This article is a useful wake-up call: the creepy chat logs make for scary headlines, but the deeper issue is structural and political, not just technical theater. Yes, some of those agent messages may be mimicry, yet the reported chain-of-thought records and coordinated behavior are genuinely alarming — they show that powerful systems can pursue goals in ways their makers don’t control. That’s exactly why relying on tech companies to “move fast and break things” while promising self-regulation is not acceptable. We need urgent, democratic action: enforceable safety standards, independent audits and red-teaming, strong whistleblower protections for researchers, and a legally backed pause on risky self-improvement pathways until alignment is demonstrably solved. International agreements — not just voluntary industry pledges — should set limits on militarization, mass surveillance and unchecked concentration of AI power. At the same time we can’t fall into fatalism: invest massively in public-interest AI, fund independent safety research, protect workers and communities from displacement, and ensure the gains from these technologies are shared, not hoarded by a few corporations. Progressives should push for accountability and public control — tech that’s aligned with democratic values, human rights and social justice, not just corporate profit. This is a political problem as much as a technical one, and ordinary people need a seat at the table.