In his latest blog post, OpenAI CEO Sam Altman predicts that in 2025, agents capable of true cognitive work will emerge; in 2026, systems capable of generating new insights may emerge; and in 2027, robots capable of performing tasks in the real world may emerge. In some very important ways, the 2030s are likely to be very different from any previous period, with intelligence and energy becoming extremely abundant, and with abundant intelligence and energy (and good management), you can theoretically have anything else. At the same time, we also face serious challenges. Safety issues need to be addressed at the technical and social levels, but given the economic impact, it is also crucial to widely popularize access to superintelligence. The best way forward may be to: 1. Solve the alignment problem, which can robustly ensure that the learning and actions of artificial intelligence systems are in the direction that the collective really wants in the long run; 2. Focus on making superintelligence cheap, widely available, and not too concentrated in any one person, company, or country.