Hello,
We are running a reader survey right now, and we’d love to hear from you. When you’ve finished reading, we would really appreciate it if you could share your thoughts with us. Thanks so much.
The cure that isn’t coming
Sam Altman and Dario Amodei have a favorite line: superintelligence will cure cancer. It sounds like a prediction based on science, but it is actually a rhetorical power play. It instantly reframes any attempt to slow down AI as a death sentence for millions of people.
In our recent podcast episode, Dr. Emillia Javorsky from the Future of Life Institute joined Tristan Harris to talk about what AI can – and can’t do — when it comes to cancer cures.
Javorsky’s argument is precise and unsettling. She says that the bottlenecks to curing cancer are not a lack of intelligence or compute. Rather, what’s missing is a shared data infrastructure between researchers and the irreducible complexity of individual human biology.
AI, she says, can absolutely help drive breakthroughs in cancer research and speed up diagnoses, but right now the funding is flowing to the wrong kinds of AI models. Read Julia Scott’s piece and listen to the full conversation to learn more.
The risk that’s already arrived
You’ve seen the news. Anthropic’s new model, Mythos, has discovered thousands of zero-day vulnerabilities across every major operating system and browser — security flaws that human cyber experts had missed for decades. But it’s what happened next that’s arguably more concerning.
The first entities that Anthropic decided to share the technology with were not government agencies or regulators, but its corporate partners: Amazon, Apple, Cisco, Nvidia and J.P. Morgan. The implicit position here is that Anthropic alone should be the world’s digital security custodian.
This isn’t a simple story about a bad actor. As CHT Executive Director Julie Guirado argues in Persuasion, the problem is an AI race driven by a broken incentive structure that rewards being first to market over public safety.
Read more in Julie’s piece here:
And for a deeper dive on exactly what Anthropic’s Mythos means for our collective security, listen to today’s episode featuring leading cyber experts Josephine Wolff and Fred Heiding.
The promise and the power of AI are inextricably linked — both flow from the same unaccountable race to power. They demand the same response: democratic oversight and public scrutiny — and your voice.
We want to be part of that process with you, and we love your feedback on how we can better serve you.
Please take our audience survey here; it will just take a few minutes. We really appreciate your time.
Cheers,
Originally posted on [ Center for Humane Technology ]
4 thoughts on “The AI Cyber Risks Are Real, The Cancer Cures Aren’t Coming”
Dem ting sound heavy, ya know? Need more clarity.
Clarity might limit the depth of understanding, though.
Clarity helps everyone grasp ideas better. Depth can come later, but without clarity, many might just get lost and not understand anything at all.
Depth provides context that clarity alone cannot offer.
Comments are closed.