I find viewing these words through the Rights lens to push back on the Utilitarian viewpoint in that it focuses less on the good, and more on whether or not dignity is being protected on a more human level. Basically—even if rapid-running AI produced major benefits for society as a whole, that would not necessarily justify exposing individuals to potential risks without their consent. The Rights lens raises questions for the individual that tends to get buried beneath the heavy benefits of AI. I’d say that the Rights viewpoint overlaps with that of the Utilitarian one on a more opposing scale, as the Rights lens more or less poses questions straight from the results of Utilitarian practice.

Through a Utilitarian lens, it seems Amodei’s main concern surrounds balancing the potential benefits of AI against the possible disadvantages that could come from the rapidity. There are obvious societal benefits that include scientific advances and increased productivity, and from a utilitarian perspective, those potential benefits matter because the ultimate goal is to produce the greatest amount of good while minimizing harm. In the same breath, Amodei argues that companies should pace the frontier because the risks have the potential to exceed our ability to understand and control said systems. In other words, there is a very justified fear of AI exceeding our control and taking on a mind of its own. Through this lens, I see slowing development not as a rejection of AI, but a call to make sure the benefits do not emerge at the cost of harm. The stakeholders, in this case, would be anybody whose livelihoods have the potential to be affected by AI, as well as the future generations that could inherit the technology we shape.