses will look back on this period as the one where they made the right decision. The businesses that do not will look back on it as the one where they did not, from a position where the options have narrowed considerably.

The physician who declines to integrate AI into diagnostic work is, in my view, making a choice that affects patient outcomes, and they should be willing to say so explicitly rather than framing it as professional judgment.

I run a veterinary practice alongside my wife, Dr. Louise Morgott. We use AI to analyze X-rays, to evaluate ultrasound imaging, to cross-reference symptom patterns against the available literature faster than any human practitioner working alone could manage. Every animal that comes through our door receives the benefit of that additional layer of analysis. The diagnostic quality is higher. The margin for error is smaller. The outcomes are better. That is not a claim. It is the result of using the tools that exist rather than the ones we were trained on.

The same applies in human medicine, in every specialty, at every level of clinical practice. The radiologist who reads imaging studies without AI assistance is working with a smaller dataset than the one available. The internist who does not use decision support tools for differential diagnosis is relying on memory and pattern recognition alone when better pattern recognition is available. At some point, declining to use tools that improve accuracy stops being a professional preference and starts being something harder to justify. That point is arriving faster than most of the medical profession is willing to acknowledge.

Amazon is not running a pilot program. It is running a template that every large company operating at sufficient scale is studying and will eventually replicate.

Warehouse robots work around the clock. They do not organize, they do not negotiate, they do not require compliance with labor protection legislation designed for workers with biological limitations, and they do not take inventory home in their pockets. The economic logic of replacing human labor with robotic labor, for the categories of work that robots can perform, is not a future calculation. It is a present one, and the companies that have already run those numbers are not waiting for a better moment. They are deploying.

The same logic applies everywhere that the work is repetitive, volumetric, and does not require judgment in the ways that human judgment is actually irreplaceable. Which is, it turns out, a very large percentage of the work that currently employs a very large percentage of the workforce.

I drive a BMW iX M60. It is an exceptional machine in the ways that German engineering has always been exceptional: the build, the dynamics, the refinement in the things that can be engineered to a specification. Its artificial intelligence is an embarrassment. The voice system does not understand me reliably. The processing hardware behind the interface is underpowered to a degree that produces visible, frustrating latency in a vehicle that costs what it costs. The gap between the ambient promise of a modern premium automobile and the actual intelligence it deploys is not subtle.

I recently spent time in a Chinese SUV. The AI handled every language I gave it, instantly, fluently, with processing speed that reflected genuine hardware investment by engineers who understood that the intelligence layer of a modern vehicle requires serious silicon. The conversation was natural. The latency was negligible. The contrast with what I drive was not something I was expecting to find so stark.

German automotive engineering remains extraordinary in its traditional disciplines. Its AI strategy, with respect to the companies involved, is not keeping pace, and the companies that believe the brand equity they built over the last century will carry them through the software decade are making an assumption I would not want to have to defend. BMW has not asked me for input on this. That is, of course, their choice.

The hackers are using AI too. I watch it in real time on my own infrastructure, every day. The attacks that arrive are more patient, more adaptive, and more precisely targeted than anything I was seeing five years ago. The phishing emails reaching people's inboxes today are not crude, misspelled attempts. They are composed by systems trained on successful deceptions, optimized for what works, and refined continuously based on the response rate. Many of them are very good. Some of them are better than what a human social engineer could produce working alone.

Go to the dark web. It is entirely possible to purchase the login credentials for Gmail accounts, for GMX accounts, for banking portals and corporate systems, belonging to people you know, including your clients, your business partners, and members of your family. The data is there because the systems it came from were not adequately protected, and the people it belonged to had no idea it was available.

In the early years of my forensic career, I ran brute force attacks on accounts as part of penetration testing work. That methodology is no longer necessary. The credentials are already indexed, already searchable, already available for less money than a cup of coffee, because the people they belong to use the same password for every account they own and have never enabled two-factor authentication. Not because they cannot. Because no one explained to them convincingly that the threat is real, that the barrier to exploiting it is lower than they imagine, and that the cost of inaction is not theoretical.

The company running customer facing systems, internal portals, and operational infrastructure without regular security audits is not managing acceptable risk. It is accumulating liability. DSGVO makes that liability explicit and financial. A data breach is not an embarrassing incident. It is a legal exposure, a reputational event, and a forensic case that will be assembled from evidence that was, in many cases, entirely preventable. I have built a significant portion of my forensic career on the wreckage of systems whose owners treated security as an optional expense. The pattern is consistent and it is not slowing down.

And while we are on the subject of infrastructure: Windows is a platform I have spent forty years watching fail in ways that were preventable. The machines that cost the most to recover, that went down hardest after an update, that provided the richest hunting ground for the attackers I spent my career studying, ran Windows. That is not a coincidence, and it is not an advertisement for an alternative. It is an observation that has held up across four decades of professional experience and I am not going to stop making it because it is inconvenient.

Not panic. Not the resignation of someone who has looked at the problem and decided it is too large to address. What this moment requires is the willingness to engage seriously, with someone who has been inside these systems long enough to know what works and what does not, and who can help you move in a direction that will still make sense when the landscape looks different again in eighteen months.

The transition is happening. The pace is not slowing. The window in which acting is still a choice, rather than a response to crisis, remains open, but it is not permanently open, and the assumption that there will be a more convenient moment to address this has been wrong for the past three years running.

I have been living inside this field since before most of the people currently disrupting it were born. I have never seen the pace of change move the way it is moving right now. That is not something I say to create urgency. It is something I say because it is true, and because the people who are positioned well for what comes next are, without exception, the ones who decided to take it seriously before they had to.