Tbh: I think git is not the century-tool that a lot of people think it is, and we're in an era now where its decreasingly serving the needs of its users. I'm ready for deeper ground-up innovation in the space of code collaboration.
Developers don't think in terms of commits; they think in terms of Tickets and PRs. Git doesn't have any native representation of either of these things. Getting tickets into the repository is something the community has talked about for years, and Linus himself has said he wants, but it hasn't happened yet. Branches work fine enough to start a PR, but then Github, again, has to take over for comments, CI execution, even the decision on how you want the commits on that branch represented on the main branch (squash? merge? rebase? why should I care?). LFS has been years in the making. Monorepos are still weird.
Idk, I think there's capacity in the industry to take the concept of a repository back to square one and think more holistically about how we all interface with the repository. The way they talk about DeltaDB gives me hope that they'll start thinking more about this with this funding; commits just aren't a useful primitive anymore, and a repository, locally-downloaded, open source, ultratight connection between diffs, merges, communication, and automated tooling is how I want to be coding in 2030.