Is anyone interested in a curiosity-driven"brain OS"core for AI?(not an expert)

Why this idea? When we compare the human brain and modern PCs, computers actually have much higher memory/storage and calculation speeds per “tick” than the brain’s moment-to-moment thinking. But humans are amazingly creative and flexible with extremely little working memory or raw data. Yet most current AI tries to “stuff” everything inside the model, keep all data/statistics, and just scale up. What if we built a SMALL core that works more like a human “brain OS”:

makes its own questions (“Why?” “What’s weird here?” “What’s interesting?”)

only keeps knowledge as loose indexes (like “I think I heard this somewhere”) and looks up info as needed

links outcomes/decisions with simple feelings or experience (“That felt good/bad/weird/curious”)

works on normal hardware, not just massive clusters

Why do this? I feel that trying to make everything “bigger” misses the key: Human thinking is about curiosity, flexible judgment, and ongoing motivation—not about big memory or giant models. So maybe AI should have a small, motivation-based “core” instead!

Example pseudocode (for illustration):

python class BrainOS: def __init__(self): self.curiosity = 1.0 self.memory_index = {} # Index only self.emotion = 0

    def motivate(self):
        if self.curiosity > 0.5:
            return "Why?"
    
    def recall(self, query):
        return self.memory_index.get(query, None)

    def learn(self, result, feeling):
        self.memory_index[result] = feeling
        self.emotion += feeling

So…

Has anyone tried to build something like this?

Am I missing obvious research or past attempts?

Is anyone interested in talking or collaborating to try a prototype?

Any advice, feedback, or info welcome!

Thanks for reading—I’m passionate but not technical, so any kind of comments or help are greatly appreciated!