That same link puts AI risk under the "far future" category, basically the same category as "threats to global food security" and asteroid impact risks. What's unreasonable about that?