We are currently in a very similar “this seems overblown” phase of a transition that is arguably much larger and more permanent than what we experienced during that global health crisis. While many people have tinkered with basic artificial intelligence (AI) tools and found them mildly amusing or perhaps a bit unreliable, a massive gap has opened between public perception and the reality of what is currently happening in the laboratories of the world’s leading technology companies. For those working on the front lines of this field, the ground is already shaking. They aren’t just making predictions about a distant future; they are describing what has already happened to their own daily workflows and warning that the same shift is about to reach everyone else.
The history of AI has often been described as a series of steady, incremental improvements. There were occasional jumps, but they were spaced far enough apart that society had time to absorb the changes. However, beginning in 2025, new techniques for building these systems unlocked a pace of progress that began to accelerate exponentially. Each new version was not just slightly better than the previous one; it was better by a much wider margin, and the time between releases began to shrink.
A definitive turning point occurred in early February 2026, when two major AI laboratories released their latest models on the same day: GPT-5.3 Codex and Claude Opus 4.6. This was the moment when the “rising water” of technology reached a level where it could no longer be ignored. For professionals who had been using these tools for guidance or editing, the shift was profound. Instead of a back-and-forth process where a human guides the machine, the relationship flipped. It became possible to simply describe an outcome in plain English and walk away.
Imagine wanting to build a complex software application. In the past, this would require weeks of planning, designing, and coding by hand. Now, a user can describe the app’s purpose and general look, and the AI takes over entirely. It doesn’t just write a draft of the code; it writes tens of thousands of lines, opens the application itself, clicks through the buttons like a human user, identifies bugs, and refines the design until it meets its own standards. Only then does it present the finished, often perfect, product to the user. This isn’t just a tool executing instructions; it is an agent making intelligent decisions based on what feels like “judgment” or “taste” – the very qualities many believed a machine could never replicate.
The Reality Of The “Ancient History” Trap
A common refrain from those who are skeptical of this shift is, “I tried AI, and it wasn’t that good”. This reaction is understandable because, until recently, it was true. If you experimented with ChatGPT in 2023 or 2024, you likely encountered “hallucinations” (nonsense answers presented as facts) or mediocre output that required heavy human correction.
However, in the context of the current pace of development, 2024 is ancient history. The models currently available are unrecognizable compared to what existed even six months ago. Evaluating the state of modern AI based on an experience from two years ago is like trying to understand the potential of a smartphone by using a flip phone from 2005. Furthermore, there is a significant divide between those using the free versions of these tools and those using the paid, state-of-the-art versions. The free tiers are often more than a year behind the cutting edge, meaning most people are judging the technology based on its past, not its present.
Measuring The Acceleration
To understand why this is moving so fast, it helps to look at concrete data rather than just anecdotes. Organizations that track the capabilities of AI models use metrics based on how long it takes a human expert to complete a real-world task.
- In early 2022, AI struggled with basic arithmetic, famously claiming that 7 × 8 equals 54.
- By 2023, it was passing the bar exam.
- By 2024, it was writing functional software and explaining graduate-level science.
- By late 2025, elite engineers were handing over the majority of their coding tasks to the machine.
The length of tasks these models can successfully complete without human help has been doubling approximately every seven months, and some data suggests that this rate is accelerating to every four months. In 2022, a model might handle a task that takes a human 36 seconds. By late 2025, that jumped to tasks requiring over five hours of expert human labor. Recent models released in 2026 are already handling projects that would take a person 17 hours or more. If this trend holds – and it has shown no signs of slowing – we are looking at AI that can work independently for days within the next year, and handle month-long projects autonomously within three years.
The Intelligence Explosion: AI Building AI
The most significant and perhaps least understood development is the fact that AI is now being used to build the next generation of AI. This creates a powerful feedback loop. When the latest version of a model like GPT-5.3 Codex was released, the technical documentation revealed that the AI itself was “instrumental in creating itself”. The developers used earlier versions of the model to debug training, manage deployment, and diagnose test results.
This isn’t a theoretical prediction; it is current practice. At major AI companies, AI is already writing much of the code used to build its successors. Researchers call this an “intelligence explosion”. Each generation of AI helps build the next, which is smarter and builds the next even faster. This flywheel is spinning with immense momentum, making it increasingly difficult for human observers to keep up with the rate of change.
Which Jobs Are Most Affected?
This transition is different from previous waves of automation. When factories were automated, workers could often move into office-based administrative roles. When the internet disrupted retail, workers shifted into logistics and services. But the current wave of AI isn’t replacing a single specific skill; it is a general substitute for cognitive work. It gets better at almost everything simultaneously. This means there isn’t a “safe” harbor to retrain for, because AI is improving in those fields too.
The most safety-focused leaders in the industry have predicted that as much as 50% of entry-level white-collar jobs could be eliminated within the next one to five years. Some expect unemployment rates could eventually rival those of the Great Depression as a result of this widespread automation.
While it will take time for these changes to ripple through the global economy, the underlying capability for this disruption is already arriving. Consider these specific examples:
- Legal Work: AI can already draft briefs, summarize case law, and analyze contracts at a level that rivals or exceeds junior associates. One managing partner at a large firm described it as having a team of associates available instantly, and he expects it will soon be able to do, after years of his own experience, most of what he does.
- Software Engineering: Large portions of the job are already being automated. AI is no longer just writing snippets of code; it is managing complex, multi-day projects. This means far fewer human programming roles will likely be needed in the coming years.
- Financial Analysis: AI is becoming highly competent at building financial models, writing investment memos, and generating reports.
- Medicine: In many areas, AI is already approaching or exceeding human performance when it comes to reading medical scans, reviewing literature, and suggesting diagnoses based on lab results.
- Content and Customer Service: From marketing copy and journalism to genuinely capable AI customer service agents that can handle multi-step problems, the quality of AI output has reached a point where it is often indistinguishable from human work.
The common belief that “human judgment” or “empathy” will be a permanent shield for jobs is becoming harder to defend. Modern models are making decisions that feel like intuitive judgment, and people are already beginning to rely on AI for emotional support and advice. If a job involves reading, writing, analyzing, or communicating through a screen, significant parts of it are likely to be affected in the near term.
A Global Security Paradigm Shift
The implications of this technology extend far beyond the job market. To grasp the scale of the risk, some industry leaders use a thought experiment: imagine that by 2027, a new country appears overnight with 50 million citizens, every one of whom is smarter than any Nobel Prize winner who has ever lived. These “citizens” think 10 to 100 times faster than humans, they never sleep, and they can operate anything with a digital interface.
In such a scenario, the national security implications would be staggering. This represents a level of power that humanity may not yet be mature enough to handle. The upside is incredible: we could see a century’s worth of medical research compressed into a single decade, potentially solving cancer, Alzheimer’s, and infectious diseases within our lifetimes. However, the downside is equally real, including the potential for AI to be used in creating biological weapons or enabling unbreakable surveillance states. The people building this technology are often the ones most frightened by it, yet they believe it is too important to abandon and too powerful to stop.
Practical Strategies For Adaptation
In the face of such massive change, feeling helpless is a natural reaction, but the single biggest advantage an individual can have right now is being early to adapt. The window where most people are still ignoring these developments is a brief but valuable opportunity.
Move Beyond Basic Search: Most people treat AI like a more conversational version of Google. This is a mistake. To understand its true power, you must push it to do actual work. If you are a lawyer, don’t just ask it a research question; give it a contract and ask for a counter-proposal. If you are in finance, give it a complex spreadsheet and ask it to build a model. If you are an accountant, give it a full tax return and ask what it can find.
Use The Best Tools Available: Do not judge the technology based on free versions. Investing the $20 a month for the most advanced models (like the paid versions of ChatGPT or Claude) is essential if you want to see what is actually possible. Furthermore, check the settings to ensure you are using the most capable model, as these apps often default to faster but less intelligent versions.
The One-Hour-A-Day Rule: A simple commitment can put you ahead of 99% of the population: spend one hour every day actually using AI. Don’t just read about it or watch videos; get into the tools and try to get them to solve a problem you aren’t sure they can handle. If you do this consistently for a few months, you will have a better grasp of the coming reality than almost everyone around you.
Build The Muscle Of Adaptability: The specific AI tools of today will be obsolete in a year. The goal isn’t just to master one specific software, but to become comfortable with the pace of change itself. You must get used to being a beginner over and over again. Those who refuse to engage because they feel it diminishes their expertise or because they believe their field is “special” will be the ones who struggle the most.
Re-evaluate Long-Term Planning: The standard playbook of getting good grades to land a stable professional job is now aimed directly at the roles most exposed to AI disruption. For the next generation, the most important skills will be curiosity, adaptability, and the ability to use these tools to build things they care about. Financial resilience also takes on a new importance; it is wise to be cautious about taking on long-term debt that assumes your current career path will remain unchanged for decades.
The Closer Horizon Of Dreams
While much of the discussion around AI focuses on the threats to the status quo, there is another side to this story. The barriers to building things have largely vanished. If you have ever wanted to create an app but lacked the coding skills, you can now describe it and have a working version in an hour. If you have wanted to write a book or start a business but felt overwhelmed by the technical or administrative hurdles, those obstacles are being dismantled.
The best tutor in the world is now available to anyone for a small monthly fee – one that is infinitely patient and available 24/7. Knowledge is essentially free, and the tools to build are cheaper than they have ever been. In a world where old career paths are being disrupted, the people who thrive may be those who use this time to pursue what they are genuinely passionate about, using AI as a force multiplier for their own creativity.
Conclusion
We have moved past the point where AI is merely an interesting topic for a dinner party conversation about its future potentials. The technology works, it is improving at an exponential rate, and the most powerful institutions in history are committing trillions of dollars to its development. The “future” is now; it just hasn’t reached everyone’s doorstep yet.
The next few years are likely to be disorienting. But those who start to engage with it – not with fear, but with a sense of urgency and curiosity – will be in the best position to navigate the transition. This is a unique window of time to get a head start before the rest of the world catches up. The shift is happening, and the best strategy is to be the person who understands how to navigate the new landscape to be ahead of the curve.
Original Article link here.
Jeff Rense Commentary on Article – MP3 link here.