This episode of Warning Shots explores the AI stories that slipped under the radar: a new White House initiative, insurers distancing from AI risk, models hitting gifted-level IQ, and AI artists topping the charts. The guests argue these trends reveal how quickly AI is reshaping jobs, safety, and culture—and why society needs guardrails now.
In this week’s episode of Warning Shots, John Sherman sits down with Liron Shapira and Michael to break down the AI developments that—once again—barely made a ripple in the national conversation. While each news item looks small on its own, the guests argue that, together, they reveal an accelerating shift in how AI is shaping government, industry, jobs, and even culture.
Below are the key trends the guests highlighted, why they matter, and what they believe could come next for AI safety and public policy.
The Biden administration announced the “Genesis” scientific accelerator, an effort to use AI and national lab compute to speed up research and innovation.
Some see this as a positive step—especially in areas like medical discovery and scientific analysis. Guests on the episode noted that many people urgently need these advances.
But they also raised concerns about the framing. According to the guests, parts of the program resemble a large-scale, competitive push intended to outpace China. When governments tie AI progress to geopolitical rivalry, the incentives shift toward speed instead of safety. The guests describe this as a classic coordination trap: everyone accelerates because everyone else is accelerating.
One of the most striking developments discussed in the episode: major insurance companies have begun distancing themselves from high-risk AI deployments.
According to reporting referenced by the guests, some insurers believe the potential financial losses from advanced AI systems—particularly when embedded in supply chains, infrastructure, and automated decision-making—are too large to price.
The guests argued that when the industry whose entire business is evaluating risk steps back, it’s an early signal that the systems are becoming difficult to predict or contain. They emphasize that this isn’t about science fiction; it’s about practical, systemic vulnerabilities that institutions are already concerned about.
Another discussion centered on new testing showing AI systems scoring around 130 on human-normed IQ metrics. The guests stressed that IQ alone isn’t a perfect measure—but the trend is clear: capability benchmarks continue rising.
If models reach levels associated with creative problem-solving or scientific reasoning, the guests believe this could reshape the job landscape much faster than institutions can adapt. Knowledge work—long considered “safe”—may be first to feel the impact.
They also noted early indicators: drops in junior hiring, AI-driven productivity, and signs that recent graduates are already struggling to find entry-level positions. For the guests, these trends suggest that the economic effects are no longer theoretical.
One of the most surprising stories covered in the episode: an AI-generated artist recently hit #1 in country and gospel charts. These are genres highly associated with identity, tradition, and lived experience.
The guests explained why they see this as a turning point. Once AI starts outperforming humans in creative spaces long seen as deeply personal, culture itself could begin to shift. They worry about a future where AI models train on AI-created content, creating a feedback loop where authenticity becomes harder to find—and harder to define.
As one guest put it, if synthetic art becomes dominant, society may slowly lose touch with what “real” means.
Across the episode, the guests return to the same underlying theme: the systems are accelerating, but the world is not preparing at the same pace. Whether it’s government racing ahead, industries backing away, or AI reshaping culture faster than society realizes, the direction is consistent.
The conversation doesn’t claim certainty about timelines—but it does point to a growing gap between the speed of progress and the depth of public understanding. And this, the guests argue, is exactly why the world needs guardrails.
If the public waits until the impacts are obvious, it may be too late to shape policy, set boundaries, or preserve human decision-making authority.
If you believe AI should remain under human control, now is the moment to speak up.
Take one minute to contact your representatives and support evidence-based AI safety policy:
👉 https://safe.ai/act
The AI Risk Network team