Cerebras Systems CEO and co-founder Andrew Feldman is set to speak at TechCrunch Disrupt 2026 about a question with growing business relevance: whether AI can continue scaling as demand for computing power, energy and infrastructure rises.

While this is an event announcement rather than a product launch or regulatory change, the subject is important for companies using or planning to adopt generative AI. The economics and availability of AI services are closely tied to the hardware and infrastructure behind them.

What happened

TechCrunch announced that Andrew Feldman will appear at TechCrunch Disrupt 2026 to discuss the growing demand for compute, energy and infrastructure required for AI systems, how Cerebras approaches those constraints, and what could happen if current AI hardware reaches practical limits.

Based on the source, no new product, policy or technical milestone has been announced. The confirmed development is Feldman’s planned discussion of AI scaling challenges at the event.

Why it matters for European businesses

For European companies, especially SMEs adopting AI through cloud platforms, SaaS tools or custom applications, AI infrastructure constraints can have direct business effects.

• Cost pressure: If compute and energy demands continue to rise, AI services may remain expensive or become harder to scale economically for routine business use.
• Service availability: Businesses relying on external AI providers may face capacity bottlenecks, slower performance or regional availability differences.
• Deployment choices: Infrastructure limits can influence whether companies use public AI APIs, smaller specialised models, on-premise systems or hybrid approaches.
• Vendor strategy: Changes in AI hardware competition may affect the pricing, speed and capabilities offered by AI vendors serving European customers.

The broader issue is that AI adoption is not only a software decision. It depends on data-centre capacity, semiconductor supply, power consumption and the ability of providers to deliver reliable performance at commercial scale.

Who may be affected

• SMEs using generative AI tools for customer support, content, search, analytics or internal productivity.
• E-commerce businesses deploying AI for merchandising, recommendations, automation or service operations.
• IT and digital teams evaluating infrastructure costs, vendor lock-in and integration roadmaps.
• Founders and operations leaders planning AI investment and looking for realistic return on investment.
• Agencies and software providers building AI-enabled services on top of third-party platforms.

What companies should consider

• Avoid assuming unlimited AI capacity: Build business cases around measurable value, not the expectation that more powerful models will always be available at lower cost.
• Review vendor dependence: Understand which AI services your tools rely on and how pricing or performance changes could affect operations.
• Prioritise efficient use cases: Many business tasks may not require the largest or most expensive models. Smaller or more targeted systems can be easier to justify commercially.
• Track infrastructure developments: Hardware and compute market changes can influence procurement decisions, especially for companies planning custom AI deployments.
• Plan for flexibility: Where possible, avoid architectures that make it difficult to switch models or providers if availability, cost or compliance requirements change.

At this stage, the announcement is mainly relevant as a signal of where the AI market conversation is heading: away from pure model performance and toward the practical limits of scaling. For European businesses, that makes AI infrastructure strategy increasingly important alongside experimentation and adoption.