- Unreliable and buggy core libraries (basic fp math, stats, autodiff) which seems to stem from an academic-style disinterest in focusing on the boring bits of language foundations

- 1 indexed arrays

- extremely slow start up times (1 min??)

- no way to create portable static binary apps (disadvantage vs rust/go)

- lack of tutorials, demos

- julia is not really faster for math when compared to jax/tf/pytorch/etc anyways