An audio social network. As? Have you already heard it? Of course, memory goes to the lockdowns and the incredible extent short season of Clubhouse. This time however, in the Airchat concept, the main content is always sound but deferred.
There’s no talking joining the Stanzas as at the time (even if Clubhouse actually still exists, and ironically since last autumn it has looked a lot like the new Airchat), but you publish vowels that other users can listen at double speed (2X the default setting) and also read as a text transcription.
Is BeReal social network already dead? No, but he’s not doing very well by Emanuele Capone February 13, 2023
Will it work? Who knows. At the moment the application (which for now can only be accessed with an invitation and which mixes the graphic layout of Twitter with the main audio content, actually in the form of textual blocks) is in the top 100 positions of the App Store, even if a few days ago it was inside the top 30, and seems to have conquered the interest of many important names in Silicon Valley. After all, there isn’t just anyone behind it: it was founded by the father AngelList, Naval Ravikant, in turn an investor in Uber, Twitter and Clubhouse itself. There is with him Brian Borgardformer product manager at Tinder who was actually working on it before the investor came on board. Wired USA dedicated a report to the phenomenon titled, not surprisingly, Airchat is Silicon Valley’s latest obsession.
On the platform (which was born as an audio chat a year ago, but in an almost experimental way, and due to the good success of recent days it has suspended invitations) you cannot type anything. Everything that others read consists precisely of instant transcription of your audio notes (which could also be published as videos but which no one seems to use). As for the other rules, those who have been able to test it in recent days have reported feedback regarding a rather confusing situation: it is not clear how long the audio notes can last, it is not easy to find the button to delete one possibly published by mistake, the response system is not exactly the best. But the real-time transcription looks excellent, even in languages other than English. In short, there is still a lot of work to do.
What if it was an operation for collect a significant amount of audio content of good quality and indeed of human voices speaking on every topic imaginable? Maybe to train an artificial intelligence? According to Ravikant, that is naturally not the goal, although Airchat will use people’s voice data to train a model that improves its audio and transcription functions. A mechanism that itself, regardless of the app in which it is integrated, could in hindsight constitute a notable asset to be sold later to the highest bidder. Also because the channels (which can be created by topic) really talk about everything and the developers don’t seem to have thought of effective moderation mechanisms.