ESRogs comments on [link] FLI’s recommended project grants for AI safety research announced