arXiv:2512.11912v1 Announce Type: new Abstract: A systematic, comparative investigation into the effects of low-quality data reveals a stark spectrum of robustness across modern probabilistic models. We find that autoregressive language models, from token prediction to sequence-to-sequence tasks, are remarkably resilient (for GPT-2, test NLL increases modestly from 2.87 to 3.59 despite 50% token corruption). By contrast, under the same levels of data corruption, class-conditional diffusion models degrade catastrophically (image-label consistency plummets by 56.81% relative to baseline), while classifiers show a moderate impact that diminishes with dataset scale. To explain these discrepancies, we analyze the results through a multi-perspective lens, integrating information theory, PAC learning, and gradient dynamics. These analyses suggest that robustness is heavily influenced by two key principles: the richness of conditioning information, which constrains the learning problem, and the absolute information content of the training data, which allows the signal from correct information to dominate statistical noise.

arXiv:2512.12413v1 Announce Type: new Abstract: Generative AI tools are increasingly embedded in everyday work and learning, yet their fluency, opacity, and propensity to hallucinate mean that users must critically evaluate AI outputs rather than accept them at face value. The present research conceptualises critical thinking in AI use as a dispositional tendency to verify the source and content of AI-generated information, to understand how models work and where they fail, and to reflect on the broader implications of relying on AI. Across six studies (N = 1365), we developed and validated the 13-item critical thinking in AI use scale and mapped its nomological network. Study 1 generated and content-validated scale items. Study 2 supported a three-factor structure (Verification, Motivation, and Reflection). Studies 3, 4, and 5 confirmed this higher-order model, demonstrated internal consistency and test-retest reliability, strong factor loadings, sex invariance, and convergent and discriminant validity. Studies 3 and 4 further revealed that critical thinking in AI use was positively associated with openness, extraversion, positive trait affect, and frequency of AI use. Lastly, Study 6 demonstrated criterion validity of the scale, with higher critical thinking in AI use scores predicting more frequent and diverse verification strategies, greater veracity-judgement accuracy in a novel and naturalistic ChatGPT-powered fact-checking task, and deeper reflection about responsible AI. Taken together, the current work clarifies why and how people exercise oversight over generative AI outputs and provides a validated scale and ecologically grounded task paradigm to support theory testing, cross-group, and longitudinal research on critical engagement with generative AI outputs.

arXiv:2512.12548v1 Announce Type: new Abstract: Patch foraging involves the deliberate and planned process of determining the optimal time to depart from a resource-rich region and investigate potentially more beneficial alternatives. The Marginal Value Theorem (MVT) is frequently used to characterize this process, offering an optimality model for such foraging behaviors. Although this model has been widely used to make predictions in behavioral ecology, discovering the computational mechanisms that facilitate the emergence of optimal patch-foraging decisions in biological foragers remains under investigation. Here, we show that artificial foragers equipped with learned world models naturally converge to MVT-aligned strategies. Using a model-based reinforcement learning agent that acquires a parsimonious predictive representation of its environment, we demonstrate that anticipatory capabilities, rather than reward maximization alone, drive efficient patch-leaving behavior. Compared with standard model-free RL agents, these model-based agents exhibit decision patterns similar to many of their biological counterparts, suggesting that predictive world models can serve as a foundation for more explainable and biologically grounded decision-making in AI systems. Overall, our findings highlight the value of ecological optimality principles for advancing interpretable and adaptive AI.

arXiv:2512.12324v1 Announce Type: cross Abstract: The rapid proliferation of Artificial Intelligence Generated Content has precipitated a crisis of trust and urgent regulatory demands. However, existing identification tools suffer from fragmentation and a lack of support for visible compliance marking. To address these gaps, we introduce the \textbf{UniMark}, an open-source, unified framework for multimodal content governance. Our system features a modular unified engine that abstracts complexities across text, image, audio, and video modalities. Crucially, we propose a novel dual-operation strategy, natively supporting both \emph{Hidden Watermarking} for copyright protection and \emph{Visible Marking} for regulatory compliance. Furthermore, we establish a standardized evaluation framework with three specialized benchmarks (Image/Video/Audio-Bench) to ensure rigorous performance assessment. This toolkit bridges the gap between advanced algorithms and engineering implementation, fostering a more transparent and secure digital ecosystem.

arXiv:2512.12500v1 Announce Type: cross Abstract: Artificial intelligence (AI) is increasingly permeating healthcare, from physician assistants to consumer applications. Since AI algorithm's opacity challenges human interaction, explainable AI (XAI) addresses this by providing AI decision-making insight, but evidence suggests XAI can paradoxically induce over-reliance or bias. We present results from two large-scale experiments (623 lay people; 153 primary care physicians, PCPs) combining a fairness-based diagnosis AI model and different XAI explanations to examine how XAI assistance, particularly multimodal large language models (LLMs), influences diagnostic performance. AI assistance balanced across skin tones improved accuracy and reduced diagnostic disparities. However, LLM explanations yielded divergent effects: lay users showed higher automation bias - accuracy boosted when AI was correct, reduced when AI erred - while experienced PCPs remained resilient, benefiting irrespective of AI accuracy. Presenting AI suggestions first also led to worse outcomes when the AI was incorrect for both groups. These findings highlight XAI's varying impact based on expertise and timing, underscoring LLMs as a "double-edged sword" in medical AI and informing future human-AI collaborative system design.

arXiv:2512.12620v1 Announce Type: cross Abstract: We study syllogistic reasoning in LLMs from the logical and natural language perspectives. In process, we explore fundamental reasoning capabilities of the LLMs and the direction this research is moving forward. To aid in our studies, we use 14 large language models and investigate their syllogistic reasoning capabilities in terms of symbolic inferences as well as natural language understanding. Even though this reasoning mechanism is not a uniform emergent property across LLMs, the perfect symbolic performances in certain models make us wonder whether LLMs are becoming more and more formal reasoning mechanisms, rather than making explicit the nuances of human reasoning.

arXiv:2512.12950v1 Announce Type: cross Abstract: Accurately mapping legal terminology across languages remains a significant challenge, especially for language pairs like Chinese and Japanese, which share a large number of homographs with different meanings. Existing resources and standardized tools for these languages are limited. To address this, we propose a human-AI collaborative approach for building a multilingual legal terminology database, based on a multi-agent framework. This approach integrates advanced large language models and legal domain experts throughout the entire process-from raw document preprocessing, article-level alignment, to terminology extraction, mapping, and quality assurance. Unlike a single automated pipeline, our approach places greater emphasis on how human experts participate in this multi-agent system. Humans and AI agents take on different roles: AI agents handle specific, repetitive tasks, such as OCR, text segmentation, semantic alignment, and initial terminology extraction, while human experts provide crucial oversight, review, and supervise the outputs with contextual knowledge and legal judgment. We tested the effectiveness of this framework using a trilingual parallel corpus comprising 35 key Chinese statutes, along with their English and Japanese translations. The experimental results show that this human-in-the-loop, multi-agent workflow not only improves the precision and consistency of multilingual legal terminology mapping but also offers greater scalability compared to traditional manual methods.

arXiv:2512.13438v1 Announce Type: cross Abstract: While Large Language Model (LLM) agents show great potential for automated UI navigation such as automated UI testing and AI assistants, their efficiency has been largely overlooked. Our motivating study reveals that inefficient UI representation creates a critical performance bottleneck. However, UI representation optimization, formulated as the task of automatically generating programs that transform UI representations, faces two unique challenges. First, the lack of Boolean oracles, which traditional program synthesis uses to decisively validate semantic correctness, poses a fundamental challenge to co-optimization of token efficiency and completeness. Second, the need to process large, complex UI trees as input while generating long, compositional transformation programs, making the search space vast and error-prone. Toward addressing the preceding limitations, we present UIFormer, the first automated optimization framework that synthesizes UI transformation programs by conducting constraint-based optimization with structured decomposition of the complex synthesis task. First, UIFormer restricts the program space using a domain-specific language (DSL) that captures UI-specific operations. Second, UIFormer conducts LLM-based iterative refinement with correctness and efficiency rewards, providing guidance for achieving the efficiency-completeness co-optimization. UIFormer operates as a lightweight plugin that applies transformation programs for seamless integration with existing LLM agents, requiring minimal modifications to their core logic. Evaluations across three UI navigation benchmarks spanning Android and Web platforms with five LLMs demonstrate that UIFormer achieves 48.7% to 55.8% token reduction with minimal runtime overhead while maintaining or improving agent performance. Real-world industry deployment at WeChat further validates the practical impact of UIFormer.

arXiv:2511.17729v3 Announce Type: replace Abstract: We present M^3-Bench, the first benchmark for evaluating multimodal tool use under the Model Context Protocol. The benchmark targets realistic, multi-hop and multi-threaded workflows that require visual grounding and textual reasoning, cross-tool dependencies, and persistence of intermediate resources across steps. We introduce a similarity-driven alignment that serializes each tool call, embeds signatures with a sentence encoder, and performs similarity-bucketed Hungarian matching to obtain auditable one-to-one correspondences. On top of this alignment, we report interpretable metrics that decouple semantic fidelity from workflow consistency. The benchmark spans 28 servers with 231 tools, and provides standardized trajectories curated through an Executor & Judge pipeline with human verification; an auxiliary four large language models (LLMs) judge ensemble reports end-task Task Completion and information grounding. Evaluations of representative state-of-the-art Multimodal LLMs (MLLMs) reveal persistent gaps in multimodal MCP tool use, particularly in argument fidelity and structure consistency, underscoring the need for methods that jointly reason over images, text, and tool graphs. Our Benchmark's anonymous repository is at https://github.com/EtaYang10th/Open-M3-Bench

arXiv:2503.11950v4 Announce Type: replace-cross Abstract: The increasing integration of artificial intelligence (AI) in digital ecosystems has reshaped privacy dynamics, particularly for young digital citizens navigating data-driven environments. This study explores evolving privacy concerns across three key stakeholder groups-young digital citizens, parents/educators, and AI professionals-and assesses differences in data ownership, trust, transparency, parental mediation, education, and risk-benefit perceptions. Employing a grounded theory methodology, this research synthesizes insights from key participants through structured surveys, qualitative interviews, and focus groups to identify distinct privacy expectations. Young digital citizens emphasized autonomy and digital agency, while parents and educators prioritized oversight and AI literacy. AI professionals focused on balancing ethical design with system performance. The analysis revealed significant gaps in transparency and digital literacy, underscoring the need for inclusive, stakeholder-driven privacy frameworks. Drawing on comparative thematic analysis, this study introduces the Privacy-Ethics Alignment in AI (PEA-AI) model, which conceptualizes privacy decision-making as a dynamic negotiation among stakeholders. By aligning empirical findings with governance implications, this research provides a scalable foundation for adaptive, youth-centered AI privacy governance.

arXiv:2505.05019v2 Announce Type: replace-cross Abstract: The generation of synthetic clinical trial data offers a promising approach to mitigating privacy concerns and data accessibility limitations in medical research. However, ensuring that synthetic datasets maintain high fidelity, utility, and adherence to domain-specific constraints remains a key challenge. While hyperparameter optimization (HPO) improves generative model performance, the effectiveness of different optimization strategies for synthetic clinical data remains unclear. This study systematically evaluates four HPO objectives across nine generative models, comparing single-metric to compound metric optimization. Our results demonstrate that HPO consistently improves synthetic data quality, with Tab DDPM achieving the largest relative gains, followed by TVAE (60%), CTGAN (39%), and CTAB-GAN+ (38%). Compound metric optimization outperformed single-metric objectives, producing more generalizable synthetic datasets. Despite improving overall quality, HPO alone fails to prevent violations of essential clinical survival constraints. Preprocessing and postprocessing played a crucial role in reducing these violations, as models lacking robust processing steps produced invalid data in up to 61% of cases. These findings underscore the necessity of integrating explicit domain knowledge alongside HPO to generate high-quality synthetic datasets. Our study provides actionable recommendations for improving synthetic data generation, with future work needed to refine metric selection and validate findings on larger datasets.

arXiv:2507.06268v3 Announce Type: replace-cross Abstract: Information technology is in the midst of a revolution in which omnipresent data collection and machine learning are impacting the human world as never before. The word ``intelligence'' is being used as a North Star for the development of this technology, with human cognition viewed as a baseline. This view neglects the fact that humans are social animals and that much of our intelligence is social and cultural in origin. Moreover, failing to properly situate aspects of intelligence at the social level contributes to the treatment of the societal consequences of technology as an afterthought. The path forward is not merely more data and compute, and not merely more attention paid to cognitive or symbolic representations, but a thorough blending of economic and social concepts with computational and inferential concepts at the level of algorithm design.

arXiv:2508.09219v2 Announce Type: replace-cross Abstract: Recent advances in AI applications have raised growing concerns about the need for ethical guidelines and regulations to mitigate the risks posed by these technologies. In this paper, we present a mixed-methods survey study - combining statistical and qualitative analyses - to examine the ethical perceptions, practices, and knowledge of individuals involved in various AI development roles. Our survey comprises 414 participants from 43 countries, representing various roles such as AI managers, analysts, developers, quality assurance professionals, and information security and privacy experts. The results reveal varying degrees of familiarity and experience with AI ethics principles, government initiatives, and risk mitigation strategies across roles, regions, and other demographic factors. Our findings underscore the importance of a collaborative, role-sensitive approach that involves diverse stakeholders in ethical decision-making throughout the AI development lifecycle. We advocate for developing tailored, inclusive solutions to address ethical challenges in AI development, and we propose future research directions and educational strategies to promote ethics-aware AI practices.

arXiv:2508.15811v2 Announce Type: replace-cross Abstract: Generative query suggestion using large language models offers a powerful way to enhance conversational systems, but aligning outputs with nuanced user preferences remains a critical challenge. To address this, we introduce a multi-stage framework designed for progressive alignment between the generation policy and user intent. Our pipeline begins with prompt engineering as a cold-start strategy, followed by the Supervised Fine-Tuning stage, in which we introduce a distillation method on click logs to create a robust foundational model. To better model user preferences while capturing their inherent uncertainty, we develop a Gaussian Reward Model (GaRM) that represents user preferences as probability distributions rather than point estimates. Finally, we employ reinforcement learning to align the generation policy with these preferences, guided by a composite reward function that integrates GaRM with auxiliary heuristics to mitigate reward hacking. To maintain training stability, this process is enhanced by a novel out-of-distribution regularization method and a two-stage reward fusion technique. Extensive experiments demonstrate that our framework significantly outperforms baselines on both automatic and human evaluations and yields a 34\% relative increase in user engagement as measured by click-through rate in live A/B tests.

arXiv:2510.02967v3 Announce Type: replace-cross Abstract: This paper presents the development and evaluation of a Retrieval-Augmented Generation (RAG) system for querying the United Kingdom's National Institute for Health and Care Excellence (NICE) clinical guidelines using Large Language Models (LLMs). The extensive length and volume of these guidelines can impede their utilisation within a time-constrained healthcare system, a challenge this project addresses through the creation of a system capable of providing users with precisely matched information in response to natural language queries. The system's retrieval architecture, composed of a hybrid embedding mechanism, was evaluated against a corpus of 10,195 text chunks derived from three hundred guidelines. It demonstrates high performance, with a Mean Reciprocal Rank (MRR) of 0.814, a Recall of 81% at the first chunk and of 99.1% within the top ten retrieved chunks, when evaluated on 7901 queries. The most significant impact of the RAG system was observed during the generation phase. When evaluated on a manually curated dataset of seventy question-answer pairs, RAG-enhanced models showed substantial gains in performance. Faithfulness, the measure of whether an answer is supported by the source text, was increased by 64.7 percentage points to 99.5% for the RAG-enhanced O4-Mini model and significantly outperformed the medical-focused Meditron3-8B LLM, which scored 43%. Clinical evaluation by seven Subject Matter Experts (SMEs) further validated these findings, with GPT-4.1 achieving 98.7% accuracy while reducing unsafe responses by 67% compared to O4-Mini (from 3.0 to 1.0 per evaluator). This study thus establishes RAG as an effective, reliable, and scalable approach for applying generative AI in healthcare, enabling cost-effective access to medical guidelines.

arXiv:2512.00696v3 Announce Type: replace-cross Abstract: Artificial intelligence (AI) is reshaping computational and network biology by enabling new approaches to decode cellular communication networks. We introduce Hierarchical Molecular Language Models (HMLMs), a novel framework that models cellular signaling as a specialized molecular language, where signaling molecules function as tokens, protein interactions define syntax, and functional consequences constitute semantics. HMLMs employ a transformer-based architecture adapted to accommodate graph-structured signaling networks through information transducers, mathematical entities that capture how molecules receive, process, and transmit signals. The architecture integrates multi-modal data sources across molecular, pathway, and cellular scales through hierarchical attention mechanisms and scale-bridging operators that enable information flow across biological hierarchies. Applied to a complex network of cardiac fibroblast signaling, HMLMs outperformed traditional approaches in temporal dynamics prediction, particularly under sparse sampling conditions. Attention-based analysis revealed biologically meaningful crosstalk patterns, including previously uncharacterized interactions between signaling pathways. By bridging molecular mechanisms with cellular phenotypes through AI-driven molecular language representation, HMLMs establish a foundation for biology-oriented large language models (LLMs) that could be pre-trained on comprehensive pathway datasets and applied across diverse signaling systems and tissues, advancing precision medicine and therapeutic discovery.

Enterprise leaders are pressing ahead with artificial intelligence, even as some early results remain uneven. Reporting from the Wall Street Journal and Reuters shows that most CEOs expect AI spending to keep rising through 2026, despite difficulty tying those investments to clear, enterprise-wide returns.

The tension highlights where many organisations now sit in their AI journey. The technology has moved beyond trials and proofs of concept, but it has yet to settle into a reliable source of value. Companies are operating in an in-between phase, where ambition, execution, and expectations are all under strain at the same time.

Spending continues, even as returns lag

AI budgets have climbed steadily in large enterprises over the past two years. Competitive pressure, board oversight, and fear of being left behind have all played a role. At the same time, executives are more open about the limits they are seeing. Gains often show up in pockets rather than in the business, pilots fail to spread, and the cost of connecting AI systems to existing tools keeps rising.

A Wall Street Journal survey of senior executives found that most CEOs see AI as central to long-term competitiveness, even if short-term benefits are hard to measure. For many, AI no longer feels optional. It is treated as a capability that must be developed over time, rather than a project that can be paused if results disappoint.

That view helps explain why spending remains steady. Leaders worry that cutting back now could weaken their position later, especially as rivals improve how they use the technology.

Why pilots struggle to scale

One of the main barriers to stronger returns is the jump from experimentation to day-to-day use. Many organisations have launched AI pilots in different teams, often without shared rules or coordination. While these efforts can generate insight and interest, few translate into changes that affect the wider business.

Reuters has reported that companies trying to scale AI frequently run into issues with data quality, system links, security controls, and regulatory requirements. The problems are not only technical, but reflect how work is organised. Responsibility is often split in teams, ownership is unclear, and decisions slow down once projects touch legal, risk, and IT functions.

The result is a pattern of heavy spending on trials, with limited progress toward systems that are embedded in core operations.

Infrastructure costs reshape the equation

The cost of infrastructure is also weighing on AI returns. Training and running models demands large amounts of computing power, storage, and energy. Cloud bills can rise quickly as use grows, while building on-site systems requires upfront investment and long planning cycles. Executives cited by Reuters have warned that infrastructure costs can outpace the benefits delivered by AI tools, particularly in the early stages. This has led to tough choices: whether to centralise AI resources or leave teams to experiment on their own; whether to build in-house systems or rely on vendors; and how much waste is acceptable while capabilities are still forming.

In practice, these decisions are shaping AI strategy as much as model performance or use-case selection.

AI governance moves to the centre of CEO decision-making

As AI spending increases, so does scrutiny. Boards, regulators, and internal audit teams are asking harder questions. In response, many organisations are tightening control. Decision rights are shifting toward central teams, AI councils are becoming more common, and projects are being linked more closely to business priorities.

The Wall Street Journal reports that companies are moving away from loosely connected experiments toward clearer goals, measures, and timelines. This can slow progress, but it reflects a growing belief that AI should be managed with the same discipline as other major investments.

The shift marks a change in how AI is treated. It is no longer a side effort or a curiosity but is being brought into existing operating and risk structures.

Expectations are being reset, not abandoned

Importantly, the persistence of AI spending does not signal blind optimism. Instead, it reflects a reset in expectations. CEOs are learning that AI rarely delivers immediate, sweeping returns. Value tends to emerge gradually, as organisations adjust workflows, retrain staff, and refine data foundations.

Rather than abandoning AI initiatives, many enterprises are narrowing their focus. They are prioritising fewer use cases, demanding clearer ownership, and aligning projects more closely with business outcomes. The re-calibration may reduce short-term excitement, but it improves the likelihood of sustainable returns.

What CEO AI strategy signals for 2026 planning

For organisations shaping their plans for 2026, the message for every CEO is not to retreat from AI, but to pursue it with more care as AI strategies mature. Ownership, governance, and realistic timelines matter more than headline spending levels or bold claims.

Those most likely to benefit are treating AI as a long-term shift in how the organisation works, not a quick route to growth. In the next phase, advantage will depend less on how much is spent and more on how well AI fits into everyday operations.

(Photo by Ambre Estève)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post CEOs still betting big on AI: Strategy vs. return on investment in 2026 appeared first on AI News.

npj Digital Medicine, Published online: 15 December 2025; doi:10.1038/s41746-025-02218-5

H&E-based MSI/MMR testing with AI in colorectal cancer: a multi-centred blinded evaluation

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.

In July, a widely cited MIT study claimed that 95% of organizations that invested in generative AI were getting “zero return.” Tech stocks briefly plunged. While the study itself was more nuanced than the headlines, for many it still felt like the first hard data point confirming what skeptics had muttered for months: Hype around AI might be outpacing reality.

Then, in August, OpenAI CEO Sam Altman said what everyone in Silicon Valley had been whispering. “Are we in a phase where investors as a whole are overexcited about AI?” he said during a press dinner I attended. “My opinion is yes.” 

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

He compared the current moment to the dot-com bubble. “When bubbles happen, smart people get overexcited about a kernel of truth,” he explained. “Tech was really important. The internet was a really big deal. People got overexcited.” 

With those comments, it was off to the races. The next day’s stock market dip was attributed to the sentiment he shared. The question “Are we in an AI bubble?” became inescapable.

Who thinks it is a bubble? 

The short answer: Lots of people. But not everyone agrees on who or what is overinflated. Tech leaders are using this moment of fear to take shots at their rivals and position themselves as clear winners on the other side. How they describe the bubble depends on where their company sits.

When I asked Meta CEO Mark Zuckerberg about the AI bubble in September, he ran through the historical analogies of past bubbles—railroads, fiber for the internet, the dot-com boom—and noted that in each case, “the infrastructure gets built out, people take on too much debt, and then you hit some blip … and then a lot of the companies end up going out of business.”

But Zuckerberg’s prescription wasn’t for Meta to pump the brakes. It was to keep spending: “If we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously. But I’d say the risk is higher on the other side.”

Bret Taylor, the chairman of OpenAI and CEO of the AI startup Sierra, uses a mental model from the late ’90s to help navigate this AI bubble. “I think the closest analogue to this AI wave is the dot-com boom or bubble, depending on your level of pessimism,” he recently told me. Back then, he explained, everyone knew e-commerce was going to be big, but there was a massive difference between Buy.com and Amazon. Taylor and others have been trying to position themselves as today’s Amazon.

Still others are arguing that the pain will be widespread. Google CEO Sundar Pichai told the BBC this month that there’s “some irrationality” in the current boom. Asked whether Google would be immune to a bubble bursting, he warned, “I think no company is going to be immune, including us.”

What’s inflating the bubble?

Companies are raising enormous sums of money and seeing unprecedented valuations. Much of that money, in turn, is going toward the buildout of massive data centers—on which both private companies like OpenAI and Elon Musk’s xAI and public ones such as Meta and Google are spending heavily. OpenAI has pledged that it will spend $500 billion to build AI data centers, more than 15 times what was spent on the Manhattan Project.

This eye-popping spending on AI data centers isn’t entirely detached from reality. The leaders of the top AI companies all stress that they’re bottlenecked by their limited access to computing power. You hear it constantly when you talk to them. Startups can’t get the GPU allocations they need. Hyperscalers are rationing compute, saving it for their best customers.

If today’s AI market is as brutally supply-constrained as tech leaders claim, perhaps aggressive infrastructure buildouts are warranted. But some of the numbers are too large to comprehend. Sam Altman has told employees that OpenAI’s moonshot goal is to build 250 gigawatts of computing capacity by 2033, roughly equaling India’s total national electricity demand. Such a plan would cost more than $12 trillion by today’s standards.

“I do think there’s real execution risk,” OpenAI president and cofounder Greg Brockman recently told me about the company’s aggressive infrastructure goals. “Everything we say about the future, we see that it’s a possibility. It is not a certainty, but I don’t think the uncertainty comes from scientific questions. It’s a lot of hard work.”

Who is exposed, and who is to blame?

It depends on who you ask. During the August press dinner, where he made his market-moving comments, Altman was blunt about where he sees the excess. He said it’s “insane” that some AI startups with “three people and an idea” are receiving funding at such high valuations. “That’s not rational behavior,” he said. “Someone’s gonna get burned there, I think.” As Safe Superintelligence cofounder (and former OpenAI chief scientist and cofounder) Ilya Sutskever put it on a recent podcast: Silicon Valley has “more companies than ideas.”

Demis Hassabis, the CEO of Google DeepMind, offered a similar diagnosis when I spoke with him in November. “It feels like there’s obviously a bubble in the private market,” he said. “You look at seed rounds with just nothing being tens of billions of dollars. That seems a little unsustainable.”

Anthropic CEO Dario Amodei also struck at his competition during the New York Times DealBook Summit in early December. He said he feels confident about the technology itself but worries about how others are behaving on the business side: “On the economic side, I have my concerns where, 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.”

He stopped short of naming Sam Altman and OpenAI, but the implication was clear. “There are some players who are YOLOing,” he said. “Let’s say you’re a person who just kind of constitutionally wants to YOLO things or just likes big numbers. Then you may turn the dial too far.”

Amodei also flagged “circular deals,” or the increasingly common arrangements where chip suppliers like Nvidia invest in AI companies that then turn around and spend those funds on their chips. Anthropic has done some of these, he said, though “not at the same scale as some other players.” (OpenAI is at the center of a number of such deals, as are Nvidia, CoreWeave, and a roster of other players.) 

The danger, he explained, comes when the numbers get too big: “If you start stacking these where they get to huge amounts of money, and you’re saying, ’By 2027 or 2028 I need to make $200 billion a year,’ then yeah, you can overextend yourself.”

Zuckerberg shared a similar message at an internal employee Q&A session after Meta’s last earnings call. He noted that unprofitable startups like OpenAI and Anthropic risk bankruptcy if they misjudge the timing of their investments, but Meta has the advantage of strong cash flow, he reassured staff.

How could a bubble burst?

My conversations with tech executives and investors suggest that the bubble will be most likely to pop if overfunded startups can’t turn a profit or grow into their lofty valuations. This bubble could last longer than than past ones, given that private markets aren’t traded on public markets and therefore move more slowly, but the ripple effects will still be profound when the end comes. 

If companies making grand commitments to data center buildouts no longer have the revenue growth to support them, the headline deals that have propped up the stock market come into question. Anthropic’s Amodei illustrated the problem during his DealBook Summit appearance, where he said the multi-year data center commitments he has to make combine with the company’s rapid, unpredictable revenue growth rate to create a “cone of uncertainty” about how much to spend.

The two most prominent private players in AI, OpenAI and Anthropic, have yet to turn a profit. A recent Deutsche Bank chart put the situation in stark historical context. Amazon burned through $3 billion before becoming profitable. Tesla, around $4 billion. Uber, $30 billion. OpenAI is projected to burn through $140 billion by 2029, while Anthropic is expected to burn $20 billion by 2027.

Consultants at Bain estimate that the wave of AI infrastructure spending will require $2 trillion in annual AI revenue by 2030 just to justify the investment. That’s more than the combined 2024 revenue of Amazon, Apple, Alphabet, Microsoft, Meta, and Nvidia. When I talk to leaders of these large tech companies, they all agree that their sprawling businesses can absorb an expensive miscalculation about the returns from their AI infrastructure buildouts. It’s all the other companies that are either highly leveraged with debt or just unprofitable—even OpenAI and Anthropic—that they worry about. 

Still, given the level of spending on AI, it still needs a viable business model beyond subscriptions, which won’t be able to  drive profits from billions of people’s eyeballs like the ad-driven businesses that have defined the last 20 years of the internet. Even the largest tech companies know they need to ship the world-changing agents they keep hyping: AI that can fully replace coworkers and complete tasks in the real world.

For now, investors are mostly buying into the hype of the powerful AI systems that these data center buildouts will supposedly unlock in the future. At some point the biggest spenders, like OpenAI, will need to show investors that the money spent on the infrastructure buildout was worth it.

There’s also still a lot of uncertainty about the technical direction that AI is heading in. LLMs are expected to remain critical to more advanced AI systems, but industry leaders can’t seem to agree on which additional breakthroughs are needed to achieve artificial general intelligence, or AGI. Some are betting on new kinds of AI that can understand the physical world, while others are focused on training AI to learn in a general way, like a human. In other words, what if all this unprecedented spending turns out to have been backing the wrong horse?

The question now

What makes this moment surreal is the honesty. The same people pouring billions into AI will openly tell you it might all come crashing down. 

Taylor framed it as two truths existing at once. “I think it is both true that AI will transform the economy,” he told me, “and I think we’re also in a bubble, and a lot of people will lose a lot of money. I think both are absolutely true at the same time.”

He compared it to the internet. Webvan failed, but Instacart succeeded years later with essentially the same idea. If you were an Amazon shareholder from its IPO to now, you’re looking pretty good. If you were a Webvan shareholder, you probably feel differently. 

“When the dust settles and you see who the winners are, society benefits from those inventions,” Amazon founder Jeff Bezos said in October. “This is real. The benefit to society from AI is going to be gigantic.”

Goldman Sachs says the AI boom now looks the way tech stocks did in 1997, several years before the dot-com bubble actually burst. The bank flagged five warning signs seen in the late 1990s that investors should watch now: peak investment spending, falling corporate profits, rising corporate debt, Fed rate cuts, and widening credit spreads. We’re probably not at 1999 levels yet. But the imbalances are building fast. Michael Burry, who famously called the 2008 housing bubble collapse (as seen in the film The Big Short), recently compared the AI boom to the 1990s dot-com bubble too.

Maybe AI will save us from our own irrational exuberance. But for now, we’re living in an in-between moment when everyone knows what’s coming but keeps blowing more air into the balloon anyway. As Altman put it that night at dinner: “Someone is going to lose a phenomenal amount of money. We don’t know who.”

Alex Heath is the author of Sources, a newsletter about the AI race, and the cohost of ACCESS, a podcast about the tech industry’s inside conversations. Previously, he was deputy editor at The Verge.