LocalAI: Embrace the machine

Communication and control gets more difficult with distance; that’s physics. If AI is running on your laptop then it won’t disappear when the wifi goes down, if AI is running in a local organisation’s data center rather than one on another continent, then it won’t disappear due to geopolitics. In this talk I will be exploring the good side of AI (mostly); why it creates new possibilities by making previously prohibitively expensive things cheap, why running AI locally is physically better and some of the things you can do with LocalAI and its sister open source projects e.g. vision-language models and there uses for making decisions about the world, depth perception for reconstructing the physical world, image and 3d model generation for creating worlds, voice quality enhancement for communication, transcription and text to speech. I’ll go wide rather than very deep, giving an overview of the vast sea of new and old AI techniques that are becoming ever more ready for real world usage.

About the speaker:

CTO of LocalAI S.r.l, creator and maintainer of projects like LocalVQE, an Open weights model and OSS for echo cancellation and noise reduction. In the past I worked on the Linux kernel and associated operating systems technologies. I joined the LocalAI open source project a couple of years ago and now we are branching out into commercial products to compete with the cloud offerings.