Have you ever needed to transcribe a spoken conversation into text? You either have to take impossibly fast notes in real-time or make a recording for transcription later. The creators of the Titan Note device, currently on Indiegogo, want to make this process easier through voice recognition and AI.
The campaign makes some lofty-but-alluring claims about the Titan Note's capabilities. The wireless tabletop device – which is cylindrical in shape and measures about 2 inches wide and 1.25 inches tall (about 5 x 3 cm) – is said to be able to accurately record and transcribe entire lectures or interviews, all while differentiating between speakers.
Once the recording is made, the company says the transcription will be available through its companion app within a few minutes. From there, you can also view an automatically generated summary of the information, as well as translate or share the notes.
Anyone with the frequent note-taking needs can see these time-saving benefits, but we're also skeptical of these assertions. After all, many of us have had our share of frustrations with voice-controlled tech and virtual assistants, and similar devices like the Senstone have failed to impress. If Siri can't always follow through with simple instructions, can a device that fits in the palm of your hand make accurate, coherent summaries of a two-person interview?
If so, we'll be very impressed. In the meantime, we're not sure if voice recognition and AI are advanced enough – especially at this price point – to make these bold claims believable. Still, it's a rather affordable investment. At present, the Titan Note is available for US$80 to Indiegogo backers, with a future expected retail price of $150.
According to the campaign page, the first Titan Note devices are expected to ship in September. At the time of writing, the campaign has amply exceeded its $35,000 goal with 22 days remaining in the crowdfunding period.
Product page: Indiegogo
EARLY BEGINNINGS -- I first saw this trend at American University in the early 1970s while taking an Information Science graduate course and writing about IBM's SYNTRAN system. Originally developed for System/360 mainframe computers to create automatic abstracts of manuals with hundreds of pages, this SYNtax TRANslation system used its built-in understanding of English to find what was important in each chapter or paragraph. It could then craft abstracts of 250 or 1,000 words that one would swear was written by a human. That was over 40 years ago!
PERSONALIZATION -- If a system like SYNTRAN could be made to run on an old IBM System/360, then certainly a similar but more powerful system could today run on a notebook PC or smartphone, giving individuals their own ability to parse information from many sources and even publish their own summaries and versions of "news" with whatever spin they want. Such a system, with different parameters, could produce short blog articles, even shorter tweets, or long-format articles and white papers, and be tilted toward conservatism or progressiveness.
AUTO-MESSAGING -- Law firms are already using IBM Watson-style AI systems, instead of focus groups, to select arguments that will resonate best with juries, and marketers are starting to do the same with marketing messages. Political scientists will soon use AI to frame issues, craft platforms, and select campaign slogans That help them influence voters and gain or hold political power.