OpenAI is in a bit of a tough spot at the moment. Following the company’s decision to sign a deal with the Pentagon for AI, users have started boycotting ChatGPT. While OpenAI CEO Sam Altman has clarified that the company will not allow the US Military to cross red lines with its systems, it seems that not everyone is happy with the deal, even inside OpenAI. OpenAI research scientist Aiden McLaughlin has slammed the move saying it wasn’t worth it.
On X, McLaughlin shared a post, giving a rather clear cut view on OpenAI’s deal with the US Department of Defense. He wrote, “I personally don’t think this deal was worth it.” In a follow-up post, he said internally, the volume and thoughtfulness of discussion around the deal have been “overwhelming” and that he was proud to work for a company where people could speak their mind.
Aiden’s comments come at a time when OpenAI is receiving criticism from users online who — ever since the company made the defense deal with the Pentagon — are ditching ChatGPT for competing chatbots particularly Anthropic’s Claude. ChatGPT uninstallation reached a whopping 295 per cent over the weekend, per data released by Sensor Tower even as Claude became number one app on Apple’s App Store in the US.
Why is OpenAI’s deal not worth it?
While Aiden McLaughlin does not explain the reason behind his opinion on this deal, the situation for OpenAI seems grim. Sam Altman recently accepted that even though the company hoped to de-escalate the situation between the Pentagon and the AI industry, rushing the agreement made it look “opportunistic and sloppy.”
OpenAI announced its deal just a few hours after the US Military cancelled its contact with Anthropic and quickly moved to categorise it as a supply chain risk, similar to what it has done with China’s Huawei. Though Altman does believe that the Dario Amodei-led startup should not be labelled a supply chain risk by the US government – something that would force all defense contractors to cut their ties with Anthropic.
Altman has clarified that the contract will not allow the US Military to use OpenAI’s AI systems for mass domestic surveillance or the development of autonomous weapons – the same concerns that led to Anthropic’s termination.
However, a report from The Verge suggests that the agreement between OpenAI and the Department of Defense is not as strict as what Anthropic had demanded.
OpenAI scientist says the company allows for free discussions
McLaughlin was asked by a user on X if the current situation made him feel conflicted. The OpenAI researcher stated that he does feel strongly over various issues within the company. He said, “The volume and thoughtfulness of discussion internally is overwhelming, and I feel incredibly proud to work somewhere where people can speak their mind.”
OpenAI safety researcher Cameron Raymond quoted McLaughlin’s original post and wrote, “Idk how the dust will settle but for now I feel similarly.”
While Aiden McLaughlin does not agree with OpenAI’s recent deal, some of his colleagues have a different view. OpenAI CEO of Applications Fidji Simo said in a separate thread on X, “Despite the cost, it remains the right thing to have done.”
Additionally, OpenAI Codex engineer Tibo wrote that the company was “setting the right guardrails to ensure that deployment is aligned with what is considered acceptable and safe.”
