OpenAI CEO Sam Altman has issued a “code red” to refocus the company on ChatGPT’s core user experience. The move will reportedly delay several planned features — including a personal assistant named Pulse, new shopping agents and advertising tools — to prioritize making the chatbot faster, more reliable and more personalized.
The pivot follows the release of Google’s Gemini 3, which has earned top marks on industry benchmarks and strong reviews. Experts told CNBC that Google’s resurgence is fueled by new “Ironwood” chips and its vast YouTube data library.
Download the SAN app today to stay up-to-date with Unbiased. Straight Facts™.
Point phone camera here
The Wall Street Journal reports that Gemini’s monthly user base jumped to 650 million in October, driven by its “Nano Banana” image tool, though OpenAI still leads with 800 million weekly users.
What OpenAI is pausing — and promising next
To support the new focus, Altman reportedly encouraged staff to temporarily transfer teams and instituted a daily coordination call for ChatGPT development. The memo confirmed that projects like the Pulse assistant and AI shopping agents would be paused.
Following the internal announcement, ChatGPT lead Nick Turley posted on X that the goal is to make the product “even more intuitive and personal.”
OpenAI has wrestled in recent months with making ChatGPT safer while staying engaging, the Journal reported, noting user complaints about GPT-5’s tone and basic tasks before a later update aimed to make it “warmer” and more responsive.
Rivals, users and money: the stakes
The stakes are high for OpenAI, which remains unprofitable. According to The Journal, internal projections suggest the company needs to reach $200 billion in revenue by 2030 to turn a profit, given its commitment to data-center investments.
In contrast, CNBC reports that Google is already rolling out Gemini 3 across its enterprise and consumer lines, leveraging its new Ironwood chips to challenge Nvidia’s dominance in AI infrastructure.
Altman told staff a new reasoning model planned for next week is “ahead of” Google’s latest Gemini, according to the Journal. Google, for its part, is rolling out Gemini 3 across products, including search features, as users and developers keep testing models on coding, multimodal tasks and complex reasoning.
