SoerenE comments on Does Kolmogorov complexity imply a bound on self-improving AI?