Anthropic devoted 94 pages of its initial public offering prospectus to risk factors, detailing risks that advanced artificial intelligence could pose catastrophic or existential threats to humanity