Benevolent Steward or Corporate Operator? A Scientist Questions the Altman Narrative

For years, Sam Altman of OpenAI has projected an image of a selfless guide for humanity's journey with artificial intelligence. He discusses safety and societal benefit with the ease of a practiced speaker. Yet according to prominent AI critic Gary Marcus, that image is fracturing. In a recent analysis, Marcus contends that a record of broken commitments, internal conflicts, and aggressive corporate tactics reveals a different story.

The core of the critique centers on OpenAI's evolution. Founded as a non-profit to develop AI for public good, the company's 2019 restructuring to a 'capped-profit' model opened doors to billions in investment from Microsoft. Marcus identifies this as the moment the mission began to blur, creating inherent tension between safety mandates and investor expectations.

This tension erupted publicly in late 2023 when OpenAI's board fired Altman, citing a lack of candor. Though swiftly reinstated, the specific concerns were never fully explained. Subsequently, board members focused on safety, like Helen Toner, were replaced by figures such as economist Larry Summers. Marcus views this as the neutralization of internal oversight.

The departure of key safety researchers, including co-founder Ilya Sutskever and Superalignment lead Jan Leike, provided further evidence of internal discord. Leike stated safety had become secondary to 'shiny products.' Reports that his team never received promised computing resources undermined Altman's public safety commitments.

Other incidents have drawn scrutiny. OpenAI required exiting employees to sign perpetual non-disparagement agreements under threat of losing vested equity - a practice Altman claimed was unknown to him. A voice feature for ChatGPT, named 'Sky,' bore a striking resemblance to Scarlett Johansson's voice after she declined Altman's offer to license it. These episodes, Marcus argues, reveal a pattern of overreach followed by retreat only under pressure.

Marcus also examines Altman's lobbying for AI regulation, suggesting it could create barriers for smaller competitors, thus protecting OpenAI's market position. While Altman has long presented himself as having no direct OpenAI equity, discussions about restructuring for him to gain a stake, alongside his holdings in affiliated funds, challenge the narrative of pure altruism.

The collective weight of these events, Marcus concludes, erodes the essential trust required from a company steering a technology of such profound consequence. The gap between humanitarian rhetoric and corporate hardball grows ever wider, raising fundamental questions about who is truly being served.

Source: Webpronews