Observed Signal · May 7, 2026 · Conference Panel Discussion · Source: techcrunch · Impact: 3/5 · Sentiment: Neutral
AI leaders warn of chip, energy and architecture bottlenecks
At a Milken Global Conference panel hosted by TechCrunch, five executives across the AI supply chain — Christophe Fouquet (ASML), Francis deSouza (Google Cloud), Qasar Younis (Applied Intuition), Dimitry Shevelenko (Perplexity) and Eve Bodnia (Logical Intelligence) — discussed structural constraints facing the AI industry. They highlighted near-term chip supply limits, growing energy and cooling challenges (including exploration of orbital data centers), and debate over model architectures. Google Cloud cited rapid revenue and backlog growth, ASML warned the market will be supply‑limited for years, and Applied Intuition emphasized real‑world data scarcity for physical autonomy. Bodnia described energy‑based models (EBMs) as a different architecture that uses far fewer parameters and updates online. Perplexity outlined agent products with granular permissioning and approval flows. Panelists also raised geopolitics and sovereignty concerns for physical AI systems and noted the potential societal and workforce impacts of accelerating AI capabilities.
Panel comments from leaders at ASML, Google Cloud and other AI firms highlight concrete, near‑term infrastructure constraints (chips, energy) and alternative architectures that affect compute costs, deployment strategies and geopolitical risk — relevant but not an immediate industry‑shifting policy or product launch.
Track Perplexity Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Christophe Fouquet (CEO of ASML) said the AI chip market will be supply‑limited for the next two to five years.
- Francis deSouza (COO, Google Cloud) said Google Cloud revenue exceeded $20 billion last quarter, growing 63%, and reported a backlog increase from $250 billion to $460 billion.
- Google is exploring data centers in space to address energy constraints and heat‑dissipation challenges.
- Eve Bodnia’s startup Logical Intelligence is building energy‑based models (EBMs) with a largest model of ~200 million parameters that she says runs thousands of times faster than large LLMs and updates without full retraining.
- Perplexity launched Perplexity Computer (a ‘digital worker’) and described agent permission granularity and an agent (Comet) that presents action plans for user approval.
Connected Companies & Entities
7 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
AI Vibe Shift: Caution and Supply Shortages Arrive
The article argues that a major narrative shift is occurring in AI on two levels. First, frontier labs are moving from rapid public releases toward precautionary access restrictions after Anthropic said its newest model, Claud Mythos, is too powerful for broad release and has been limited to select partners (including Microsoft and Apple); OpenAI is also reportedly restricting access to its most advanced models. Second, the market narrative has shifted from fearing an AI capex bubble to confronting severe compute supply constraints: consumer demand is outpacing hyperscalers' ability to provide chips, data centers and electricity. The author suggests AI may be entering a long-term infrastructure growth phase analogous to early‑20th‑century electricity rather than a speculative bubble.
Executives: AI Demand 'Almost Unlimited' Amid Valuemaxxing
AI executives told CNBC that demand for compute tied to generative AI remains extremely strong and, in many cases, exceeds available capacity across chips and data centers. Leaders including Pat Gelsinger, Marc Boroditsky, Andrew Feldman and others said energy, data center space and component supply are primary bottlenecks. At the same time, enterprises are shifting from carefree 'tokenmaxxing' usage toward 'valuemaxxing' — focusing on return on investment and rationalizing AI spending by choosing models and workloads that match value. Market moves such as Meta and xAI selling or renting excess compute capacity have sparked debate about overcapacity, but several executives said those are isolated cases and do not reflect a broad slowdown.
CEO Dinner Highlights AI Risks: Chips, Institutions, Safety
A March 18 CEO Dinner in San Francisco convened sixteen founders, investors and researchers to discuss risks that could slow AI progress and the human skills likely to remain durable. Attendees grouped threats into three categories: geopolitical (chip concentration in Taiwan and energy/vulnerability risks), institutional/human (adoption friction, foot-dragging, scams and civil unrest), and technical (rushed model releases, prompt injection and novel attack vectors like "LLM viruses"). The room expected a bumpy three- to five-year transition before sustained acceleration. Participants argued that the most resilient human advantages are relational and emotional capabilities — judgment in noisy domains, reading a room, cultivating long-term trust, humor and generating self-transcendent leadership — which AI is unlikely to fully replicate in the near term. The report synthesizes conversation highlights and does not attribute views to named attendees under Chatham House Rule.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.