Mike Smith
@selfawarepatterns.bsky.social
Programmer; aspiring fiction writer; interested in science, philosophy, history, science fiction, fantasy, skepticism
blog at: selfawarepatterns.com
blog at: selfawarepatterns.com
I agree machine consciousness will be different. And I often say that consciousness is in the eye of the beholder. When we call something "conscious" we're basically saying it's like us to one degree or another. There's no bright line between like-us and not like-us, just a broad blurry spectrum.
October 15, 2025 at 10:28 PM
I agree machine consciousness will be different. And I often say that consciousness is in the eye of the beholder. When we call something "conscious" we're basically saying it's like us to one degree or another. There's no bright line between like-us and not like-us, just a broad blurry spectrum.
Might depend on what we mean by "ego." Modeling self in world, along with preferences about the states the models represent, is far from an easy engineering problem. But I don't see any barriers in principle to solving it.
October 15, 2025 at 2:17 PM
Might depend on what we mean by "ego." Modeling self in world, along with preferences about the states the models represent, is far from an easy engineering problem. But I don't see any barriers in principle to solving it.
To me, it seems like a system is whatever it is regardless of its lineage. The hip joint evolved yet we can replace it with a designed product. I do think it's reasonable to ask why we'd bother giving AI an ego with everything that comes with it. But I struggle to find any barrier to doing it.
October 13, 2025 at 9:46 PM
To me, it seems like a system is whatever it is regardless of its lineage. The hip joint evolved yet we can replace it with a designed product. I do think it's reasonable to ask why we'd bother giving AI an ego with everything that comes with it. But I struggle to find any barrier to doing it.
I don't think there is any barrier in principle to an AI having those things. But I agree that an ego, not just self modeling, but also self concern, won't happen automatically. And I don't think we don't want to introduce it, since it could lead to much of what the AI doomers worry about.
October 12, 2025 at 7:37 PM
I don't think there is any barrier in principle to an AI having those things. But I agree that an ego, not just self modeling, but also self concern, won't happen automatically. And I don't think we don't want to introduce it, since it could lead to much of what the AI doomers worry about.