#68 (35) Progress, Priorities and Likely Future Paths for AI