I think Linus and Junio for git is a no brainer. If they don't get that by 2030 it would be silly.

Wolfram also is a no brainer.

Deep learning will get at least one more. Chris Lattner and the llvms perhaps? I don't know what work led to the M1 but it feels ground breaking.

I'd also like to see perhaps aaronsw and alexandra elbakyan win one.

I think my collaborators will win won for 2-D/3-D languages if this prediction comes true: https://longbets.org/793/ . Sounds arrogant and probably a mistake in my math but I'm betting on it.