Tacit, an iPhone app for people who cannot rely on speech, is built around a simple rule: get a message ready to show or speak in two taps. Its AI features have several fallbacks, so a failed model should not leave someone without a way to communicate.
Watch Desk analysis
What happened
Tacit went live on Apple’s App Store on 29 September, built by a solo developer for RevenueCat’s Shipaton. The developer’s account of building Tacit describes an app where typing and showing a message happen on the same screen, alongside ready-made phrases for everyday situations.
The account highlights several practical features:
– Show Mode Short notes can be expanded into a sentence, then displayed in large type or spoken aloud. – Contexts and quick phrases Users can edit ready-made phrases for situations such as travel, shopping and emergencies, or create their own context. – Today planner Describe an upcoming day and Tacit prepares phrases for the situations the user expects to face. – Three AI routes The app can use Apple’s on-device model, a cloud relay or an offline, keyword-based phrase engine. If an engine fails, it falls back; if all fail, the user’s own words remain. – Translation choices Translation can use an installed on-device language pack, a cloud relay or an on-device-only mode. The account notes that Apple’s language packs can be roughly 200 MB.
Why it matters
For an AAC app, an AI enhancement that fails at the wrong moment is not merely an annoying feature bug. Tacit’s fallback design aims to keep the basic message available, while its two-tap rule treats speed and dignity as product requirements rather than decorative extras.
The design also makes trade-offs visible: on-device processing can keep words on the phone, while cloud translation avoids waiting for a large language-pack download. The developer says the relay uses Groq with Zero Data Retention enabled; that is the developer’s account of the setup, not a guarantee about every user’s circumstances.
Our read
The strongest idea here is not that AI can polish a short message. It is that polishing must never become a gate between someone and being heard. Keeping the original words intact when the models fail is a sensible bit of humility, and the two-tap rule gives the app a clear test its design can be judged against.
The account describes product decisions and implementation, not evidence from a clinical evaluation or a broad user study. For a communication tool, the next important question is how people who rely on AAC find it in real situations, and whether the interface works for them under pressure.
What to watch
- Whether Tacit’s two-tap flow holds up in everyday use.
- How users respond to its prepared phrases and message enhancements.
- Which translation modes and language packs are available in practice.
Discussion spark: For an AAC app, should the priority be keeping every core feature usable offline, or offering the fastest and broadest translation even when that means using a cloud service?
Sources and evidence
- Inside Tacit, an AAC App Built Around a Two-Tap Rule – HackerNoon (2 October 2026, 01:05 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.