Anthropic’s ‘Mythos’ AI proves that obsessing over AGI is folly

# Anthropic's New AI Shows Why We're Focused on the Wrong Thing Anthropic released a new AI called Claude Mythos that's unexpectedly good at finding security holes in software—better than most human experts—even though that wasn't what they trained it to do. This reveals an important truth: instead of obsessing over creating a super-powerful AI that can do literally everything, we should pay attention to the very real problems AI is already creating in specific areas like cybersecurity and data protection. The real-world impact matters more than the sci-fi scenario.