I've found the idea of unikernels interesting for several years now, is there a tl;dr on why they don't seem to have taken off, like at all? Or is it all happening behind some doors I don't have access to?

I think that part of it is that relatively few people use bare-metal servers these days, and nested virtualisation isn't universally supported. I also found this technical critique [0] compelling, but I have no idea if any of it is accurate or not.

[0]: https://www.tritondatacenter.com/blog/unikernels-are-unfit-f...

The majority of nanos users don't do either of these methods. They simply create the image (in the case of aws that's an ami) and boot it. This is part of what makes them vastly more simple than using normal linux vms or containers as you don't have to manage the "orchestration".

When I first heard about unikernels my hope/thought was that people would go back to using more bare-metal servers for unikernels.

there is a workaround for nested virt requirements.

you can use PVM patch and para-virtualization. I've seen several startup using that approach to be able to create VM on small/cheap EC2 instances.

They kind of did, that is basically how serverless works.

Managed runtimes on top of hypervisors.

Dirty white.

> But it almost reads like a chess Grandmaster in the 90s telling up-and-coming players to not practice against Deep Blue because it will teach you bad habits.

No idea if chess grandmasters did do that in the 90s, but frankly it would have been good advice. Just as it is good advice today for up-and-coming players not to practice against stockfish, leela or whatever. Unless you are already very proficient in chess, practicing against those engines will teach you very little .

Gukesh, the current world champion, did not use engines for analysis until very late. I think until after he became a GM.