What's 1.05 x 0? Even if you run it 1000 times, it's still 0.

Are you trying to say that the capability of modern LLMs is zero?

Because if so, I'm pretty sure any frontier LLM is better at evaluating AI capabilities than you are.