Breakdown of OpenAI's new safety statement

I know what the “Open” in “OpenAI” stands for! Open questions about AI safety. We should call out OpenAI's new safety statement!

➤ OpenAI’s GPT-4o release included a brief safety statement

🤐As usual, the statement is vague at best and intentionally ignores huge risks at worst.

🤐 Additional documents may be in the works. But, if they are to come, why isn’t safety information important enough to be published with the release?

➤ My overall takeaway

😒 The main message -- "Just trust us. We evaluated ourselves against our own standards and we’re good."

❌ This is simply not good enough. Not for any company, and especially not a company boasting that they can build artificial general intelligence and care about humanity.

📣 Let's demand better! Call them out!

➤ Highlighting seven open questions:

➤ Join the conversation in the LinkedIn post!