Why AI Agents Concern Me
If Meta had its way, all of us would hand off the mental bandwidth required for buying goods and services to a digital Muse agent that’s named as if it’s our pet. Everything from toilet paper and birthday cakes to car insurance and home repairs would be run through our Muse who would make the necessary choices, decisions, and trade-off analysis on our behalf. We are then left merely to consume the fruits of our Muse’s labor.
If OpenAI had its way, we would train various Dots to do all the things that we don’t like or want to do ourselves. The sky would be the limit. We could create our own virtual Dot workforce complete with a chief of staff, VPs, and analysts, presumably each given its own pet-like name, with everyone working 24/7 on our behalf. To complete the CEO feeling, we would be able to call any of our Dots on a whim and either reprimand them or have heartfelt discussions regarding how they are performing.
My charitable reaction to all of this is that we are seeing early iterations of technology that has come a long way in just the past year or two but which is very raw and in search of problems to solve. The pace at which companies have taken OpenClaw and set their sights on broader markets beyond early adopters makes me nervous. Both Meta and OpenAI are desperate to begin monetizing their significant AI investments. Both companies think it is better to ship something now and ask questions later rather than wait for something more thought out to come along.
A more heartfelt reaction to all of this is that we are bein
This article is available exclusively to Above Avalon members. Become a member using the payment forms ($20/month or $200/year) below. Members can continue reading this article here.
An audio version of this article is available to members who have the Podcasts add-on (+$10/month or +$100/year) attached to their Above Avalon membership.
