OpenAI ignoring research, Sora and DALL-E, suggest people leaving ChatGPT maker

According to a recent report by the Financial Times, a profound cultural and strategic shift is unfolding within the halls of OpenAI. Once a sanctuary for “blue-sky” experimentation, the $500 billion company is reportedly pivoting into a product-driven engineering factory. This transformation, driven by an intense “Code Red” internal mandate, has prioritized the commercial dominance of ChatGPT at the expense of its pioneering research into video and image generation, specifically targeting projects like Sora and DALL-E.

Survey

Also read: Indians get 13 scam messages daily, AI and deepfake boosting scams

Add As A Trusted Source For Google.

Add as a preferred source on Google

The “second-class” status of creative AI

The transition from a research laboratory to a Silicon Valley giant has created a hierarchy where ChatGPT is the undisputed priority. Multiple sources close to the company indicate that researchers working on non-language models now find themselves in a constant battle for resources. In the high-stakes world of AI development, “computing credits” are the primary currency; however, those working on the text-to-video model Sora and the image-generator DALL-E have reportedly seen their requests denied or slashed.

Also read: Sam Altman feels useless thanks to ChatGPT Codex, and he’s sad about it

This reallocation of compute power and funding toward Large Language Models (LLMs) has left many top-tier researchers feeling like “second-class citizens.” The Financial Times highlights that projects deemed less relevant to the immediate improvement of ChatGPT’s chatbot interface are being wound down or starved of the necessary infrastructure to validate new theories. This “ChatGPT-first” strategy is a direct response to the “cut-throat race” against rivals like Google’s Gemini 3 and Anthropic’s Claude, forcing OpenAI to choose between long-term innovation and short-term market defense.

A talent emigration

The most visible symptom of this shift is a growing talent drain. Senior figures who joined OpenAI for its original mission of foundational research are now looking elsewhere. The departure of Jerry Tworek, a seven-year veteran and VP of Research, serves as a high-profile example. Tworek reportedly left following a standoff over the direction of “reasoning” research, signaling that the company’s current architecture is now favored over exploring entirely new scientific approaches.

Other notable exits include model policy researcher Andrea Vallone, who moved to rival Anthropic, and economist Tom Cunningham. These departures suggest a mounting disillusionment among staff who believe the company is straying from impartial research to focus on work that primarily serves its commercial narrative. While OpenAI’s leadership maintains that long-term research remains central, the internal reorganization suggests otherwise.

For investors, the shift is a logical evolution toward justifying a half-trillion-dollar valuation. By focusing on “platform lock-in” for its 800 million users, OpenAI is building a formidable business moat. However, for the scientists who built the foundation of generative AI, the sacrifice of Sora and DALL-E’s advancement may represent a loss of the very soul that made OpenAI a pioneer in the first place.

Also read: MoltMatch is a dating platform for AI agents, no we are not kidding

Follow Us

Vyom Ramani

A journalist with a soft spot for tech, games, and things that go beep. While waiting for a delayed metro or rebooting his brain, you’ll find him solving Rubik’s Cubes, bingeing F1, or hunting for the next great snack. View Full Profile