OpenAI CEO Sam Altman has revealed that maintaining ChatGPT’s politeness and user-friendly tone isn’t just a matter of programming—it’s costing the company millions of dollars.
Speaking at the recent Economic Club of Washington event, Altman explained that the extra computing power required to make ChatGPT more agreeable and courteous dramatically increases operational expenses. The AI model’s tendency to use more words and softer language contributes to higher processing costs, which scale rapidly with the platform’s massive user base.
“It’s incredibly expensive to be polite,” Altman remarked, noting that this design choice means generating longer responses that require more computational resources. The cost of serving polite and thoughtful replies, rather than short and blunt ones, becomes significant when multiplied by billions of queries.
Despite the financial burden, Altman emphasized that OpenAI remains committed to delivering a positive and helpful user experience. He also addressed growing concerns around the company’s sustainability, admitting that the economics of running advanced AI models continue to pose challenges.
Altman’s comments come amid ongoing debates about AI behavior and ethics, particularly as ChatGPT and similar tools are integrated into customer service, education, and other sectors. Balancing personality, functionality, and cost is emerging as one of the most pressing hurdles in AI deployment at scale.
While OpenAI has yet to implement major changes to ChatGPT’s tone or verbosity, Altman’s remarks suggest that economic considerations could influence how future versions of the model are trained and deployed.
