Where the Economy Thrives After AI

**Nathaniel Whittemore** (0:00)
Today on The AI Daily Brief, we're talking about the part of the economy that will thrive after AI.
The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.
All right, friends, quick notes before we dive in. Yesterday, if you haven't seen yet the bonus operators episode, we launched our latest free AI DB training program. This one is a NuFar Gaspar joint. She took a bunch of what I had built with ClawCamp and turned it into a more general, more extensible, adaptable, ongoing agentic operating system program. So if you have dabbled with OpenClaw or have wanted to, but want an agentic system that can slot in Clawed code or Codex or Cursor or any other model or harness that you're using, check out Agent OS. There are links to it from aidailybrief.ai. And when you sign up for Agent OS, you can actually do it with your new AIDB handle. I plan on doing a lot of these free programs. And so instead of you having to sign up with emails and passwords every time, we've created now an AIDB single sign-on, with of course, cool vanity handles. So even if you aren't signing up for a program yet, but you just want to squat on your favorite three or four digit handle, you can do that once again from aidailybrief.ai or directly at aidblabs.ai.
But with all that out of the way, let's talk about AI and the economy.
Welcome back to the AI Daily Brief and happy weekend. Listen, it will probably surprise none of you that I have what I would call major beef with the AI jobs discussion.
It is not just that I think that the labs themselves are doing an unfathomably horrible job pitching the value of AI. I've made the comparison in the past to pharmaceutical commercials. If you watch a 60-second pharma commercial on cable, which is basically the only commercials they have at this point, the first 45 or 50 seconds are spent around how this new drug or treatment is a miracle cure for some really painful, awful thing that people have otherwise had to just live with. Now, yes, at the end, there are those 10 to 15 seconds of disclosures around potential side effects, some of which can be very, very bad. But the emphasis is on the better life that you get to lead that justifies all of those risks. In AI, we're basically exactly the reverse. We spend the first 45 to 50 seconds of our metaphorical commercial, i.e. every time we get a chance to talk to anyone about AI, talking about how damn bad society is going to be after AI gets its hooks in, with only this tiny little consideration of the potential benefits, which are presented as very hand wavy at best. Anyways, that critique is not actually the core of my frustration with the jobs discourse. The core of my frustration with the jobs discourse is one, that I fundamentally think that all of these people who are claiming that we're going to see 15, 20, 30, 50% unemployment are just straight up wrong, and they're wrong in part because of our failure to actually try to spend even any time thinking about what jobs and employment and the economy looks like on the other side of AI and agents actually being fully integrated. To put it differently, as with any new technology, we are in a period of creative destruction. Always it is the case that the destruction is easier to see before the creation because the destruction tends to happen earlier. The creation is what gets born and it's very hard to see the shape of the new thing before it actually starts to come into being. And yet that does not mean that we can't spend time trying to understand by following reasonable traces and first principles, what might happen in the future that would change the shape of how we think about any destruction that happens in the present. The big area that I'm interested in, and I'm working on a bunch of things around that will come to the show at some point, is the idea that in an AI world, the constraints in the economy shift from supply, how much we can make of stuff, to demand and consumption capacity, how much we can actually consume of things, which in many cases is going to be a vector of time and attention.
In a world where the constraint shifts from the supply side to the demand side, the really imperative question becomes, in what areas could we consume a lot more than we are now?
Then of course from there, what would it look like to consume more? What are the types of consumption that's happening, and what is the infrastructure that's needed to build or provide those new things that we want to consume? The example that I'm spending a bunch of time on is healthcare. I think we consume a vanishingly small amount of the health services that we would if our system wasn't so expensive and so reactive. There are an immense number of opportunities for preventative care, data monitoring, tracking, and a whole infrastructure that could go around them, enabled by the decreases in cost that come from widespread implementation of AI into the health system. That argument isn't done yet, but it's why I was so interested to see an essay from economist Alex Imas start to really go viral and impact the jobs conversation, at least in our little corner of AI Twitter. Alex isn't exploring exactly the same themes, but as you'll hear, there are some similar interests here.