AI/ML consciousness requires emergent needs

In a prior posting (https://news.ycombinator.com/item?id=23048919) I speculated that "consciousness is emergent need". That is, to quote: ------------------- "First, understand the 'emergent' behavior. Traffic jams "emerge". They are not related to cars. They can happen with people pushing through an exit or sand clogging up a hole. Second, understand "need". Things that impact a system and threaten its smooth functioning create a "need". For instance, you "need" to urinate. A cell "needs" to eliminate waste. A city "needs" to remove trash. Needs generate a response. The response (urinating, cell waste dumping, trash collection) is a systematic attempt to reduce the "need". "Consciousness" is the "awareness of the need" giving focus to a response. Computers show no signs of having any need and thus are not conscious. " ------------------- AI/ML systems focus on sensing and recognition. I am suggesting that there should be a focus on defining and creating the idea of a "need" that an AI/ML system must fulfill. Define a system that is homeostatic, and needs to create an equalibrium. This already exists in some robot systems that have a goal. I'm suggesting multiple, competing goals where the AI/ML "needs" to maintain the balance. A paperclip factory "needs" to schedule trucks to bring inventory but also to remove machine shavings (waste), resharpen blades, schedule packaging material, etc. How complex must a system be to "define new needs" based on environmental change. Insects don't seem to adapt to damage of environmental change. "Higher level" organisms, such as an Octopus, will adapt. Ultimately, it seems to me that "self-modifying systems" are at the highest level of adaptation, and thus, of consciousness.