On the table, a piece of half-eaten cake, in the room of our garden shed, smoke still lingers from blown-out birthday candles. I’ve just officially turned the numbers on the dial of my life – I’m celebrating my 43rd birthday. Of course, I won’t tell you what I wished for when blowing out the candles, as I don’t want to ruin the chance of it coming true. But it’s an appropriate moment to tell you something else, equally important.
Lately, we’ve been living in strange times. Progress in technology has accelerated so rapidly that even among experts in the tech field, a deep divide has emerged. A chasm that has split society into viewers and initiated insiders gasping for breath. We’re all flooded with information about artificial intelligence progress (hereinafter sometimes referred to simply as “AI” in this text). Mostly, we’re all tired of hearing about it, because AI literally jumps out at us from every refrigerator. However, there’s also a subgroup of “initiated gaspers” who, besides the equally annoying bombardment of AI news, must also deal with artificial intelligence in their daily work. Picture this: Everyone aboard an aircraft experiences that unsettling stomach-churning sensation during rapid ascent after takeoff, watching billowy clouds drift past their windows. But a select few in the cockpit don’t need to speculate among themselves—the array of instruments surrounding them reveals exactly where this flight is headed.
It’s a precarious informational advantage, heavily outweighed by crushing responsibility. Ask yourself honestly: How many would trade that panoramic view of clouds for the pilot’s burden of responsibility for every soul on board? You noted, Filip, that those in the cockpit aren’t just “informed”—they’re also “gasping for breath.” Absolutely. Only those few in the cockpit know we struck something during takeoff (a flock of birds, perhaps?) and that some instruments now display false readings due to the impact. The altimeter insists we’re maintaining altitude, yet through the windows, we can clearly see ourselves pulling away from the airport. Meanwhile, passengers in the cabin anticipate their usual in-flight service or escape into their entertainment screens. While you’re absorbed in TikTok or the latest blockbuster, those in the cockpit know the coming hours “won’t be pleasant”—and things may well turn dire.
What we collided with during ascent was Artificial Intelligence. And it will fundamentally reshape all our lives. Like the aircraft scenario, there’s no immediate catastrophe post-impact—nothing’s on fire, engines continue running. Even the entertainment system works perfectly; life proceeds as normal. But the pilots know our fuel won’t last forever and, crucially, we’re already airborne. If we can’t execute a safe landing ourselves, gravity will gladly handle our return to earth—without hesitation. A competent pilot doesn’t sprint down the aisle screaming “We’re doomed! We might all perish!”—even if that weren’t entirely false. Instead, a skilled pilot must, precisely in such moments, engage rational thinking, anticipate what lies ahead, and communicate those crucial realities to the crew.
Just as pilots understand that regardless of their efforts, fuel reserves will sustain flight for only so many hours. Pinpointing the exact minute becomes irrelevant and futile. The critical truth is this will happen far sooner than you originally expected to die. Much sooner. I just celebrated my 43rd birthday, and artificial intelligence’s effects will overtake us before I reach fifty. I’m being this direct because it means we’ll all spend an unavoidable portion of our lives navigating AI’s consequences.
In a real airplane, the pilot would have your full attention – any announcement from the cockpit interrupts your entertainment on the screen. However, I don’t have that advantage, so I just want to ask you: Stop your TikToks, pause the plot of your life movie and listen to this important warning. In the following lines, you learn what this AI hit will likely bring.
The Great Blow to Self-Confidence
AI’s arrival brings one fundamental reality: artificial intelligence now surpasses humans in virtually all basic cognitive abilities—reading text, recognizing images, retaining facts, analyzing hypotheses, reasoning, creating text, music, videos. You’ll still encounter plenty of skeptics who’ll reassure us that certain tasks remain beyond AI’s reach. “Besides,” they’ll add, “AI hallucinates.”
Should you require proof that AI has already surpassed human benchmarks, I’m happy to provide citations in the endnotes. However, I’d suggest those hours spent reading through the proofs might be better invested elsewhere. Consider this: chess programs aren’t evaluated against random street player, they must defeat grandmasters and Go champions. Similarly, when AI reaches the 98th percentile of human capability, it has long since exceeded what we, ordinary mortals, can achieve. This text won’t focus on “convincing” you of this reality. If, after reviewing all available evidence, you remain doubtful that AI has surpassed us, that’s perfectly acceptable. I respect your position. However, the remainder of this piece probably isn’t intended for you. Sit and enjoy the in-flight entertainment, until it stops.
The most devastating blow to our functioning will be acknowledging this “AI dominance.” Understanding this correctly: It’s not about whether you’ll actively declare or publicly communicate this power shift to surrounding people. The truly difficult battle occurs where you reign supreme: in your private decision-making. Genuine acceptance that AI “has the edge” only emerges when we begin delegating real tasks to it. Those “I’ll handle this small thing myself” arguments stem from both weak(er) AI tools command skills and resistance to accepting that someone else -or something else could handle this type of work (as well).
Work With and Without AI
We dedicate the lion’s share of our adult lives (second to sleep only) to work. It’s where we attempt to monetize our abilities, time, and effort in exchange for salary. And this is why we should pay close attention here, as artificial intelligence can match or exceed our offerings in all three areas. We’ve already examined the capabilities question (see “The Great Blow to Self-Confidence”). If you harbor lingering doubts, remember that most of us don’t even utilize 100% of our human knowledge during work hours—setting a relatively low bar for AI. Furthermore, artificial intelligence operates continuously without complaint, even on mind-numbingly repetitive tasks. AI doesn’t require vacation time or sick leave; it has no children to care for.
Unsurprisingly, management teams responsible for task allocation will actively investigate how to assign substantial work portions to this “always-on” AI workforce. The speed and intensity of that shift remain up for debate, but early experiments are already underway. When just few of your industry competitors embrace this approach, they’ll gain such a significant cost advantage that ignoring the trend becomes impossible, even if you (and your managers) prefer to avoid AI participation.
Naturally, a “counter-movement” will emerge, rallying under banners like “We proudly employ humans!” and pursuing the opposite philosophy. Some AI-unprepared nations (CEE countries could easily fit this category) may attempt to make virtue of necessity, reassuring citizens “We won’t sell you out to artificial intelligence” (more on this in “Pressure on Politicians”). Yet this will largely mirror destiny of film photographers after digital cameras arrived; Or miners proudly succumbing to lung disease from inadequate working conditions.
Your specific profession in your particular region might survive this AI workplace integration wave. However, for most knowledge and office-based roles, the job market will certainly become more competitive.
In discussions about AI’s impact on the job market, 3 statements often appear that I’d like to briefly explain and comment on:
“Artificial Intelligence won’t replace you at work. You’ll be replaced by people who can use it.” This statement documents what mood we’re in today. We’re in such an early stage of AI deployment that (based on recent studies) only ~4% of Slovaks use AI so often that they need to buy a paid version of one of the AI models. So, if you’re already among those progressive AI users today, it indeed seems true that you won’t lose your job and will rather even “blow it away” from someone else. Learning how to use individual AI tools on the market is therefore certainly a worthwhile idea to follow. However, as an imaginary pilot in the cockpit, I see that the validity of this statement will pass in just a few quarters. Artificial intelligence is moving into a realm where higher-level orchestration AI bots assign tasks to flocks of smaller models (let’s call them apprentices) and thus accelerate the overall effect of solutions. For humans, therefore, knowledge of “apprentice-level AI models” may be interesting and exciting, but not very effective for saving their workplace. After AI changes catch up, workers won’t sit in the same place and just assign their tasks to artificial intelligence. The final stage of AI automation will be that the main AI robot assigns tasks to other AI robots. And so, to be completely precise: Your job won’t be taken by AI (nor by those who will know how to work with AI tools), but by those who will know how to configure flocks of AI robots.
“Just as AI will eliminate/replace jobs, it will surely create new jobs too.” I’m really sorry, but this statement is just wishful thinking and can be easily mathematically disproven. Count with me: Let the tandem “human+AI tool” be able to do work in a third of the current time. So the same performance can be achieved by a company with 1/3 of its employees. Of course, those redundant 2/3 of employees could find similar work (where a new, efficient model will already be waiting for them), in which they’ll do the equivalent of work for 3 people. But this means that capacity will be created for 2/3 of people (which is 2x more than remained working) *3-fold work = 6-fold of current production. In other words, humanity would suddenly need a sudden 600% increase in demand for services and goods to be able to sustain the entire workforce. And it’s obvious that this can’t happen even within a year or two. AI automation will therefore certainly lead to a (temporary) downturn in the job market, and even sincere efforts to replace work for those affected by this trend won’t prevent it.
“But it’s humans who pay for goods and services. So firing too many people will take the demand down, as jobless people can’t buy the same volume of goods and services.” This argument seems much stronger on first view but actually is also yet another fallacy. Let me take you for a walk on thought experiment: Let’s take some essential service like mobile carrier. Almost everybody pays an amount to his mobile phone operator on a monthly basis (btw. even the already unemployed people, but less leave that aside). Now let’s assume that your mobile phone operator manages to displace 20% of their stuff (e.g. from technicians and programmers) with AI, while keeping the quality of the service comparable. Thus, you, as an end-customer, do not feel any change in the service, the service is still essential for you as before. Now the mobile phone operator can even give discount to those “threatening quit” as the service is cheaper to operate. The demand for mobile phone services is not likely to fall. Nor the bread, milk or washing powder. Unless all companies implement AI on the same day (leading to COVID like shock), the demand will not go through any sudden shocks.
Education
Surpassing human capability thresholds will seriously impact another cornerstone of our lives. Since ancient times, knowledge accumulation has commanded respect and recognition. Consider councils of elders or royal advisory bodies. Deep expertise in any field – whether as a professor or cited expert – still confers social status today. This is just about to change dramatically.
To avoid misunderstanding, we must distinguish between knowledge’s value and utility. Even in the AI era, knowing things remains useful. Constantly consulting artificial intelligence for basic general knowledge would inflate service bills and slow daily decision-making. Moreover, internalized knowledge helps detect AI hallucinations and errors. Thus, retaining knowledge stays practical in the AI age. But, …
Education itself will be acquired differently and serve different social functions. When AI in your smartphone possesses knowledge comparable to university professors across all disciplines, knowing 50-70% of your professor’s knowledge (typical result of university graduate studies) provides no extra informational advantage. Reaching approximately 100% of professorial knowledge and advancing science through research retains value, but only a fraction of students achieve this threshold – those pursuing doctoral studies or professional research careers. Currently, merely 2-4% of university students earn PhDs. The remaining 98% graduate to obtain knowledge-level “certificates” required by employers. For this majority, current educational models will cease making sense.
For employers, knowledge-level certificates (which AI surpasses effortlessly) will lose value. Additionally, five-plus years of study duration will seem prohibitively long and expensive for student (families) compared to benefits received. We’ll retain fields where states or societies mandate official diplomas (medicine, for instance), but many positions (including AI-related roles) won’t require them. University education will face fundamental disruption.
University educators should contemplate this trend most seriously. Higher education systems have guaranteed them steady student streams (albeit declining in quality). Their positions derived from serving as authoritative program guarantors – the outside world equivalent of respect for WHAT they knew. In the new AI era, however, WHAT becomes easily replaceable. Don’t misunderstand, I’m not suggesting they’ll become worthless. They must recognize their added value shifting (from WHAT) towards “content creation” focused on HOW material is PRESENTED. Anyone can dump entire syllabi into ChatGPT requesting lectures or summaries without professors involved. Yet people will still pay for “excellent human presentation” of complex content. If you’re among university educators or experts, begin experimenting immediately with YouTube channels, podcasts, or other digital content formats in your expertise area.
Education’s changing value brings another consequence that compounds the job market impact discussed earlier. Until now, qualification levels and work sophistication served as high ground during economic “floods”. Unemployable in the job market? Increase your qualifications. Complete university, improve foreign language skills. Education as an income advancement tool. This role of education has been creating pyramid-like impressions -lower social strata can climb toward better long-term salaries and status. Simultaneously, this creates false impressions that higher qualification levels offer better protection from job market contractions (which AI will bring).
That perception will rapidly prove mistaken – really within quarters. AI will leap beyond most position requirements in raw competencies. Highly qualified programmers will become as replaceable as supermarket cashiers. When this trend emerges, it will devastate unprepared office workers with dismay and bitterness. Above-average qualified and intelligent individuals will suffer most, perceiving this as profound “injustice.” They’ve invested in education their entire lives – how can they end up equivalent to those who neglected their learning?
Even Deeper Cesspool of Online Content
While job market effects will impact individuals at varying times and intensities, social phenomena will affect us all equally – and equally poorly. Artificial intelligence introduces one major shift to digital content creation.
Consider the combination: today’s already-low content consumption standards (especially on social networks) plus effortless AI content generation (AI-written articles, AI-generated video and audio content). Previously, content volume was limited by creation and verification time requirements. AI’s arrival definitively opens these floodgates – AI can produce vastly more content in no time. The vicious circle completes itself: mainstream content quality being already alarmingly poor, now combined with tireless new content generation, for this mass-produced AI content to survive, it must target current audience topics and desires. The inevitable triple-ingredients leads to AI churning out not just massive content volumes, but massive volumes of mass taste (low-quality) content. If you or others around you already now spend hours by watching Netflix or TikTok videos, understand- it’s gonna get worse soon.
What Can I Do About It ?!
Truly useful warnings shouldn’t merely frighten or paralyze with anxiety. Thus, allow me to share several “cockpit observations” about managing this AI arrival at the individual level.
Era of Active Self-Supporters
During emergency aircraft landings, passengers who freeze in terror and waiting for rescue most likely perish. Crisis survival requires active participation in your own preservation. If we accept that many jobs will become AI-replaceable (potentially threatening employee status itself), those who mentally prepare now for securing livelihood through active engagement or entrepreneurship will suffer less damage. I’m not suggesting you immediately file for business licenses or incorporate companies. However, if you’ve spent your career as an employee where supervisors determined your daily and weekly tasks, you must abandon this mindset. In the AI era, your responsibilities encompass only what you actively engage with. Everything people won’t spontaneously tackle will be filled by robots and AI. Thus think about yourself as a contractor who actively pitches your next week’s tasks to your superiors. (or you might find yourself watching your tasks done by AI alternative provider).
Real Estate of the AI Age
Every human epoch has defining resources. Initially land, then mineral resources and oil, later machinery, buildings, and real estate. Land ownership during feudalism provided advantages. Factory ownership during the industrial revolution yielded advantages. This raises the question: what constitutes AI-era real estate?
Currently leading AI companies offer clues to this answer. Giants like OpenAI, Google, and XAI possess abundant capital, brilliant minds, and training data for AI model development. However, their greatest limitation is graphics processors’ (GPU) computational power.
GPU computational power is up for rental already today and will likely remain available so indefinitely. But it quite resembles contemporary land ownership. Vegetable farming or house construction benefits strongly from own land ownership. Even if your GPU lies “fallow” initially for several quarters. I realize this sounds bizarre – owning a home GPU. But remember how people reacted in the 1980s: Why on Earth would anyone need a home computer? Since we’ll all consume AI computational power, owning GPUs will resemble owning wells, boilers, or solar panels. You’ll survive without it but having them you enjoy significant cost savings. An keep in mind the historic real estate lessons: when everyone wants same resource simultaneously, purchasing it becomes dramatically more expensive than pre-owning it.
Pressure on Politicians
Regardless of how rapidly and harshly the AI era unfolds, it will generate challenges only states can address by nature: mass unemployment, universal income, AI usage regulations in society. Therefore, political representation must be both knowledgeable about core problems and ACTIVELY implementing countermeasure. And shall do so even more critically than ever before.
Daily reality reveals that politicians and public officials know virtually nothing specific about artificial intelligence and show no interest in the topic (with rare exceptions of few technocratic governments). As individuals, we must contact our politicians and pressure them to make this a priority issue – not merely electoral talking points. If the tsunami of impacts on work, education, and content consumption catches us as a society with undebated options, losses will be even more severe. What’s sufficient action from your end? Write letters, emails (or carrier pigeons) to politicians you’ve supported or plan to support. Express concerns about your job, livelihood, and social recognition with regards to AI. Ask them what their solution to AI’s arrival is. Only when politicians receive five to ten such letters monthly, it pushes them toward meaningful discussion on AI. Spontaneity is not there, we need to push them.
Pilot’s announcement in nutshell
Artificial intelligence will impact me before I reach fifty. Along with me, all of you in productive years will feel these effects. Given average life expectancy (77+ years) and retirement age (65+ years), we face at least 15 years between when AI impacts start affecting us and when they become irrelevant to our lives. That’s super extended period – you cannot hide 10+ years from it, neither in job market or daily life. We must inevitably accept fundamental life changes and switch from spectator to active mode. Don’t let anyone comfort you with false assurances that this won’t affect you, that it will somehow pass you by. It’ll come soon and if we can’t secure a safe landing ourselves, gravity will gladly handle our return to earth.
Publikované dňa 14. 7. 2025.
Späť na domovskú stránku
