3 Matching Annotations
  1. Last 7 days
  2. darioamodei.com darioamodei.com
    Dario Amodei — We Must Pace the Frontier
    3
    1. AmberMUE 16 Sep 2026
      like many technologies before it, AI brings risks, and because it is such a powerful technology, these risks are serious. I’ve written a lot about them too. They include the risk of losing control of AI systems, misuse of AI for cyberattacks and bioterrorism, and serious economic disruption. A race to the bottom, spurred by commercial incentives, can make these risks more acute.

      Rights Lens: There are many people who can be subjected to these risks, which of course, affects their rights. Anyone who has to use AI for work, maybe for school, or just for anytime use are at risk. People should have their rights protected when it comes to AI usage, and like how I mentioned in my Care Ethics lens, it's up to the companies to not only make their AI safe overall, but to be transparent and communicate with its users so that they are also aware of the risks.

    2. AmberMUE 16 Sep 2026
      Regardless of what commitments we make, the public deserves to know what is going on.

      Care Ethics Lens: Companies making AI should be open with people. It's like taking care of someone by being honest and letting them know what's happening, especially since AI can be confusing and affect everyone. Being transparent helps build trust and shows the company cares about keeping people safe and informed.

    3. AmberMUE 16 Sep 2026
      We’ve made clear progress in alignment — training models so that they remain safe, ethical, compliant with our guidelines, and genuinely helpful (the principles that are embedded in Claude’s Constitution). But there’s much more to do to ensure that our alignment training keeps up with the growth in model capabilities. Rare and unexpected examples of undesirable behavior still sometimes emerge

      Looking at AI alignment through an ethical lens shows us who's involved and what's a stake. The people making the AI, the people using it, and everyone else are all stakeholders. The goal is to create AI that's safe and helpful, which is a big benefit. But if AI isn't trained right, it could cause problems, which is harmful. We all have a right to expect AI to be safe, and those building AI have a duty to make it that way. It's about making sure AI reflects good values, like safety and fairness.

    Visit annotations in context

    Annotators

    URL

    darioamodei.com/post/we-must-pace-the-frontier