Are new proof of work ideas still being developed, or is everything effectively moving to PoS at this point?

The consensus conversation has been dominated by proof of stake for a few years now. Most newer projects default to PoS because the infrastructure overhead is lower and the scaling story is cleaner.

That said, Bitcoin's dominance means PoW is not going anywhere, and there are still meaningful projects running PoW that aren't Bitcoin.

What I've been trying to understand is whether anyone is still experimenting with PoW at an architectural level - not just different hashing algorithms, but actual changes to what the mining infrastructure is doing. The question isn't just "PoW vs PoS" - it's whether PoW itself can evolve.

While reading into this I came across Qubic which seems to be testing approaches where mining hardware contributes to useful computation beyond hashing. Interesting to see someone actually building this rather than just theorizing.

Do you think PoW has meaningful architectural innovation left, or is the development energy essentially gone from that side of the space?