Google Could 'Eat OpenAI's Lunch,' Warns Cal Newport

Alphabet Inc. (NASDAQ:GOOGL) could “eat OpenAI’s lunch” by offering cheaper artificial intelligence through its existing products, potentially undermining the ChatGPT maker’s business, according to Georgetown University computer science professor Cal Newport. Speaking on the Prof G Markets podcast Friday, Newport warned that Google’s custom chips, cloud infrastructure and enormous reach could threaten OpenAI as AI becomes increasingly integrated into everyday software. Why Newport Thinks OpenAI’s Big Bet Could Fail Newport believes OpenAI and Anthropic have bet that building increasingly powerful language models will eventually produce artificial general intelligence, capable of handling almost any intellectual task. “That was their bet and I think that bet is not going to pay off,” Newport said. He argues that recent improvements have concentrated in programming, mathematics and cybersecurity, where models can learn from large amounts of information with verifiable answers. Instead, he expects cheaper AI systems designed for particular tasks to become increasingly important, weakening the case for spending enormous sums on general-purpose models. Asked what should keep OpenAI CEO Sam Altman up at night, Newport pointed to Google. Google’s Structural Advantage Google already controls Search, Gmail and Docs allowing it to bring AI directly to existing customers. On Thursday, Google unveiled a universal Gemini agent that works across Workspace and can select between Google’s own models and Anthropic’s Claude. That means Google can offer customers AI services without always relying on its own model, and route simple tasks to cheaper models. “The best model for the task is not always the largest one,” Google Cloud CEO Thomas Kurian said. Newport argues that if users can get affordable AI assistance directly through Search and workplace applications, OpenAI risks losing demand currently flowing through ChatGPT. “That would be a financial nightmare scenario for OpenAI,” he said. He even speculated that Microsoft Corp. (NASDAQ:MSFT) could eventually acquire OpenAI if Google’s strategy succeeds. What Traders Expect On Polymarket, traders give Google approximately 70% odds of having the highest-ranked AI model at the end of October, with Anthropic at 26% and OpenAI at about 1%. Google’s Gemini 4 Argon, announced Sept. 30 but not yet widely available to the public, currently tops Arena’s text leaderboard, which the market resolves on, although its score remains preliminary. Other benchmarks paint a different picture. On Artificial Analysis, Argon scores 53 on the Intelligence Index, behind Anthropic’s Claude Opus 5.5 at 58. Whichever model tops the leaderboard, Newport’s warning is that Google could make ChatGPT less necessary. Alphabet shares were up nearly 1% in Friday’s premarket trading. Image: Shutterstock Read Also: TSMC Plans Up to $64 Billion in Capex and Still Isn’t ‘Building Enough,’ Says Applied Materials Veteran