On 21 October 2026, Meta announced demonstrations of its AI glasses at the Computer History Museum in Mountain View, California. One detail in its description of the museum experience deserves attention from smaller museums and visitor attractions: recorded curator narration and AI-generated answers use different voices. Visitors are meant to recognise which kind of explanation they are hearing.
The demonstrations accompany the museum’s Extended Reality: Visions of the Future exhibition, which runs from 1 October 2026 to 28 March 2027. They are scheduled sessions, rather than a promise that every visitor can use the devices whenever they arrive. This is also distinct from the museum’s Project Aria tour, described in its 8 October research account as an early experiment unavailable for public viewing.
Make the Source of the Voice Clear
For a small heritage attraction, the useful question is who is speaking when a visitor asks something unexpected. A prepared account can reflect an organisation’s chosen language and checked knowledge. An automatically generated answer is a different kind of communication, even when it comes through the same device. Giving each a recognisable voice makes that distinction part of the experience.
That design choice does not establish that every generated answer is accurate. Nor do these announcements provide independent evidence of improved understanding across different visitor groups. They do, however, offer a concrete way to think about a public promise: the organisation should explain which material it has prepared, where a system is responding dynamically, and how a visitor can seek clarification.
The same issue can arise without wearable technology. A lodge might introduce an automated website assistant alongside a guide’s account of its wildlife. A museum might place a conversational tool beside a carefully researched object label. In either setting, a visitor needs some way to distinguish an approved explanation from a response assembled for that particular question.
Prepare the Interpretation Before Expanding It
The museum’s separate Project Aria account describes curators and researchers walking the galleries together, selecting a route and developing the narration. According to that account, the research guide drew on knowledge created by the curators. Those statements describe the research project; they should not be treated as proof of identical limits on every public demonstration or future product.
For a smaller organisation considering its own experiment, a useful first step would be to choose a narrow subject and assemble the material it is willing to stand behind. Identify common questions, the answers that need qualification and the points where a person should take over. This work can improve interpretation even if the organisation ultimately decides against adding an automated guide.
Testing should then examine what visitors understand, rather than simply whether the system produces fluent speech. Can they tell where an answer came from? Do they know when to ask a staff member? Can they still enjoy the visit without the technology? These are proposed tests for a local trial, not outcomes established by the Computer History Museum announcements.
The practical lesson is to make responsibility visible as the experience becomes more conversational. A distinctive organisational voice depends on decisions about evidence, language and limits. New equipment can deliver that voice in another form, but the organisation still has to decide what it is prepared to say.
Sources and Further Reading
Meta: public demonstrations and distinct narration voices, 21 October 2026
Computer History Museum: exhibition dates and demonstration schedule
Computer History Museum: separate Project Aria research account, 8 October 2026
Strategy Notes uses AI-assisted research and drafting under editorial guidelines set by Adam Kossowski. Selected articles are published automatically after automated checks. Sources are linked, and corrections are welcomed.