A Sputnik moment for AI?
After the release of DeepSeek-R1 on January 20 triggered a massive drop in chipmaker Nvidia's share price and sharp declines in various other tech companies' valuations, some declared this a 'Sputnik moment' in the Sino-American race for supremacy in artificial intelligence. While America's AI industry arguably needed shaking up, the episode raises some difficult questions.
The US tech industry's investments in AI have been massive, with Goldman Sachs estimating that 'mega tech firms, corporations, and utilities are set to spend around $1 trillion on capital expenditures in the coming years to support AI.' Yet for a long time, many observers, including me, have questioned the direction of AI investment and development in the United States.
With all the leading companies following essentially the same playbook (though Meta has differentiated itself slightly with a partly open-source model), the industry seems to have put all its eggs in the same basket. Without exception, US tech companies are obsessed with scale. Citing yet-to-be-proven 'scaling laws, ' they assume that feeding ever more data and computing power into their models is the key to unlocking ever-greater capabilities. Some even assert that 'scale is all you need.'
Before January 20, US companies were unwilling to consider alternatives to foundation models pretrained on massive data sets to predict the next word in a sequence. Given their priorities, they focused almost exclusively on diffusion models and chatbots aimed at performing human (or human-like) tasks. And though DeepSeek's approach is broadly the same, it appears to have relied more heavily on reinforcement learning, mixture-of-experts methods (using many smaller, more efficient models), distillation, and refined chain-of-thought reasoning. This strategy reportedly allowed it to produce a competitive model at a fraction of the cost.
Although there is some dispute about whether DeepSeek has told us the whole story, this episode has exposed 'groupthink' within the US AI industry. Its blindness to alternative, cheaper, more promising approaches, combined with hype, is precisely what Simon Johnson and I predicted in Power and Progress, which we wrote just before the generative-AI era began. The question now is whether the US industry has other, even more dangerous blind spots. For example, are the leading US tech companies missing an opportunity to take their models in a more 'pro-human direction'? I suspect that the answer is yes, but only time will tell.
Then there is the question of whether China is leapfrogging the US. If so, does this mean that authoritarian, top-down structures (what James A Robinson and I have called 'extractive institutions') can match or even outperform bottom-up arrangements in driving innovation?
My bias is to think that top-down control hampers innovation, as Robinson and I argued in Why Nations Fail. While DeepSeek's success appears to challenge this claim, it is far from conclusive proof that innovation under extractive institutions can be as powerful or as durable as under inclusive institutions. After all, DeepSeek is building on years of advances in the US (and some in Europe). All its basic methods were pioneered in the US. Mixture-of-experts models and reinforcement learning were developed in academic research institutions decades ago; and it was US Big Tech firms that introduced transformer models, chain-of-thought reasoning, and distillation.
What DeepSeek has done is demonstrate success in engineering: combining the same methods more effectively than US companies did. It remains to be seen whether Chinese firms and research institutions can take the next step of coming up with game-changing techniques, products, and approaches of their own.
Moreover, DeepSeek seems to be unlike most other Chinese AI firms, which generally produce technologies for the government or with government funding. If the company (which was spun out of a hedge fund) was operating under the radar, will its creativity and dynamism continue now that it is under the spotlight? Whatever happens, one company's achievement cannot be taken as conclusive evidence that China can beat more open societies at innovation.
Another question concerns geopolitics. Does the DeepSeek saga mean that US export controls and other measures to hold back Chinese AI research failed? The answer here is also unclear. While DeepSeek trained its latest models (V3 and R1) on older, less powerful chips, it may still need the most powerful chips to achieve further advances and to scale up.
Nonetheless, it is clear that America's zero-sum approach was unworkable and ill advised. Such a strategy makes sense only if you believe that we are heading toward artificial general intelligence (models that can match humans on any cognitive task), and that whoever gets to AGI first will have a huge geopolitical advantage. By clinging to these assumptions – neither of which is necessarily warranted – we have prevented fruitful collaboration with China in many areas. For example, if one country produces models that increase human productivity or help us regulate energy better, such innovation would be beneficial to both countries, especially if it is widely used.
Like its American cousins, DeepSeek does aspire to develop AGI, and creating a model that is significantly cheaper to train could be a game changer. But bringing down development costs with known methods will not miraculously get us to AGI in the next few years. Whether near-term AGI is achievable remains an open question (and whether it is desirable is even more debatable).
Even if we do not yet know all the details about how DeepSeek developed its models or what its apparent achievement means for the future of the AI industry, one thing seems clear: a Chinese upstart has punctured the tech industry's obsession with scale and may have even shaken it out of its complacency. — Project Syndicate, 2025
The writer is a 2024 Nobel laureate in economics and Institute Professor of Economics at MIT, is a co-author of Why Nations Fail: The Origins of Power, Prosperity and Poverty
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Observer
a day ago
- Observer
AI's arrival at work reshaping employers' hunt for talent
TOM BARFIELD Predictions of imminent AI-driven mass unemployment are likely overblown, but employers will seek workers with different skills as the technology matures, a top executive at global recruiter ManpowerGroup said at Paris's Vivatech trade fair. The world's third-largest staffing firm by revenue ran a startup contest at Vivatech in which one of the contenders was building systems to hire out customisable autonomous AI 'agents', rather than humans. Their service was reminiscent of a warning last month from Dario Amodei, head of American AI giant Anthropic, that the technology could wipe out half of entry-level white-collar jobs within one to five years. For ManpowerGroup, AI agents are "certainly not going to become our core business any time soon," the company's Chief Innovation Officer Tomas Chamorro-Premuzic said. "If history shows us one thing, it's most of these forecasts are wrong." An International Labour Organization (ILO) report published in May found that around "one in four workers across the world are in an occupation with some degree of exposure" to generative AI models' capabilities. "Few jobs are currently at high risk of full automation," the ILO added. But the UN body also highlighted "rapid expansion of AI capabilities since our previous study" in 2023, including the emergence of 'agentic' models more able to act autonomously or semi-autonomously and use software like web browsers and email. Soft skills Chamorro-Premuzic predicted that the introduction of efficiency-enhancing AI tools would put pressure on workers, managers and firms to make the most of the time they will save. "If what happens is that AI helps knowledge workers save 30, 40, maybe 50 per cent of their time, but that time is then wasted on social media, that's not an increase in net output," he said. Adoption of AI could give workers "more time to do creative work" -- or impose "greater standardisation of their roles and reduced autonomy," the ILO said. There's general agreement that interpersonal skills and an entrepreneurial attitude will become more important for knowledge workers as their daily tasks shift towards corralling AIs. Employers identified ethical judgement, customer service, team management and strategic thinking as top skills AI could not replace in a ManpowerGroup survey of over 40,000 employers across 42 countries published this week. Nevertheless, training that adopts those new priorities has not increased in step with AI adoption, Chamorro-Premuzic lamented. "For every dollar you invest in technology, you need to invest eight or nine on HR, culture transformation, change management," he said. Tomas Chamorro-Premuzic, Chief Innovation Officer AI hiring AI? One of the areas where AI is transforming the world of work most rapidly is ManpowerGroup's core business of recruitment. But here candidates are adopting the tools just as quickly as recruiters and companies, disrupting the old way of doing things from the bottom up. "Candidates are able to send 500 perfect applications in one day, they are able to send their bots to interview, they are even able to game elements of the assessments," Chamorro-Premuzic said. That extreme picture was not borne out in a survey of over 1,000 job seekers released this week by recruitment platform TestGorilla, which found just 17 per cent of applicants admitting to cheating on tests, and only some of those to using AI. Jobseekers' use of consumer AI tools meets recruiters doing the same. The same TestGorilla survey found almost two-thirds of the more-than-1,000 hiring decision-makers polled used AI to generate job descriptions and screen applications. Where employers today are focused on candidates' skills over credentials, Chamorro-Premuzic predicted that "the next evolution is to focus on potential, not even skills even if I know the skills you bring to the table today, they might be obsolete in six months." — AFP
Times of Oman
2 days ago
- Times of Oman
Indian consumer durable companies likely to see moderate growth in this quarter: Goldman Sachs
New Delhi: Consumer durable companies in India are expected to report moderate growth in the first quarter of the financial year 2025-26 (April-June), according to a recent report by Goldman Sachs. The report mentioned that the overall growth for consumer durables and electrical companies is likely to slow down in the current quarter. It stated "We expect consumer durables and electricals companies' growth to moderate this quarter. While B2B categories have a better outlook". While the business-to-business (B2B) categories are performing better due to strong government capital expenditure in April and continued work on major government-led projects like PM-KUSUM, RDSS, and Bharatnet, the business-to-consumer (B2C) categories are expected to see slower growth during this period. According to the report, demand from B2C segments is still weak, and the real estate-driven demand has not yet picked up. This is especially visible in the cooling products category like air-conditioners and refrigerators, where sales have been on the lower side. As a result, the overall growth expectations for companies covered under consumer durables have come down for this quarter. Goldman Sachs added that although some positive macroeconomic factors like lower inflation, tax cuts, and favourable interest rates are present, their impact on demand and consumer sentiment is likely to be seen only in the second half of FY26. The report also highlighted that companies are facing margin pressures due to negative operating leverage, which means their fixed costs remain high while revenue growth is not strong enough to offset them. However, this margin pressure is being partially managed by lower commodity prices and selective price increases in products. In terms of raw material prices, copper and aluminium have seen a month-on-month increase in May and June. This may lead to some restocking by companies.
Observer
2 days ago
- Observer
Oil prices jump after Israel's attack on Iran
BEIJING:Oil prices surged on Thursday after Israel said it attacked Iranian nuclear sites in Natanz and Arak overnight, as investors grappled with fears of a broader conflict in the Middle East that could disrupt crude supplies. Brent crude futures rose 88 cents, or 1.15%, to $77.58 a barrel by 0708 GMT, after gaining 0.3% in the previous session when high volatility saw prices fall as much as 2.7%. US West Texas Intermediate crude for July rose $1.11, or 1.48%, to $76.25 a barrel, after settling up 0.4% in the previous session when it dropped as much as 2.4%. The July contract expires on Friday. The more active August contract rose 92 cents, or 1.25%, to $74.42 a barrel. There is still a "healthy risk premium baked into the price as traders await to see whether the next stage of the Israel-Iran conflict is a US strike or peace talks," said Tony Sycamore, market analyst at IG. Goldman Sachs on Wednesday said a geopolitical risk premium of about $10 a barrel is justified, given lower Iranian supply and risk of wider disruption that could push Brent crude above $90. Trump on Wednesday told reporters that he may or may not decide whether the US will join Israel in its attacks on Iran. The conflict stretched into its seventh day on Thursday. The risk of major energy disruptions will rise if Iran feels existentially threatened, and US entry into the conflict could trigger direct attacks on tankers and energy infrastructure, said RBC Capital's analyst Helima Croft. Iran is the third-largest producer among members of the Organization of the Petroleum Exporting Countries, extracting about 3.3 million barrels per day (bpd) of crude oil. About 19 million bpd of oil and oil products move through the Strait of Hormuz along Iran's southern coast, and there is widespread concern the fighting could disrupt trade flows. Separately, the US Federal Reserve kept its interest rates steady on Wednesday but pencilled in two cuts by the end of the year. Chair Jerome Powell said cuts would be "data-dependent." — Reuters
