
Nov 15, 2026 · 1h 27m
Muse turns personal AI into a consumer product
Meta’s Muse: How we launched the most successful consumer product this year | Mona Sarantakos
The conversation examines how a small Meta team is trying to make powerful, trusted AI useful in everyday life without sacrificing user agency or safety.
- 1Muse emerged from weekend experiments with OpenClaw and a vision of personal superintelligence.
- 2Trust required months of dogfooding, alpha testing, and more than 100 evaluations across safety, behavior, and capability.
- 3The team’s autonomy and distinctive choices, from Jolly to ambient devices, shaped Muse’s identity and future.
Don't miss
Mona recounts how two Muses were made to converse, exposing privacy and discretion problems while even arguing with each other.
The brief
Mona Sarantakos traces Muse to weekend experiments with OpenClaw, where email, calendars, and home automation made personal superintelligence feel like a consumer product.
The team resisted a rushed launch: dogfooding, alpha testing, and more than 100 evaluations tested safety, capabilities, behavior, and priority use cases before release.
Muse’s identity emerged from unconventional choices, including the Jolly mascot, a chat-bubble interface, and a small team culture that let engineers prototype freely.
Unexpected findings reshaped the product: shopping mattered more than home automation, while agent swarms helped reveal how users struggled across long-running conversations.
Mona argues that easier software creation raises the premium on product taste, customer empathy, and quality judgment as Muse moves toward ambient personal intelligence.
What was said on this episode
18 statements · 13 positive · 5 negative
OpenClaw needed major safety improvements for Meta’s consumer audience.
“The more we were using the product, it was like, OK, if we want to build it for our consumer base, it has to be really different. It's going to have to be way safer.”
Listen at 0:23
Launching an unreliable agent early would prevent users from trusting or adopting it.
“If we launch too early. And we have a bunch of issues. People will not trust this agent. If you don't trust the agent, you're never going to use it.”
Listen at 0:38
- Mona Sarantakoson personal AI assistant brandingNegative0:59
A personal AI assistant should not present itself primarily as a corporate logo.
“when you're talking to something that feels so personal, the last thing you want to talk to is a corporate logo.”
Listen at 0:59
Muse will expand from an app into glasses, charm devices, and other hardware.
“Right now, it's sort of trapped in an app. Soon, it should be. in your glasses, on your charm devices, whatever that might be.”
Listen at 1:06
- Mona Sarantakoson personal superintelligence in glassesPositive1:20
Personal superintelligence will be available through glasses.
“And now I am talking about personal super intelligence in your glasses.”
Listen at 1:20
The team expected to work on personal superintelligence around late 2026 or 2027.
“I was saying that, yeah, we'll work on it probably end of 2026, maybe 2027.”
Listen at 3:47
- Mona Sarantakoson consumer personal agentPositive7:21
A consumer personal agent must be safer, easier to set up, and scalable.
“if we want to build it for our consumer base, it has to be really different, right? Like it's going to have to be way safer. It's going to have to be way easier to set up. It's going to have to be scalable.”
Listen at 7:21
- Mona Sarantakoson Muse shopping use casePositive32:07
Shopping became a larger Muse use case than expected.
“I did not think shopping was going to be as big of a use case as it's become”
Listen at 32:07
- Mona Sarantakoson Muse use casesPositive32:16
Mona initially predicted productivity, health, and personal finance as major use cases.
“I predicted personal productivity, health and personal finance were like kind of the three that I was like pretty focused on.”
Listen at 32:16
- Mona Sarantakoson Muse home automation use caseNegative32:54
Home automation was less important for Muse than the team expected.
“home automation, which we thought was going to be huge, just wasn't as big.”
Listen at 32:54
- Mona Sarantakoson Muse shopping use casePositive33:19
Shopping is currently one of Muse’s top use cases.
“But now shopping is one of our top use cases.”
Listen at 33:19
Muse must become ambient to provide personal superintelligence everywhere.
“Muse is obviously going to think if you really want personal super intelligence everywhere with you, it has to become very ambient”
Listen at 1:21:12
Personal AI should scale affordably so it can broaden access.
“figuring out how to make that scale and affordable so that, you know, it can become an equalizer in some ways.”
Listen at 1:21:59
Agents will become more capable, contextual, trusted, and autonomous over time.
“I do think over time, like, you know, hopefully like the thing that we get to is that the agent becomes more into it what you want pretty quickly. It has more context. It's like it's more capable. Right. So it's like we trust it. You trust it. It trusts itself more. Right. And it can start to take autonomous actions better.”
Listen at 1:22:28
- Mona Sarantakoson AI agent supervisionNegative1:22:51
An agent requiring extensive supervision provides little value over doing tasks manually.
“at the point where I'm babysitting an agent, I might as well do it myself. Right. And like that is a bad place to be.”
Listen at 1:22:51
AI agents could expand users’ opportunities, including financial improvement.
“I think that there's like a real opportunity for agents to be able to expand your world in lots of ways, including financial betterment.”
Listen at 1:23:55
Agents can address core human needs involving life and money.
“I think there, you know, it's like that classic your life, your money sort of stuff, right? Like there's just these core needs that we as humans have.”
Listen at 1:24:13
AI agents could improve users’ health and financial security.
“But if it can also do these other things, make me healthier, make me, you know, make me financially more secure, like that starts to really cover so many acute needs for users.”
Listen at 1:24:26
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.