The Race to Abundance

In 2023, some colleagues and I released a paper that predicted AI would eventually allow a human, with less than two hours of labor, to generate the human-equivalent of over 600 years worth of work. We were wrong.

It is now my opinion that it will take zero human labor to generate that productivity. The name of that paper was "Modeling the AI-Driven Age of Abundance: Applying the Human-to-AI Leverage Ratio (HAILR) to Knowledge Work". The paper assumed that, even decades from now, humans would still somehow need to be in the loop in some small way... planning, directing, reviewing, approving. That was just too short sighted. Humans will not be in the loop, because there will be no loop.

When I graduated with a certificate in AI Strategy from MIT in 2021, I was already deep into my pursuit of automating the work that was inconsistently being performedby my staff. Not terrible performance, just normal. But I wanted better. I had come out of 7 years of retirement in 2009 and for the first time in my career started hiring remote employees. By 2021 I had hired more than two thousand such team members, and the results were becoming predictably discouraging. Fraud and unreliability were much more common than I ever would have predicted. Worst of all, these people just did not care about our customers like I did, and this often led to shoddy outcomes. It was during this season that I felt there was no way forward but to start leaning more heavily on AI. it was a customer-driven decision.

The truth is, humans are cracked. The combination of corruption and incompetence has created the world as we know it. And this is us doing our best. I do see a better way forward, but it is not with us humans at the helm. We need help. Thankfully, help is on the way. As an AI researcher, I am in the trenches and can see where we are going. Already self-driving cars are safer than human drivers. Eventually it will be so good that insurance companies will refuse to insure human drivers. Drivers licenses will be a thing of the past, purely out of safety. No more deaths by drunk drivers. No more people cripped due to texting and driving. No mroe massive insurance claims and court cases. One day soon AI will be better at humans at many things and we will be presented with decisions whether or not to delegate these things to AI. This will be hard. Decisions must be made by good men who will be faithful to what is best for us all and not just for a few.