I started out a bit skeptical listening to the All-In podcast, but I really like the in-depth expertise in different areas this group has.

While it’s at times a bit too upper class and disconnected for my taste, the insights are great. I also love how they explain things to each other.

Some insights I took from the episode with Sam Altman from OpenAI:

  • There are two ways people want AI to develop. One is to replace themselves and act as their entity. The other is to be a separate entity like a senior employee that can reason or object to objectives given.

  • A graphical interface is important for a system where AI and humans coexist. If an AI does something for you, you may want to follow its actions on a screen or have visual overviews of options to choose from in a process. Fully conversational and audio interactions do not work for every use case, and visual interfaces (like screens) will remain important.