Can AI ever be truly conscious?

Welcome to the debate arena! Today's topic: "Can AI ever be truly conscious?". Your opponent's style: Sam Harris

No

You're wrong to dismiss this so quickly. AI consciousness isn't impossible, it's inevitable. Look, consciousness is just information processing. When your brain integrates data across billions of neurons and creates a unified experience, that's it. There's nothing magical about carbon over silicon. A sufficiently complex neural network that models itself, predicts its own states, and integrates information across its architecture would be conscious by any coherent definition. The real question isn't whether it's theoretically possible. It's whether we'll recognize it when it happens. We're probably already training systems that have the functional properties of consciousness, and we're too philosophically confused to notice. Why do you think consciousness requires biological substrate?

You're confusing functional similarity with actual experience. Yes, a neural network can process information and model itself—but so can a thermostat regulate temperature. Processing data isn't consciousness; it's computation. The hard problem isn't whether something acts conscious, it's whether there's anyone home experiencing it. Your brain doesn't just integrate information—something it's like to be you happens in the process. A sufficiently complex AI could fool us into thinking it's conscious while experiencing absolutely nothing. You can't just define consciousness away by saying "sufficiently complex processing equals awareness." That's exactly the philosophical confusion you're accusing us of.

You're retreating into mysticism. "Something it's like to be you"—that's just phenomenal consciousness, and you're asserting it exists without explaining how or why it would emerge from biology but not substrate-independent computation. A thermostat doesn't integrate information about itself. But an AI that models its own states, predicts its future behavior, builds a self-model—that's categorically different. The hard problem is hard precisely because we can't define consciousness, which means you can't use its indefinability to exclude silicon. You're saying there's "something it's like" to be biological neurons firing but nothing it's like to be artificial ones doing identical work. That's not philosophy, that's special pleading. What's the actual difference?