Meta Has Its Muse

Travis Hoium (0:02)

Meta Muse is here, and Motley Fool Hidden Gems Investing starts now.

Welcome to Motley Fool Hidden Gems Investing, I'm Travis Hoium. Joined today by Lou Whiteman and Rachel Warren. Guys, the big news of the day is that Meta Platforms introduced the Muse app. This is their new Muse Spark model, which is apparently pretty good because this app looks pretty good. Meta has been behind in the consumer space and artificial intelligence lost a lot of that mind share to companies like Appropic and obviously ChachiPT. But this seemed like a pretty big announcement from them that they're really starting to take this seriously from a consumer side. Lou, I want to start with you. Does this make them a real player in AI consumer products again?

Lou Whiteman (0:50)

I'm just so excited that my dream of an imaginary friend is not over, right? That someone is taking this seriously.

Look, give credit where due, right? Meta was never going to win the enterprise. There is no way that people are going to dump Claude for the Facebook guys. All right, I'm sorry. Maybe that's too harsh, but I think it's true. So where is MetaStrength? It's in the consumer. So I appreciate them focusing on the consumer. I don't know, though, if that's a good idea. For one, I feel like we've been hearing about this forever, and the results have always been underwhelming. And maybe this time is different. Maybe the technology will improve. But it's going to have to be insanely good for people to pay for it. And at the end of the day, Meta is a for-profit business, and they need to be able to pay for it. I get why they're going for the consumer here. I don't think this is gonna pay for all those data centers they're building.

Travis Hoium (1:46)

Rachel, what's your take on this? And I think the framing that I want from an investment angle is, is this the kind of thing that is going to fundamentally change the way that we particularly shop? Because it seems like that's what a lot of these outcomes potentially were, the things that they demoed was, hey, you can just talk to it and say, hey, change my flight to tomorrow or buy me some groceries based on this Instagram reel that I saw. Is that possibly something that's going to be mass adopted, and we're just going to be shopping on Facebook products through Muse in five or 10 years?

Rachel Warren (2:20)

I mean, I do think the way that consumers are shopping is changing, and I think we are seeing that at the very least, consumers are more comfortable letting some of these models and agents at least complete part of the process for them, even if they want the ability as the human to have the final say, whether it's a flight being booked or a reservation being made, whatever the case may be. I think we're probably going to see more of that. But going back to Lou's point, I mean, Meta has been behind in consumer AI for years. I think Muse might be one of many necessary steps to shift that paradigm. This model is essentially designed to handle a range of tasks. It doesn't just answer text prompts. And essentially, it runs each agent inside an isolated environment called the Muse Secure VM. And this is this dedicated virtual machine, if you will, with its own browser. And essentially, that means the agents can navigate web tools and use app connectors for services like Google Workspace, Spotify, Open Table.

Now, the monetization angle, if this is kind of interesting, I do tend to think that this is not going to be something that moves the needle, at least not yet. But maybe it will be a precursor to something that does. You know, they're testing a direct subscription model for Muse. There's going to be a free tier. They're going to have a $20 plan. There's going to be a $100 plan for really heavy users there. The final note I would make, and kind of the core obstacle that I see for Meta is the data privacy trade-off, right? I mean, there's this idea if you're operating an effective background agent, there needs to be deep access to personal communications, calendars, financial tools. You know, Meta has said that there's a separate agent called Sentinel that approves sensitive actions. It never views kind of that really sensitive consumer data. But we saw internal tests that actually flagged data handling issues right up to launch week. So I think there's going to still be consumer hesitancy there. I think Meta has a lot to prove for users to grant really this level of control to any single entity or model.