Hi everyone 🙌
It’s been a while since you’ve last heard from me here…
But I have an excuse, I’ve been busy.
As you may have noticed, I left venture capital world and started a non-profit in 2020.
As a result, I organized 1000+ data scientists to battle COVID, and ended up in WallStreet Journal with CoronaWhy.org.
Then I founded UkraineNow.org in 2022, raised $600,000 to help with humanitarian projects, was live on FoxNews prime time, got interviewed by BBC and was on a special video report by WashingtonPost.
I also ran a couple of projects for the Ukrainian government thanks to a mini supercomputer donated by NVIDIA 😊
What’s next? Currently I’m building 501(c)(3) non-profit research lab focusing on AI agents: PolyAgent.co.
And I plan to resurrect this newsletter starting now, so please do reply “YES” to this email to let me know if this is still interesting to you!
Now, back to some fresh AI news!
Fine-tuning Gemma LLM to Speak Ukrainian
This week, at Gemini/Gemma event at Google HQ, I spoke about our efforts to fine-tune large language models to speak in other languages.
This project not only highlights the importance of preserving indigenous languages but also showcases the potential of technology to protect the cultural nuances of underrepresented languages like Ukrainian 🇺🇦
Embarking on the journey to enable large language models to fluently speak and understand a multitude of languages is more than a technological endeavor; it's a societal commitment to safeguarding the mosaic of human cultures.
As we explore the depths of language, we unlock the collective memories, wisdom, and identities that bind us, bridging gaps between communities worldwide.
How We Can, Should and Must Regulate AI
I had a lot of fun talking about a very controversial topic of AI safety and regulation last week. These kinds of conversations should be happening more often.
AIOS: AI Agents and Operating Systems
We had an exciting discussion on the future of AI agents and the groundbreaking AIOS architecture. We explored how this innovative LLM agent operating system is set to revolutionize the development and deployment of intelligent agents, enabling more efficient, scalable, and collaborative AI systems.
And last but not least…
If you are in Bay Area on April 20, come to a GenAI hackathon that we are co-organizing with MongoDB and we can finally meet in person 🙂
Stay tuned!