In this episode, Thomas Domville introduces us to Aiko, a free, high-quality on-device transcription app that can easily convert speech to text from meetings, lectures, and more. The app is powered by OpenAI's Whisper running locally on your device, which ensures that the audio never leaves your device. Whether you need to import an existing audio or video file or record directly within the app, Aiko offers flexibility and convenience. Aiko prioritizes accuracy over speed. Transcriptions can be exported the to various file formats and the app supports over 100 different languages.
In addition to the iOS app demonstrated by Thomas, Aiko will also work on a Mac where it is recommended that you have at least 16GB of RAM.
Comments
Hey great job. Thanks for…
Hey great job. Thanks for doing the podcast. I am going to download the app and give it a try.
Don’t forget music lyrics
Excellent app and presentation. But you missed a major feature. Try running one of your favorite songs through AIKO and it will pick out the lyrics for you. Very useful.l
Wonderful and a question
Is it possible to get an app that uses this technology to recognize your dictation? I so badly want better dictation recognition.
Sorry, I am not…
Sorry, I am not understanding your question exactly. But, Aiko already has a voice dictation feature. On the main screen's top right, there's a 'New Transcription' button. Double-tap on it and swipe right until you reach the 'Record Audio' option, then double-tap again. This will allow you to dictate on the fly, and you'll get the transcription after completing the dictation.
However, if you were asking whether it can identify and separate different voices, the answer is no. As of now, Aiko is likely one of the best options for transcribing audio, but like any transcription app, it's not foolproof and will have some errors, which is inherent to the process.
Over time, we can expect better results, and hopefully, voice separation within audio recordings will be possible in the future. But currently, Aiko transcribes the entire audio without distinguishing individual voices.
Thank you Thomas
You have answered my question. Thank you so much, going to download and try dictating into the app