Reading the mixed reviews on GPT-5 and the endless debates around the “intelligence” of today’s GenAI tools, I asked myself if we are in an AI bubble as many people including opinion leaders believe. My answer: probably not. But if someone asks whether there is an AGI (Artificial General Intelligence, i.e. human-level AI) bubble, I would say probably yes.
Why? Because many startups and labs are still promising AGI—or even ASI (Artificial Super Intelligence, i.e. AI smarter than humans)—simply by scaling up transformer-based AI systems with massive amounts of money and compute. In my opinion that can be a risky bet for investors.
On the other hand, AI—including GenAI—is already delivering business value. I see this not only in analyst reports but also in my everyday work, which is increasingly augmented by AI tools. Even as an AI professional who educates clients (and sometimes friends and family) about AI, I had to learn—and am still learning—how to make tools like Microsoft Copilot, ChatGPT’s advanced features, or Perplexity Pro really useful at work.
Of course, understanding how to use AI and GenAI effectively—especially in innovative ways—is much harder for businesses. Large enterprises and their leaders often face the steepest learning curve. Yet the potential rewards are significant, and in many cases, could be transformative.
That is why I suggest looking past the AGI hype and instead focus on how to learn, adopt, and implement AI in practice—for yourself, your business, and your organisation.