Just like the internet, personal computing and smartphones have defined the decades that came before, AI is quickly becoming the defining movement of the 2020s, but yet nearly four years after ChatGPT went public it’s still one of the most misunderstood technologies.
Depending on who you speak to in an organization, this broad term could mean anything from simple productivity boosts to full-on agentic automation, and it’s left company boardrooms debating over how it should be used and how realistic ROIs are in the immediate future.
Even so, that hasn’t stopped companies from investing heavily, and many now have some form of AI strategy in place.
But the world’s largest companies, including Nvidia which has grown 1,110% in five years to become the world’s most valuable company, have collectively added trillions in market value as the world bets on AI’s transformative potential.
That scale has drawn comparisons with previous tech bubbles – soaring valuations, aggressive and non-stop spending, and an increasingly crowded marketplace has me worried we could be living in another bubble, and for it to burst could be catastrophic.
Are we living in an AI bubble? What happens next?
Longer term, I also worry about what might happen when AI succeeds. Will it replace humans?
This discussion is becoming more relevant as AI systems evolve from one-way generative machines to full autonomous agents capable of writing, coding, analyzing and more with little to no human input.
There’s also the sheer pace that AI is evolving at, and we already know that even the biggest enterprises are struggling to keep up.
To explore whether the AI boom is being driven by genuine transformation potential or just excessive hype, I spoke with GoodData.AI CTO Peter Fedorocko about the growing bubble-associated risks, the role of humans and how enterprises should tackle their own strategies.
- Peter, let’s start by a blunt question. As of 2026, the magnificent 7 have a combined market value that’s bigger than the annual GDP of the EU. Are we therefore in an AI-induced bubble?
It’s a fair question and I won’t dodge it, the valuations are eye-watering. But I’d separate market exuberance from underlying value. Yes, there is speculative froth at the edges, there always is with transformative technology, like we saw with the internet in 2000.
The excitement or ‘bubble’ as you call it behind AI is mostly being caused by the impact of seeing how tech is being disrupted and then the predictions that AI will disrupt and spill over into other fields, which again is similar to the internet ‘bubble’.
The internet bubble burst because the infrastructure wasn’t ready for the promises being made. AI infrastructure, the compute, data, and models, are here and they work.
That doesn’t mean every company with “AI” in its pitch deck deserves its valuation. Many don’t. But the core technology? That’s not going anywhere, so call it a bubble if you want, but just don’t confuse the hype layer with the foundation underneath it.
- Sam Altman says that humans won’t be replaced by AI. You agree with him. The cynic in me, however, would argue the contrary and point out that AI is a very, very different beast from say, the internet or the personal computer. Can you succinctly explain your thought process?
The cynics have a point worth taking seriously because AI is different. Previous technological shifts automated physical or repetitive tasks. AI can reason, write, code, and make decisions, it’s adding intelligence to this automation, and I get why that feels more threatening.
But here’s my thought process: what AI still can’t do is want things. It has no stake in the outcome.
Businesses are systems of tasks, relationships, judgment calls, and accountability. Someone has to own the decision, carry the responsibility, and look a client in the eye. AI can supercharge the person doing that, and it is and will continue to, but it cannot be that person.
People will always want to resume control and talk to people. That’s not a limitation that we’ll engineer away, it’s fundamental to how human organisations actually function.
AI is just an efficiency amplifier, I’m not 100% sure on what new job titles AI will create but it likely will do.
- I would agree that AI seeks first and foremost to improve efficiency – ideally reaching 100%. The problem is capitalism – in the western world – depends entirely on inefficiencies. The more you make our society “efficient”, the more unequal it will become and the more obsolete and fragile our capitalism becomes. Do you disagree with that argument?
I disagree with the framing, but not entirely with the concern. Efficiency gains don’t automatically produce inequality, but the distribution of those gains does. That’s a policy and ownership question, not a technology question.
The printing press made information radically more efficient to distribute. It didn’t have to produce monopolies on knowledge, but in some cases it did, because of how power was structured around it.
AI is the same. The technology itself is neutral on that question. Where I’d push back harder is on the idea that capitalism requires inefficiency to survive, what it actually requires is demand, and demand comes from people having purchasing power and meaningful work.
If AI genuinely eliminates entire categories of jobs without creating new ones, that’s a demand crisis, not an efficiency triumph. Smart businesses understand that. The ones racing purely to cut headcount are making a short-term bet that I think they’ll regret.
- Perhaps the problem stems from how AI is being defined. Right now, this seems to include everything from blunt machine learning to Artificial Super Intelligence. How important do you think AI literacy is in light of the current debate?
Critically important, and it’s one of the most underrated problems in this entire conversation. When a journalist writes about AI taking jobs and a researcher writes about AI taking jobs, they can mean completely different things.
One might mean a rules-based automation tool, the other might mean something approaching general reasoning. The conversation then gets polluted.
Boardrooms are making billion-dollar decisions based on a vague concept they haven’t properly defined. I see it constantly. Companies come to us having committed to an “AI strategy” without being able to articulate what problem they’re actually solving or what type of AI is even relevant to their situation.
AI literacy shouldn’t be considered a nice to have, it’s the difference between genuinely transforming your business and burning money on a trend. We need to be far more precise in how we talk about this, especially in public discourse.
Also, brands are brainwashing with AI everywhere which adds gas to the fire because customers get more confused and lose the appreciation or trust in real AI that adds value. It’s a mixture of education and better regulation of what is AI and what is generated by AI vs human.
- Change, as they say, is the only constant in life (predating taxes and death). 4 years ago, ChatGPT revolutionized the way humanity interacted with AI. And just over the past 10 months, we had six (yes six) major iterations of ChatGPT. If ChatGPT is going that quick, how can enterprises cope with that rate of change?
They can’t, and they shouldn’t try to. Chasing every model release is a fool’s errand and a fast track to AI fatigue inside your organisation.
What enterprises need to do is focus on understanding what AI really is; intelligent process automation.
They also need to understand their processes deeply and where AI can be applied and then think of the economic value of this application. Not get excited about every model release, every new framework being built, and prematurely optimise and invest into the AI-ification of everything.
At the end of the day it will be about economic value as we are already starting to hear more and more about the cost side.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
https://cdn.mos.cms.futurecdn.net/QAVMXXwGtQk9mDz6SgXdeh-1600-80.jpg
Source link
desire.athow@futurenet.com (Desire Athow)
