AI's Future: Can It Keep Growing?

Get ready to hear about a potential roadblock on the AI superhighway, and why it might matter to your wallet and your digital future.

Andrew Feldman, the CEO of Cerebras Systems, is set to speak at TechCrunch Disrupt 2026. He’ll be talking about a big question: can Artificial Intelligence [AI, computer systems designed to perform tasks that typically require human intelligence] keep getting bigger and more powerful, or are we hitting some limits? Think of it like a popular new restaurant. Everyone wants a table, but the kitchen only has so many stoves, the power grid only has so much electricity, and the building only has so much space.

Feldman plans to discuss the soaring "compute" [the raw processing power needed for AI], energy, and infrastructure [the underlying systems and facilities] demands that come with building bigger and smarter AI. He’ll explain how Cerebras is trying a different approach to these challenges, essentially trying to build a more efficient kitchen or find ways to share resources better. The big question he'll tackle is what happens next if the specialized computer chips and systems we use for AI today, like those powering models from OpenAI or Google, can't keep up with the hunger for more power.

This matters because the "bigger and better" trend in AI is what gives us everything from more convincing chatbots to more advanced image generators. If AI hardware hits a wall, it could slow down the development of new AI features we've come to expect. It could also make these powerful AI tools more expensive to create and run, potentially affecting the cost of services that rely on them.

This discussion from Cerebras Systems fits into a broader pattern we’re seeing in the tech world: the search for more sustainable and efficient ways to power our increasingly digital lives. As AI becomes integrated into more everyday tools, the hidden costs of its development, particularly in energy and infrastructure, are becoming more apparent. Instead of just marveling at what AI can do, we're starting to ask how we can make it work without breaking the bank or the power grid. A concrete next step for you might be to pay attention to your own energy consumption from digital devices.

The future of AI might depend on finding smarter, not just bigger, ways to power its progress.