From Platformer News: 2024-06-04 20:23:52

Eleven current and former employees of OpenAI, along with two from Google DeepMind, have raised concerns about their inability to voice concerns about risks due to confidentiality agreements. This trend mirrors what has happened in other tech companies like Meta and Google, where idealistic workers are hired, but executives prioritize commercial incentives over safety. OpenAI, initially a nonprofit dedicated to safely building artificial general intelligence (AGI), has shifted towards commercial interests since the release of ChatGPT in 2022. CEO Sam Altman’s transformation of OpenAI to resemble a traditional startup led to tensions within the company, with employees preferring the commercial prospects it offered. Following Altman’s firing and return, new business-minded board members have stood by him, despite questions about his business dealings and conflicts of interest. The latest whistleblowers from OpenAI claim the company’s focus on profits and building AGI have led to hardball tactics to silence employees expressing concerns about their technology and products. While specific complaints may be scarce due to confidentiality agreements, the whistleblowers’ letter serves as a Rorschach test for differing views on AI safety and the potential risks associated with developing AGI. The lack of transparency and specificity in whistleblowers’ complaints highlights the ongoing challenges of addressing AI safety and the need for a clearer path towards safe development and aligning AI with human intentions.



Read more at Platformer News: What aren’t the OpenAI whistleblowers saying?