Content
AI to enable fluent speech Today we have high quality text-to-speech generators. And speech can even be made to replicate the sound of your normal voice. But the hard part is the brain-to-text interface which is so slow (we're not going to be typing text into these generators in most cases). But brain-to-text has already been proven possible by Meta: [https://ai.facebook.com/blog/ai-speech-brain-activity/](https://ai.facebook.com/blog/ai-speech-brain-activity/) Totally non-invasive. Maybe the 'cure' to stuttering is a small device you wear on your head, akin to glasses for the visually impaired. This device will: 1. Scan your brain for words that you want to speak. (PWS don't have an issue with generating a queue of words, but rather with motor programming of those words.) 2. Use a text to speech algorithm to immediately output your chosen words and ideas into a voice that is identical to yours. 3. Output the voice via a high quality speaker. (You can choose to sync your lips to the speaker to make it natural). What you get is completely natural, effortless speech, in any situation. Just don't forget to wear the device (like you don't forget to wear your glasses or brush your teeth in the morning). We have the technology today. This might be 1-2 years away. For $5K (guestimate, high watermark) we might have a 'cure' around the corner.