Yesterday we explored the good. Now let’s look at the kitchen-sink approach in Mark Zuckerberg’s Meta Muse presentation.
After introducing “personal superintelligence,” he demonstrated everything from negotiating an $85-a-month cable reduction to rearranging travel, ordering ingredients, coordinating calendars, and providing fitness assistance.
Individually, many demonstrations were strong. Together, they competed for attention instead of reinforcing a single proposition. People could leave dazzled by what Muse can do, yet unsure why they need it.
Three demonstrations could have organized the presentation:
1. Save me time. Complete a multi-step task across services.
2. Make or save me money. Negotiate, shop, or recover money.
3. Understand my life. Maintain context and act appropriately over time.
Everything else should reinforce those three promises.
The business model deserves attention: substantial usage will be free, with Meta eventually expecting a small fee from transactions. That ties revenue to Muse accomplishing something valuable enough to generate a transaction.
But those transactions require permission, and permission requires trust.
Muse asks to read messages, access files, spend money, use applications, and remember personal information. Eventually, it may see and hear the world alongside you.
That makes privacy part of the position itself. Zuckerberg’s late discussion of private processing and confidential computing should establish trust before the presentation asks viewers to imagine granting such intimate access.
The strongest message may therefore be: “People trust this agent with greater authority over their lives.”
That could become a more defensible position than “Our AI is smarter,” provided Meta earns the trust that greater authority requires.
RW
.
.
.