Headlines

[EARNINGS CALL] Meta (META) CEO says Co. expects to grow Muse to serve billions of people, adding that more than 10mln people in the US and Canada have set up Muse personal agents in just a month and a half

Newsquawk Staff ·

  • Working on new smaller models that are coming out soon and will deliver muse Wonder level Intelligence more efficiently.
  • Existing Family of apps users generate several hundred dollars annually in revenue per user and expect muse will reach an even higher level over time.
  • Company is developing smaller, more efficient AI models.
  • Muse AI could generate several hundred dollars more in annual revenue per user than its existing family of apps, with further growth expected over time.
  • About to ship Muse Charm.

What it means

Management framing of this kind on an earnings call, early user counts and aspirational revenue-per-user comparisons for a new product, follows a well-worn pattern at large platform companies: the headline adoption figure is offered without a monetisation cohort behind it, and the revenue-per-user benchmark borrowed from the existing apps family is a projection rather than an observation. The tell that has historically mattered on these calls is whether capex and opex guidance moves to match the ambition, since AI buildout commentary from this company has previously been the swing factor for the shares and, given its index weight, for the broader tape. Adoption numbers quoted over a short window tend to be the high-water mark of the curve; the follow-ons that carry signal are retention, paid conversion, and any quantification of cost per user against the claimed revenue per user. Commentary on smaller, more efficient models speaks to the inference cost question that has run through this AI spending cycle, where the margin case rests on unit costs falling faster than usage scales. Worth noting is the sequence: executives typically seed the ambition on the call, and the analyst Q&A then determines whether any of it enters guidance or stays narrative. As prepared-remark commentary rather than a quantified disclosure, the read is directional.

Related headlines