The AI industry is having its loudest debate yet about whether its technology poses an existential threat to humanity. The current discussion began when AI researcher Jacob Coxon announced his resignation from Anthropic, saying the leading AI companies are "gambling with our lives" by racing toward self-improving superintelligence without adequate safety measures.
The situation escalated when Anthropic's alignment lead publicly endorsed Coxon's warning, posting that the team "really does earnestly believe AI could kill all humans" and estimating the probability at greater than 10 percent within the next decade. The exclamation mark on that statement became one of the most-discussed punctuation marks in tech this week, underscoring how seriously even the people building these systems take the risk.
On TechCrunch's Equity podcast, the team dissected the debate from multiple angles. One host questioned whether these existential warnings are partly a "weird way of flexing to show how far advanced their company's AI model is," particularly as companies like Anthropic prepare for their initial public offerings. Another wondered how these risk disclosures would appear in an S-1 filing — imagining junior lawyers having to rewrite sections to acknowledge that their company officially believes there is a significant chance it could develop something that eradicates humanity.
The debate arrives as Anthropic CEO Dario Amodei published his own plan for "pacing the frontier," including giving independent safety evaluators access to leading models and developing common safety standards across companies. Whether this represents genuine concern or strategic positioning ahead of a major IPO remains the central question of the moment.
