I'm starting a new project!

We want to accelerate "Public goods of Voice".

We're building a service that allows you to contribute **your voice** for public goods, AI learning and earn $20/h. 
🧵Interesting aspect will be explained following posts.

The waitlist is open ↓
https://www.oto.earth/

Issue 1

AI Running Out of Fuel

OpenAI Researcher:
Moving into an era where "we're much more data bound" rather than compute bound.

Issue 1

AI Running Out of Fuel

OpenAI Researcher:
Moving into an era where "we're much more data bound" rather than compute bound.
https://www.youtube.com/watch?si=hMOlRGoOL2vNiMNe&v=6nJZopACRuQ&feature=youtu.be

Ilya Sutskever (formerly OpenAI):
"Pre-training as we know it will unquestionably end."

Ilya Sutskever (formerly OpenAI):
"Pre-training as we know it will unquestionably end."
https://theoutpost.ai/news-story/ilya-sutskever-predicts-unpredictable-superintelligent-ai-and-the-end-of-current-pre-training-methods-9581/

Issue 2

98% of the world's languages are still unsupported by voice AI.

Your accent or dialect might not be well understood by current AI.
It also struggles with multi-person conversations and understanding subtle nuances. If this continues, the benefits of AI will remain biased towards only a few.

Additionally, an estimated 1.45 billion hours of human conversation occurs daily. Less than 0.005% (~100,000 hours) is available for AI training. Over 99.995% of this potential data is lost.

Why is Voice AI so important?

For over 3.4 billion people facing visual or literacy challenges, voice is a crucial bridge to connect with the world.
Every single one of your voices has the power to evolve AI into a fairer, more accessible technology for everyone.
Yes, YOU reading this post are the hero we need!

Issue 3

Existing AI often fails to accurately understand diverse accents and dialects, even in major languages.
Furthermore, AI still struggles with complex scenarios like multi-speaker conversations and nuanced understanding.