I don't see why it's not possible.

In your example scenario, it sounds like Bob is vibe-coding a project in assembly language, and then using an existing assembler to create the final binary?

I have not tried it, but I would imagine that current LLMs are relatively weak on generating assembly language, just due to less thorough training compared to higher-level languages, but even if so, that's surmountable.

As for, what I think you are suggesting, having the LLM also do the assembly step? Again, in theory, sure, but I would think just using the existing known-good assembler would be preferable to training an LLM to convert assembly language to binary. I'm not sure what you would gain in terms of either speed or overall productivity to have the LLM itself do the assembler step?

This will essentially always be more expensive than generating code a compiler can use, because the compiler and intermediate representation rule out, ex ante, a vast logical space of binaries that will crash.

The basis of this idea is grounded on fantasy.

You understand that one wrong byte means that the program crashes? Why use a tool that is totally probabilistic to generate assembly directly which must be correct?

Why?