Researchers at OpenAI decided that a system that scores well at understanding language could too easily be manipulated for malicious intent.
No comments yet. Be the first to comment!
More in Course Chunks from WIRED
No comments yet. Be the first to comment!