Who do you believe when it comes to the potential impact of artificial intelligence? An MIT Nobel laureate economist or the former CEO of the world’s biggest tech company?
MIT economist Daron Acemoglu, for his part, says the current hype is way over the top. AI might profitably automate only 5% of tasks and add just 1% to global GDP over the coming decade, he said in a recent MIT Sloan presentation. Acemoglu also asserted that AI’s potential is less clear than the internet’s was when it began to proliferate in the 1990s. Currently, he added, AI is seen as a cost‑cutting tool rather than as a new force for innovation, as was the internet in its early days.
Also: Will AI think like humans? We’re not even close – and we’re asking the wrong question
At the same time, former Google CEO and chairman Eric Schmidt believes that, if anything, we’re underplaying AI and not grasping its full, sweeping effects. “The arrival of non‑human intelligence is a very big deal,” he told a recent TED conference. He reiterated the same points at a conference in July. Essentially, AI has become adept at developing solutions to problems—solutions never considered by humans.
On the overhyped side, “the industry has not produced applications that are critical for the production process or for generating new goods and services that are going to be hugely valuable,” Acemoglu said. The rise of the commercial internet has had a more substantial impact, he added. Even the scenario in which machines will eventually perform cognitive tasks “is not so clear on how you’re going to get AI tools into the production process.”
Also: People don’t trust AI but they’re increasingly using it anyway
The current state of the art in AI is still relegated to predictive tasks that don’t require high levels of social interaction and judgment, he continued. Roughly 20% of the economy is “either in the crosshairs of AI to be automated or could be majorly boosted by AI input.”
Schmidt, on the other hand, says AI is upending even the most forward‑looking economic models. “There are assumptions that we’ll end up with something like 30% increases in productivity per year,” he pointed out. This is even beyond the scope of traditional economics, as economists have never seen such gains. And the pace is accelerating. “As this stuff happens quicker, you will forget what was true two years ago, or three years ago,” Schmidt added.
Also: 9 programming tasks you shouldn’t hand off to AI – and why
Acemoglu is more skeptical about such claims and said it will be some time before AI is ready to take over businesses in a profound way. “Do you really think in three, four, five years’ time you’re going to have an [artificial general intelligence or AGI] that can do tasks with no human supervision — all of your accounting, or all of your marketing?” he asked. “With the current architecture, the best we can do is copy human decision‑making.”
He added that “we are not developing AI in the best possible way. That best possible way is much more human, much more targeted at working with human decision makers.”
There are plenty of voices that claim that AI is different, more profound, and world‑changing than previous technology waves. Still, it’s notable to anyone familiar with the Gartner hype cycle that all ballyhooed technologies will crest at the top of the hype wave well before implementations are in place and producing — leading to a slide down to a “trough of disillusionment.” But eventually the said technology will gain ground as implementations begin to show their worth, and the ascent up the less‑sexy “plateau of productivity” commences. At that point, something else that is relatively untested is being hyped.
Tips for seeing one’s way through the current AI hype blizzard
So, what’s a technology or business professional to do?
Should you follow the hype and avoid fear of missing out (FOMO) and falling behind, or proceed cautiously? It’s safe to assume that no one is going to immediately tear down their existing systems and analytics packages in the hope of moving up the AI ladder. There are thousands of organizations that still rely on their mainframes for core business processes.
Also: How to get rid of AI Overviews in Google Search: 4 easy ways
But at the same time, it’s clear that a lot of money is flowing to AI — a fact not lost on vendors, of course. AI spending surged to $13.8 billion in 2024, more than 6× the $2.3 billion spent in 2023, according to figures from Menlo Ventures. Still, one‑third of executives in its survey don’t have a clear vision for the direction of their AI strategy.
AI is overhyped as the solution to everything, yet underhyped as an underlying platform that will power many functions in a less‑sexy, more behind‑the‑scenes way.
1. Avoid the FOMO trap
One can be forgiven for feeling as if the whole world around them is moving onto AI. Actually, everybody is still trying to learn as much as they can about the technology and may be puzzled about where it can go. Before huge sums of money are spent, decision-makers need to consider whether more traditional systems can do the job just as well.
2. Work closely with the business
Do they really need advanced AI, or would more robust predictive analytics do the trick? Involve employees in the processes of identifying areas in need of improvement, and hear out their ideas about goods and services that AI can help produce and deliver.
3. Stay familiar with new technology developments
The availability of large language models is truly astounding, with new updates, and capacities seemingly announced every week. Fortunately, these are fairly interchangeable. But remember, today’s amazing LLM may be small potatoes a year from now.
4. Keep tabs on what others are doing with the technology
Through user groups, forums, and customer gatherings, one may get a sense of where AI is making a dent in their companies’ success and where it may be a dud.
5. View AI as amplifying human skills, not replacing them
Figure out where AI is better than humans, and where humans are better than AI — not just cutting costs.
6. Develop new types of goods and services an AI system can provide
As opposed to simply cutting costs or attempting to replace human roles with chatbots. In many cases, generative AI is simply used as a new type of search engine, not as a generator of new goods and services. The world needs more health, educational, and financial services.
Want more stories about AI? AI Leaderboard, our weekly newsletter.
Leave a Reply