Lex,

Garbage in, garbage out. That much I will agree to. The richer (more salient) the input stream, the better a system can reason about its environment. But to imply that this is exclusive to animals and animal sensory bands is ridiculous. Also, the arbitrary concentration on external input, input from the "environment" vs information from within the thinking entity seems problematic at best. It might be true that entities evolve into existence simply and precisely because their very structure reduces the scale of the calculation that has to be made in their absence. Also, there are valid arguments to be made against wide-input streams. Wide input just adds more hay that must be sifted to find the needles. I think you are confusing input method with the natural structural optimization that occurs around any input method. I argue that input method history can easily become a hindrance to the development of salience filters. Animal vs. machine? I see no difference at all. Show me how animals are not machines.

If an idea is simple, explaining it should be simple as well. I never buy the argument that an audience isn't "ready" for an idea. Make us ready! Prep us. It shouldn't be that hard to do. Stop talking about the obstacles standing in the way and start building a path. We aren't Lex,

Garbage in, garbage out. That much I will agree to. The richer (more salient) the input stream, the better a system can reason about its environment. But to imply that this is exclusive to animals and animal sensory bands is ridiculous. Also, the arbitrary concentration on external input, input from the "environment" vs information from within the thinking entity seems problematic at best. Also, there are valid arguments to be made against wide-input streams. Wide input just adds more hay that must be sifted to find the needles. I think you are confusing input method with the natural structural optimization that occurs around any input method. I argue that input method history can easily become a hindrance to the development of salience filters. Animal vs. machine? I see no difference at all. Show me how animals are not machines.

If an idea is simple, explaining it should be simple as well. I never buy the argument that an audience isn't "ready" for an idea. Make us ready! Prep us. We are the people willing to talk to the matrix within the matrix within the god within the teacup within the mote within the budding n-verse within the laughing Buddha's dream in the pure information binary all-space spun into the the infinite quantum dot you thought was a blackhead on your nose. If you can't explain an idea to AI enthusiasts, it probably isn't an idea. Shouldn't be that hard. Stop talking about the obstacles standing in the way and start building a path. We aren't opposing your ideas. We are frustrated that you seem not to care enough to walk us through them.

Randall

P.S. I enjoyed your explanation of the bottom up approach… simple to complex… that you feel is missing in AI research. Though I must say that there are a great many people working at this level. There is a crossover at this depth between AI and pure logic, set theory, mathematics, linguistics, complexity science, information science and the study of evolution. Don't confuse AI, the science, with the arbitrary group of people who frequent Monica's meetup or the discussion any of us are having in this online portal. That would be like reading Jaws and determining that all books are about sharks. Silicon Valley isn't Monica's fault.

I specifically asked you to expose your motivations as regards AI. Still waiting.

Randall