the fact that some of you are suspicious if my posts were GPT-3, frankly, tells more about you and your state of knowledge about current SOTA machine learning capabilities than this typical (not so controversial idea) post.
Actually disagree. One can do personal tests using your own brain. One which Iāve done recently is related to my car showing inaccurate location of objects on the road or the side of the road. It is clear to me the carās NN is trained relative to the car. But our brains reference trash cans and cars on the road not to our location but to the roadās location. You can test this by shaking your head and see if objects separate themselves from the surface they are sitting upon. If not, then clearly the coffee cup on the table is referencing relative to the table and not to your location. It is clear that my carās NN is not trained in the most robust fashion to reference the trash can relative to the sidewalk and the car relative to the road and thus the trash and car can can float around. To develop a more robust system, the developers need to train the NN to use the appropriate coordination reference. Introspection like this is useful to motivate development approaches for commercial systems and research.
Sorry.
1 Like
This has turned out to be one of the most interesting threads on the entire forum, but like Seinfeld, itās a thread about nothing.
2 Likes
And many more. If you canāt understand plain English definitions and see why what you say is wrong, then I have no way to persuade you.
Introspection is the examination of oneās own conscious thoughts and feelings. It is not ideation. Period.
I think thatās a good way to look at it: introspection might get the ball rolling but you need real science to make real progress. All I ever asked the OP was to express his idea in a testable scientific form.
The real problem with introspection is we think it tells us something useful but our brains lie to us. What we introspect is a sham, created by a āsubconsciousā to keep the āconsciousā satisfied that it is in control. It really isnāt.
And that can be proven scientifically.
2 Likes
I know what introspection means, Itās just not enough to know it, as I said, by those definitions, I do introspection 100% of the time and if I stop introspecting, then Iāll not be conscious.
Those plain English definitions are good at explaining the concept, but they dont define it in a sense concrete enough for us to use in a discussion like this.