Jürgen Schmidhuber

I strongly believe that by 2028, we will have developed a general AI system that surpasses human intelligence in all respects. Our rapid advancements in neural networks and machine learning are pushing the boundaries of what AI can achieve. It's only a matter of time before we create a system that can outperform any human in every aspect.

In my view, large language models are a crucial step towards achieving artificial general intelligence. Their ability to process vast amounts of data and understand complex patterns mirrors the cognitive processes necessary for AGI. By expanding these models and integrating them with other AI technologies, we can move closer to realizing true general intelligence.

As a strong proponent of AI and superintelligence, I believe entrepreneurship will not only be possible but will likely flourish in unprecedented ways. Superintelligence will augment human creativity, enable solving complex problems, and open new markets and opportunities that we can't even imagine today. Just like past technological advancements have spurred innovation, superintelligence will empower entrepreneurs to explore new frontiers.

I believe that while transparency is important, requiring model cards to be published before commercial release could impede innovation. Researchers and developers should focus on pushing the boundaries of what's possible, and overly stringent disclosure requirements might stifle creativity. It's crucial to find a balance where innovation is not hindered while addressing ethical concerns.