About me

Hi, I’m Carter. My background is in engineering and I’ve spent most of my career working on technology in healthcare, most recently at Google, Amazon, and two startups. Along the way I’ve been fortunate to help build products on each major AI/ML wave: computer Vision + classifiers, deep learning, NLP, and LLMs + Agents. Because I was the product or operations person, I’ve been focused on how these products create value for customers, what is the edge of what’s possible, and how that has expanded over time.

Rapid change in AI state of the art

I used to describe AI systems as “Really smart and really dumb.” For example, the system we designed to detect diabetic retinopathy (a type of eye disease) was as accurate as a specialist but it couldn’t spot glaucoma or any other eye disease and would be useless at identifying skin disease.

Today’s frontier models have far greater breadth and performance.

The same model can write code, analyze a legal document, and support scientific research. And when it’s doing these tasks, it’s doing them as well or better than most humans.

The result is we have technology that can now do most cognitive tasks better than most humans.

Preparedness gap

Because breadth and performance of AI has increased so dramatically we are simply not prepared. We do not have good answers about how we best utilize AI as individuals, as companies, or what that utilization means for society. There aren’t good answers inside the biggest companies that are working on AI and there aren’t good answers inside of government either.

The best way I know how to develop a point of view and make a case for it is to write so that’s why I’m doing here.

Smash that subscribe button

Subscribe to get full access to the newsletter and publication archives.