On Wednesday, some of the most influential and well-connected figures in global business and politics gathered in Manhattan for an event that drew the world’s attention. The focus of conversation was unmistakable: artificial intelligence. This technology, once confined to the domain of research laboratories and speculative fiction, now stands at the center of discussions shaping the global economy, innovation policy, and national competitiveness.
It has been three years since the public debut of ChatGPT—a launch that irrevocably transformed the public’s understanding of what artificial intelligence could accomplish. Its introduction marked a turning point, igniting a surge of interest and competition that has since evolved into a full-fledged technological arms race. Today, that race has reached unprecedented intensity, raising questions about whether the scale of this rapid expansion might have already exceeded reasonable bounds.
The occasion for these reflections was the New York Times DealBook Summit, an annual conference where the world’s most prominent leaders in finance, technology, and governance come together to discuss matters defining the global marketplace. Among the attendees were Larry Fink, CEO of the massive investment firm BlackRock, and Lai Ching-te, the President of Taiwan—two leaders representing both financial power and geopolitical significance. Their discussions revolved primarily around the state of the artificial intelligence sector and the possibility that an economic bubble may already be forming around it, echoing debates that have dominated conversations in both Silicon Valley and Wall Street over recent months.
The facts are striking. Industry titans such as Meta, Alphabet, and Microsoft are collectively pouring tens of billions of dollars into capital expenditures this year, much of it directed toward building and maintaining AI infrastructure—data centers, silicon chips, and computational networks capable of training and deploying increasingly sophisticated models. According to data cited by TechCrunch in November, forty-nine U.S.-based AI startups have each secured funding rounds exceeding one hundred million dollars in 2024 alone. This unrelenting flow of capital conveys immense confidence in the transformative potential of AI, yet it simultaneously fuels concerns about sustainability and risk.
While the powerful voices convened at DealBook concurred on one central point—that artificial intelligence is not a transient trend but a fixture of the future—they also acknowledged that the path forward will be uneven. Rapid technological revolutions rarely proceed without disruption, and this one will likely generate both spectacular victories and notable failures. Larry Fink captured this sentiment in conversation with DealBook founder Andrew Ross Sorkin: “There are going to be some huge winners and huge failures,” he warned, emphasizing that the coming years may yield both exhilaration and collapse. Yet, despite recognizing the peril, Fink expressed optimism that long-term market demand for AI resources remains strong. He described ongoing discussions with the major “hyperscalers”—companies like Amazon, Google, and Microsoft—that supply cloud computing capacity essential to AI’s operation, all of whom, he noted, are already confronting shortages in computational resources.
This cautious optimism resonated throughout the summit, as numerous participants—spanning industries and perspectives—echoed the theme of measured confidence paired with realism. Dario Amodei, the CEO and cofounder of Anthropic, an artificial intelligence company specializing in responsible model development, elaborated on the inherent volatility that defines the sector. According to Amodei, the enormous capital requirements tied to constructing the vast data centers that train and host models make the business unusually precarious. Even if the science delivers fully on its promises, he noted, a small misstep in timing or investment strategy could prove disastrous. In his words, “Even if the technology fulfills all its promises, I think there are players in the ecosystem who, if they just make a timing error—they just get it off by a little bit—bad things could happen.”
Amodei emphasized that Anthropic approaches these challenges through prudence and structural caution. The company mitigates exposure by forming partnerships with large enterprise customers and by expanding computational capacity conservatively rather than aggressively. By contrast, he suggested, certain competitors have taken a far more speculative approach, betting heavily on uncertain timelines. “There are some players who are YOLOing,” he remarked, using the colloquial shorthand for taking reckless, high-stakes leaps without measured foresight. Although Amodei declined to identify any specific companies, his remarks included a discreet jab at his former employer, OpenAI, and its CEO, Sam Altman. When he commented, “We don’t have to do any code reds,” he was alluding to a recent internal memo from OpenAI that framed Google’s emerging AI model as a critical competitive threat.
Yet amid all this talk of risk and potential failure, a broader consensus emerged: that such turbulence is not only inevitable but essential to progress. In the process of refining transformative technologies, some ventures will fail, but these failures serve as catalysts for improvement and discovery. Many leaders at the summit argued that success in AI development should be measured not merely by profitability but by its contribution to societal advancement and its ability to uphold a nation’s technological edge. As Fink put it, “If we don’t spend enough faster in AI, in digitization, and in tokenization, other countries are going to beat us.” His comments underscored a conviction shared by many Western executives—that strategic investment in artificial intelligence is not optional but imperative for maintaining global competitiveness.
Amodei extended this argument into the realm of national security and governance, asserting that government policy will play an indispensable role in determining how safely and effectively the technology evolves. He advocated for regulators to adopt a proactive stance, both to prevent reckless speculation and to safeguard technological sovereignty. One concrete measure he proposed was tightening restrictions on the export of high-performance Nvidia chips to China. Allowing cutting-edge AI infrastructure to be concentrated within authoritarian regimes, he warned, could have far-reaching implications: “If the best model is plopped down in an authoritarian country, I feel like they can outsmart us in every way: intelligence, defense, economic value, R&D.”
Taiwan’s President Lai Ching-te, likewise, stressed the importance of international cooperation in managing what he described as the dual challenge of harnessing AI’s potential while avoiding its catastrophic misuse. In his remarks, Lai urged leaders of every nation engaged in AI development to collaborate and to enact safeguards that will ensure sustainable growth. “Leaders around the world, especially those from countries with AI-related industries, should work together and take necessary measures to ensure AI develops sustainably and has a soft landing so that it can drive long-term global growth,” he said during the conference, framing technological responsibility as a shared moral and economic responsibility.
Lai’s perspective carried particular weight given Taiwan’s central role in the global AI supply chain. The island’s semiconductor industry—dominated by firms like TSMC—produces the highly specialized chips that power artificial intelligence systems around the world. This unique position makes Taiwan both a major beneficiary of the AI boom and a participant in the very speculative cycle of accelerated investment that many fear could lead to a bubble. Thus, even as Lai called for prudence and multilateral oversight, he did so as a stakeholder deeply embedded in the same ecosystem whose unchecked exuberance he cautioned against.
As the summit concluded, one overarching message became unmistakably clear: artificial intelligence stands at the crossroads of extraordinary promise and formidable risk. It represents the next frontier in human innovation, yet its evolution demands an unprecedented level of strategic foresight, cooperation, and ethical consideration. The leaders assembled in Manhattan did not simply celebrate AI’s ascent; they acknowledged the responsibility—and potential peril—that accompanies driving the world’s most transformative force forward.
Sourse: https://www.businessinsider.com/ceos-predict-ai-bubble-losers-dario-amodei-larry-fink-2025-12
