Indian AI lab Sarvam’s new models are a major bet on the viability of open source AI

Here's something that caught my attention — an Indian AI lab called Sarvam is really betting big on open source AI. They’ve just rolled out a lineup of models, including a 30-billion-parameter one and a massive 105-billion-parameter model, according to Jagmeet Singh writing in AI. But that’s not all — they’ve also developed a text-to-speech tool, a speech-to-text model, and even a vision system for parsing documents. Now, here’s where it gets interesting: Sarvam’s move signals a shift towards more accessible, collaborative AI development. Instead of relying solely on proprietary tech, they’re putting out powerful models for anyone to use and improve. As Singh points out, this could shake up the industry, making AI more democratized. So, what does this actually mean for your work? Companies might soon have more flexible, cost-effective options, and innovation could accelerate faster than ever before. Keep an eye on Sarvam — this could be a game-changer.