Before you fade Muse, use it for 72 hours

Here’s how (and why) I’d recommend investors experiment with consumer agents like Muse

I’ve spent a lot of time writing about and playing with the new wave of consumer AI agents (mainly Muse, with a little bit of Instinct mixed in). Those posts have generated a ton of inbounds, and they largely land on two extremes. On one end: “consumer agents are the most overhyped product since Google Glass / Apple Vision Pro and will have minimal impact on real businesses.” On the other: “every company that touches a consumer is a short as agents eat into their margins.” I’m not being hyperbolic; people really are that far out on both ends!

I think both camps are too extreme…. but if you forced me to pick, I’m closer to the second one. Yes, I said last week that the “consumer inertia” sell-off looked a little overdone. But I just can’t watch how easy Muse is to use, and how fast downloads have taken off, and conclude it won’t have a big impact on the consumer economy.

I’m in the minority there among my investing friends; something like ~75% of the inbounds lean “agents are overblown,” and I think there are two reasons why.

The first is readership bias. This is a value blog, and value investors are a curmudgeonly bunch (I say that lovingly as one of them!). We’re wired to fade hype and zig when everyone else zags. Usually that’s a feature, not a bug. The most dangerous words in finance are “this time is different,” and if you’re betting that a new product (like consumer AI agents / Muse) is going to reshape the consumer economy, you’re making a “this time is different” bet and those are normally good to fade. But AI is really revolutionary…. this time really might be different!

The second is the one that worries me: most of the skeptics I’ve talked to haven’t really used it. Sure, they’ve tried the product, but they haven’t immersed themselves in it. When I ask a friend who’s dismissing Muse what they’ve done with it, the answer is almost always some version of “I asked it a few questions on day one, didn’t want to hook it up to my email, and haven’t opened it since.”

That matters because of an old rule: what the power user does today is what the average user does tomorrow. Whatever the heaviest home internet users were doing a few years ago is roughly what the average household does today. The way college kids binged Netflix in the early 2010s is how everyone watched TV by the end of the decade; if you were betting that linear TV was a winner because “only some drunk college kids will binge streaming” then you had your face ripped off.

I think AI follows the same rule, but much, much faster. What the bleeding edge is doing with agents today is how everyone will use them in ~18 months (or sooner). If you haven’t really used the products extensively, it’s going to be hard to see where the puck is going or how these could change the consumer landscape.

The 72 hour challenge

So here’s my ask: download Muse, give it access to your email, and use it for everything for 72 hours. It’s easy, takes basically no effort and no skills, and once you get in the habit of asking it to do everything, you’ll start to see how fast things could change. And I’m not getting paid for telling you to do that or anything (though if Meta¹ wants to write me an influencer check, I would happily cash it!); I just really think these are going to cause a change for a lot of businesses and I don’t want any of my investing friends to miss something because they’ve been dismissive of AI agents.

That’s my ask; I’m going to spend the rest of this article detailing how I’ve tried to immerse myself in AI agents (so you can follow along / copy me if of interest). I just got the new iPhone (don’t judge; my old one was over three years old!), which lets you turn the old silent switch into a button that opens any app. I made that button connect to Muse, so literally anytime I need something, I have one tap and I’m in Muse. And I mean anything; if I have a question I’d normally Google, I pull up Muse. If I have something I want to do, or a reminder I want set, I pull up Muse and tell it.

But before I get to the examples, a request. You could read these and think “those are small potatoes! So Muse saved Andrew a few bucks here or there, or made it easier for him to find a Knicks game for his family. So what?” Remember this: these are all tricks some increasingly middle aged dude discovered within a few days of a brand new AI agent rolling out. It took a few years for killer iPhone apps like Uber to show up, for consumers to really get used to how to use an iPhone and for the iPhone to start impacting consumer habits and spending. So don’t read this and think it’s the end of what AI agents can do; read it and think “man, if a middle aged guy can do this within a week of the apps releasing, what’s going to happen in two years when the AI models are twice as smart, every iPhone is rolling out with agents preinstalled, and every Instagram story is telling you the best way to use AI agents to make your life easier?”

Request out the way, let me give a few of the ways I’ve used AI agents:

Upgrading my phone. I asked Muse for the cheapest price on the new iPhone and AirPods (again, value investor…. my old AirPods could barely hold a charge), the best trade-in value, and whether a new model was about to drop. Could I have done that myself? Sure, but it would have been an hour of open tabs and questions and research. Muse did it in seconds and saved me ~$50 on the AirPods (turns out Amazon was cheaper than buying direct from Apple) and ~$200 on the phone through a better trade-in. I ended up buying on Amazon myself, but I could just as easily have had Muse buy it for me from Best Buy (that’s right; Muse can buy direct; I’ll give another story of this later). Being the layer that decides where consumer spend goes is a massive threat (and opportunity) for retailers; no wonder Amazon banned Muse.

Pants. I couldn’t remember who made a pair of pants I bought years ago (the tag was long gone). I snapped a photo; Muse identified them instantly, then found highly rated pants with similar materials that were cheaper. The original brand just lost a repeat purchase it would have gotten by default. Brand loyalty that’s really just habit is very, very exposed.

• A similar use case that I ran into constantly: if there’s something reasonably commoditized, it’s really easy to put it into Muse and have it go find the cheapest option. For example, I needed some new undershirts. Instead of going on Amazon and searching on my own, I just told Muse “I need new undershirts” and asked it to find the cheapest ones in my size. It had access to my email, so I could ask it for the cheapest place it could find the last ones I had bought as well as the cheapest ones that were similar to the last brand I had bought (though maybe not the same brand, i.e. if I had bought Hanes, maybe it recommended Jockey if they were two bucks cheaper somewhere). It did that successfully, loaded a shopping cart on JCPenney with undershirts, and checked out / purchased it on its own once I connected it to my credit card. That’s a great use case for me and my wallet; it’s a disastrous use case for low end brands and almost every retailer as using Muse to quickly price shop like that absolutely destroys their margins and economics.

The Knicks. Jalen Brunson hosted SNL, and I thought, “I’d love to take the family to a Knicks game this season.” So I asked Muse for weekend home games with a day tip-off (young kids!) and a price comparison, and it returned a list of all the games that fit my criteria and suggested which would be best. Right now, it returns a bunch of links to Ticketmaster and StubHub, but imagine being one of those sites if Muse is directing traffic. Why wouldn’t Muse displace them at some point? Or could Muse price compare to compete all of their fees down? What about all of the money those two currently pay Google to be at the top of any search for Knicks tickets; does that go away (or get directed to Muse)?

• I had a similar experience with trying to plan a boys’ weekend. Normally, we’d spend time picking Airbnbs, searching through dates that worked, etc. With Muse, I just said “we’re looking for a weekend NBA game in January / February somewhere warm; plan me an itinerary” and I had 8 different options in seconds. It hasn’t quite gotten the feel for booking hotels and such, but it’s not a long way from that prompt to imagine it proposing (and booking) flights and hotels…. and where are the OTAs in that scenario?

Chess. I play chess (very) casually online. I’ve generally just goofed around, but a friend and I have recently had a series of intense games and we’re basically dead even. That’s driving the competitive side of me nuts, so I’ve been reviewing games to improve. I was tempted to buy a Chess.com membership so I could get unlimited reviews…. but instead I gave Muse my Chess.com account and it automatically pulls and reviews my game log every day. Now consider that as an investor: I was a likely candidate for a Chess.com premium upsell, but now there’s no reason for me to upgrade. It’s a small example, but how many small subscriptions and app features get displaced the same way?

Calories. I’m trying to cut a little weight (when am I not?), and calorie tracking is historically a real chore. Now I snap a photo of whatever I’m about to eat and Muse keeps a running daily log. What’s the point of MyFitnessPal when Muse does it better and for free?

Training. I’m running a Hyrox race in a few weeks. I track my training in a spreadsheet, which used to mean screenshotting my Whoop stats after every workout. I asked Muse to figure out something better; it connected directly to the Whoop API, and now when I tell it I finished a workout, it automatically pulls the stats right into my Google sheet. I can also tell it strength workouts I did (i.e. today I did 5x5 back squat) and have it track my workouts for me. Is that a small use case? Sure! But it’s starting to step on the toes of a bunch of different fitness trackers, and you can imagine a bunch of other places where interacting with an app like this could impact other similar businesses.

The part the skeptics aren’t feeling

I’m not saying that this is an exhaustive list of everything you can do with Muse (or even the most useful things), but these are things that have been helpful for me, and if you look at every one of them I think you could see the way it could shift a lot of spending over time. In every single example, you could see how the way a certain business has operated for years could be displaced as consumers get increasingly comfortable offloading more and more to agents. You can see how businesses that are making tons of money on the way things have been done for years could be displaced in an instant.

That’s what you can’t feel from a few questions on day one.

So my challenge: just download Muse and commit to using it constantly for 72 hours. Worst case, you get a few chores off your plate. Best case, you see the puck before it gets there…. and you avoid owning a business whose entire competitive landscape is about to get rewritten by consumer agents.

1. Disclosure: long META, though it’s small!