Raw LLM Responses

> One works with a deterministic set of instructions, and the other through symbols that inherently carry some degree of uncertainty. This is such a terrible way to distinguish carbon-based computation from silicon-based computation because not only do we already have “non-deterministic” silicon-based hardware that leverages the voltage continuum rather than just binary bits, but it has been specifically designed for machine learning / AI applications and is poised to become extremely widespread with the adoption of AI technology since it is substantially more energy-efficient for the same applications. More to the point, I really wish highly trained neuroscientists, psychologists, cognitive scientists, and philosophers in general would realize that silicon-based information processing hardware does not have to have a von Neumann or any other kind of deterministic architecture. This is the kind of mistake that people look back on decades later and think “How could they have been so dumb?” You don’t even need hindsight for this, you just need to open your eyes to the current state of research in alternative computer architectures, i.e. do your homework before opening your mouth (or in this case touching a keyboard).