OpenAI's CEO Announces 20% of Computing Resources for AI Safety

TapTechNews August 2nd news, OpenAI's CEO Sam Altman tweeted on the X platform yesterday (August 1st), announcing to provide at least 20% of computing resources for AI safety-related research and development work to ensure safety.

TapTechNews translated the tweet content as follows:

A brief update on AI safety from OpenAI

As we said in late July, we ensure to allocate at least 20% of computing resources for work in the field of AI safety.

We have reached an agreement with the American Institute for AI Safety to give it early access to our next-generation foundation model so that we can jointly promote the scientific development of AI assessment. We are very excited about this!

For current and former employees who can raise concerns about AI safety, I hope to provide reassurance.

I hereby promise that safety is crucial for any AI company, especially for us, and is also an important part of our safety plan.

We canceled the non-derogatory clause for current and former employees and the clause granting OpenAI the right to cancel vested equity rights (although never used) in May. We have been constantly working to correct this problem.