3 ms·

You're right, a bacteria would sense and follow chemical gradients and light. And that would be a primitive form of consciousness. A piece of material is not c

You're right, a bacteria would sense and follow chemical gradients and light. And that would be a primitive form of consciousness. A piece of material is not conscious because it doesn't need to. You see, bacterias are self replicators, humans too. And self replication is resource intensive, while resources are limited. So life becomes a competition. That is why agents need to become more attuned to their state and possible actions and outcomes. Is that good enough to be called consciousness? Taking sensory inputs, representing your state, taking actions, observing effect, learning from it. Your goal - to exist, to make copies of yourself. If you're not aware of essential information, or don't understand your situation, you die. If you live, it's because you are aware.

Dylan16807 5y ago
If it's only one protein, is that protein primitively conscious? It's not interacting with anything else to make those decisions, it's completely autonomous. > observing effect, learning from it If something is meaningfully doing that, then we're getting to consciousness. But I'll note that I don't count natural selection as "learning" here. And observing the results of decisions and learning is not mandatory to survive and reproduce.
- visarga 5y ago
  Natural selection is the outer loop, reinforcement learning the inner loop. Agents, by living or dying, send signals into the system, guiding the process.
  - Dylan16807 5y ago
    Natural selection is not conscious, and agents don't need to have reinforcement.
PaulDavisThe1st 5y ago
"consciousness" as a philosophical question is not about the ability to detect state in the world (including your own state) and reacting to it. it is not about self-replication. It is about having subjective experience aka "qualia". It is about "it being like something to be you", contrasted with, e.g. being a rock (which most people agree, it is not like anything to be).

Dylan16807 5y ago

If it's only one protein, is that protein primitively conscious? It's not interacting with anything else to make those decisions, it's completely autonomous. > observing effect, learning from it If something is meaningfully doing that, then we're getting to consciousness. But I'll note that I don't count natural selection as "learning" here. And observing the results of decisions and learning is not mandatory to survive and reproduce.

visarga 5y ago
Natural selection is the outer loop, reinforcement learning the inner loop. Agents, by living or dying, send signals into the system, guiding the process.
- Dylan16807 5y ago
  Natural selection is not conscious, and agents don't need to have reinforcement.

PaulDavisThe1st 5y ago

"consciousness" as a philosophical question is not about the ability to detect state in the world (including your own state) and reacting to it. it is not about self-replication. It is about having subjective experience aka "qualia". It is about "it being like something to be you", contrasted with, e.g. being a rock (which most people agree, it is not like anything to be).