Andrew Critch’s recent threat model ends with the following:
We humans eventually realize with collective certainty that the companies have been trading and optimizing according to objectives misaligned with preserving our long-term well-being and existence, but by then their facilities are so pervasive, well-defended, and intertwined with our basic needs that we are unable to stop them from operating. With no further need for the companies to appease humans in pursuing their production objectives, less and less of their activities end up benefiting humanity.
Eventually, resources critical to human survival but non-critical to machines (e.g., arable land, drinking water, atmospheric oxygen…) gradually become depleted or destroyed, until humans can no longer survive.
I occasionally see posts by people who believe that surveillance advertising is bad, and we should try to write google out of our lives. Regardless of the merit of this argument, I admire the discipline it takes to degoogle. Gmail and gdocs are really high quality, add a ton of value, which is why they’ve become so entrenched. I can scarcely imagine actually doing without google docs, at this point!
It occurs to me: should we be practicing the skill of doing without something that’s pervasively adding a lot of value, just in case that skill helps to keep us from being enfeebled by an aligned AI system, or destroyed by a misaligned AI system? Is it also practice coordinating, which would payoff more generally than AI problems?