The article further highlights a pattern of evasion and half-truths from OpenAI over the years, providing examples of statements made by the company and their actual implications. These include issues about openness, data training, finances, views on regulation, and more. The author suggests that this lack of full candity is embedded in the company's DNA and raises concerns about the company's plans to upload private data for potential resale to advertisers, as well as the potential risks of rogue AI.
Key takeaways:
- The internet is buzzing about OpenAI’s CTO Mira Murati's interview with Wall Street Journal reporter Joanna Stern, particularly her evasive answers about what data Sora was trained on.
- Many in the tech and journalism community have criticized Murati's lack of transparency, with some suggesting that her lack of knowledge about the specific data used is unlikely given her position.
- The article suggests that OpenAI has a history of being less than fully transparent, with examples of misleading or evasive statements dating back to its founding in 2015.
- As OpenAI prepares to potentially access and use private user data, the author suggests that the company's track record of half-truths and lack of full candor is a cause for concern.