Artificial intelligence could trigger a new wave of dropping out growth or innovation if it was poorly managed.
Adobe Stock
For two centuries, technology has transformed our way of living and working, from the steam to the Internet. It is tempting to assume that progress will always continue. However, the story offers a lesson that gives to think: a big step forward often stop when institutions do not adapt. Economist Carl Benedikt Frey, author of “How progression ends“Rights that artificial intelligence could either feed a new economic boom, or bring innovation to stop, according to how we react today.
Why progress can end
Frey maintains that the idea of inevitable technological progress is deeply imperfect. “If progress was inevitable, the first industrial revolution would have occurred much earlier,” he said in our recent conversation. “And if progress was inevitable, most countries in the world would be rich and prosperous today.” Many companies have experienced intense innovation periods followed by stagnation or collapse. Ancient cities like Ephesus have once prospered and then disappeared. The Soviet Union has industrialized quickly but failed to follow the start of the computer era.
The heart of the problem lies in the transition of societies between two distinct innovation phases. Frey describes an early “exploration” phase, during which new ideas are emerging in a highly decentralized environment where experimentation prosperous. Successful savings are later transformed in “exploitation”, scaling technologies, costs reduction and power consolidation. “Exploration prosperous on decentralization,” he said. “The exploitation prosperous on concentration and consolidation. But then, to access the following cycle, you again need decentralization. And this transition is very difficult to do. ”
Lessons for the age of AI
Artificial intelligence is squarely at the center of this fragile transition. The first breakthroughs, transformers at generative AI, come from an experimentation open in universities and small laboratories. However, market power is quickly concentrated. Frey notes that “Openai with Microsoft has something like 70% of the market”, and many major technological holders invest in startups that could otherwise challenge them. A return to consolidation could slow down the next wave of breakthroughs and transform AI into a tool of incremental efficiency rather than real transformation.
Frey is also critical by assuming that the scaling of existing models will provide an intelligence at the level of man. “Many people always thought you could simply take existing models and make them evolve,” he said. “It could work in a static world, but the world is changing all the time.” He underlines how even superhuman systems can fail in the face of unknown situations, citing how amateur players using laptops have recently defeated TOP GO programs by presenting new strategies. In order for AI to continue to move forward, we need new approaches and a competitive and decentralized innovation environment.
Regulations and risk of strangulation of progress
One of Frey’s strongest warnings concerns the regulations. He recognizes that some railings are necessary, for example, around training data and harmful applications. However, he believes that too complex rules can involuntarily crush smaller innovators and anchor the largest companies. “The more obstacles to the entrance, the more we add the costs of compliance, the more the risk of a few companies or even a company monopolizing technology,” he said.
He underlines the experience of Europe with the laws on data confidentiality as an example. “Take well -intentioned regulation like the GDPR,” said Frey. “Good reasons to worry about the confidentiality of the data, yes, but we must examine the involuntary consequences. Large technological companies have essentially been able to compensate for compliance costs by capturing a larger share of the market, where some small businesses had trouble competing. ” The risks of the AI Act proposed by Europe to repeat this error. If the development of AI becomes as expensive and bureaucratic as pharmaceutical products, innovation will slow down and power will be more consolidated.
Centralization, decentralization and global competition
The tension between centralization and decentralization also shapes the geopolitical race of AI. Historically, the United States has led because it encouraged competition and flexible institutions. Government funding supported exploration while antitrust actions have opened markets for new entrants. This allowed companies like Microsoft and a generation of internet innovators to emerge. Japan has excelled in refining and scaling existing technologies; However, its highly centralized business system has left it poorly positioned for the software revolution.
Frey fears that America will lose this advantage. “The autonomy of American universities is being reduced, the financing of the National Science Foundation Cup, the distribution of access to talents and immigration,” he said. “I am much more worried about America than when I wrote the book.” Meanwhile, China centralizes power and favors national security on growth, which also risks stagnation. The result is a difficult breed where none of the two parties can maintain the conditions for a sustained innovation.
The real opportunity: beyond automation
Many organizations use AI mainly for process automation and cost reduction. Frey thinks that this will not produce transformative growth. “If AI means that we make emails and calculation sheets a little more effectively and that we facilitate the way in which we reserve travel, the transformation will not be equal with electricity or the internal combustion engine,” he said. True prosperity comes from the creation of new industries and the fact of doing inconceivable things before. Managers should encourage their teams to experiment with AI to create new products and services, not only rationalizing those existing.
This requires giving employees at all levels autonomy to test and implement AI solutions. “People who experiment best understand what technology can and cannot be used,” said Frey. Companies that reward experimentation and decentralize decision -making are better placed to overcome the Wave of AI.
Inclusion and future innovation pipeline
Frey also highlights the importance of inclusion. Half of the inventive potential of humanity remains underused when large groups are excluded from innovation. “If you take the fraction of the female and male inventors in the United States, the gap, the gap, but at the rate of the closure, it would take one hundred and twenty years to close,” he noted. Role models and various networks are crucial to attract more people to science and technology. A wider participation makes innovation more resilient and more adaptable.
Uncertain
The future impact of AI is unpredictable, but Frey believes that leadership decisions will shape today if it accelerates growth or traps us in stagnation. He warns against excessive confidence that technology will simply do things. Progress has already ended when companies have failed to adapt their institutions and their state of mind.
It remains cautiously optimistic that we can promote open competition, avoid heavy regulations and allow people to experiment. “If you want to prosper as a business in the AI revolution, you must give people at low level of the organization a greater decision-making autonomy to implement the improvements they find for themselves,” he said.
The message is clear: AI could be the most powerful growth engine in generations, but this result is far from being assured. History shows that technological waves can stall when power centralizes, innovation shrinks and the regulations stifle competition. The leaders who embrace the experimentation, support diversity and are concerned against monopolization have the best chance of ensuring that the era of AI feeds a new cycle of progress rather than putting it at the beginning.
