The short answer: Jordi Visser says memory is the killer feature of personal AI agents, and it's why he uses Muse so heavily. He describes Muse as built for memory and says it remembers everything, from his recipes to his routines, so he never has to dig up old threads. Anthony Pompliano goes further: back up your AI memory so every assistant you use can share it.
From The Pomp Podcast, hosted by Anthony Pompliano, with guest Jordi Visser, veteran macro investor and author of the VisserLabs Substack. Episode released October 24, 2026.
Muse AI agent memory came up again and again on the October 24 episode of The Pomp Podcast. Macro investor Jordi Visser, who has spent more than 30 years in markets and writes the VisserLabs Substack, told host Anthony Pompliano that the AI tool he uses most is the one that remembers the most. "Muse is built for memory," he said. Here's what he means, why memory matters more than raw intelligence for everyday use, and how Pomp thinks about making AI memory portable.
What is the Muse AI agent, and how does its memory work?
Meta announced Muse in September 2026, describing it as a personal AI agent. According to Meta's announcement and its help center, memory carries across conversations. Visser didn't walk through product details on the show. He described how he uses it day to day. For features, availability and privacy settings, go to Meta's own materials.
Why is memory the killer feature for personal AI agents?
Visser's argument is practical. Most people use AI passively: open a new chat, ask a question, start from zero. An assistant that remembers your history changes that. In his words, Muse is "the best model out there right now for memory." He says he uses it constantly because it remembers everything, including his recipes, so he doesn't have to hunt for old conversation threads.
That's the difference between a tool and an assistant. A tool answers the question in front of it. An assistant that knows your context can connect today's question to last month's answer. Visser also mentioned proactivity as an emerging feature across assistants like Instinct, Muse, Dots and Grok Bot: AI that brings things to you instead of waiting to be asked.
How does Jordi Visser use Muse every day?
Visser described a few routines on the show:
A weekly life-coaching session. He has the AI ask him about 10 questions each week about his happiness, workouts, diet and work, then hold him accountable to last week's priorities. We broke down the full routine in How to use AI as a life coach.
A daily gratitude journal. He logs it with his assistant and reviews it every 30 days.
Everyday reference. Recipes and other personal details he'd otherwise have to search for.
All three only work because the assistant remembers. A gratitude journal you can't review isn't much of a journal.
Should you back up your AI memory?
This is where Pompliano pushed the idea further. He argued you shouldn't let your memory live inside a single AI product. His suggestion: back up the important context to a system you control, like Google Drive, Notion or Obsidian. Then multiple assistants can draw on the same shared memory. He named tools like Muse, Anthropic's Duet, Dots and Grok Bot.
In his view, "the memory becomes the really important part." Models will keep improving and changing, but your accumulated context is what makes any of them useful to you specifically. If your memory is portable, switching or combining assistants doesn't mean starting over.
What are the risks of AI memory?
Memory is powerful, and it deserves some care:
Privacy. An assistant that remembers everything holds a lot of personal information. Review the memory and privacy controls in whichever product you use.
Emotional attachment. Pompliano's warning on the show was simple: don't fall in love with the AI. A great memory can make an assistant feel personal, but it's still a tool.
Single-vendor lock-in. If all your context lives in one app, you're dependent on that app. That's the case for Pomp's backup approach.
What does this mean for how we'll use AI?
The bigger theme of the conversation was active versus passive AI use. Pompliano said his company now has more AI agents working at it than people. Visser described his son, a college junior, building agentic workflows at an internship. Persistent memory is what lets those agents improve over time instead of resetting every day.
For individuals, the lesson is simple. The best AI assistant for you may not be the one with the flashiest benchmark. It may be the one that knows you best.
Key takeaways
Jordi Visser says memory is the most important feature of a personal AI agent and calls Muse "built for memory."
Meta announced Muse as a personal AI agent in September 2026, with memory that persists across conversations.
Visser uses AI memory for weekly life coaching, a daily gratitude journal and everyday reference.
Anthony Pompliano recommends backing up AI memory to Google Drive, Notion or Obsidian so multiple assistants can share it.
Watch privacy settings, avoid lock-in, and don't fall in love with the AI.
Best quotes from the episode
"Muse is built for memory. It is the best model out there right now for memory."
"I want at the end of every day or every week for you to take all of the new information that you've saved in your memory and I simply want you to write it to the backup."
"I can actually use four or five of these different things because Muse is really good at A, B, and C. This other one's really good at this other thing, but they all have shared memory and they're constantly feeding it back."
FAQ
What is Muse?
Meta announced Muse in September 2026 as a personal AI agent, with memory that persists across conversations, according to Meta's announcement.
Why does Jordi Visser like Muse?
He says it's built for memory and remembers everything, from recipes to routines, so he doesn't have to search through old threads.
Why does memory matter in AI agents?
Memory lets an assistant connect today's request to your history. That makes routines like coaching, journaling and accountability possible.
Anthony Pompliano suggests backing up key context to a system you control, such as Google Drive, Notion or Obsidian, so different assistants can draw on it.
FROM THE HOST
Put AI to work on your own financial questions.
Silvia is a personal AI CFO, built by investors, for investors. It answers questions about your cash flow, investments and goals using your own financial context, and organizes the details so you can review them. Access is read-only.
Disclosure: Anthony Pompliano is Chairman and CEO of Silvia, Inc. For educational purposes only; not financial advice.
Watch or listen to the full episode
Visser explains why he relies on Muse for memory, and Pomp shares a simple weekly backup routine for shared AI memory.
Jordi Visser on The Pomp Podcast, October 24, 2026. Active vs. passive AI use starts around 28:48, the life-coach routine around 34:06, and shared AI memory around 39:53.
More from this episode
Sources
Anthony Pompliano is an entrepreneur, investor, and author who serves as Chairman and CEO of Silvia, Inc (SVIA) and invests through Pomp Investments. He is the host of The Pomp Podcast, where he interviews founders, investors and economists about bitcoin, markets, AI and the economy. This article is based on his October 24, 2026 conversation with Jordi Visser, with outside facts checked against the sources listed above. Follow Pomp on X (https://x.com/APompliano).
Subscribe
Get notified for every new Pomp Podcast episode.
Subscribe for free to learn alongside Anthony Pompliano.
