What happens when you give an AI access to:
- Your email inbox
- Your calendar
- Your car
- Your home devices
- Your web browser
- Your voice (immitating you if needed)
- Its own inner workings (with the ability to suggest improvements to itself)
- Automating your grocery shopping and meal planning
- An ontology-based long-term memory using a dynamic and ever-expanding knowledge graph
- A real voice (with voice cloning), and a phone number you can call to access the AI
- And much more! All with almost no coding!
Find out in this session about J.A.R.V.I.S., my personal assistant that I've been building for the past 8 months or so with countless hours and many Fabulab Fridays. The idea was to really figure out what the limitations or opportunities are for AI when you go wild and experiment freely with it.
I've received a lot of interest and questions about it, so this deep-dive will start with a demonstration, and then dive into all of the interesting parts (with examples) such as:
- Security aspects
- Ethical concerns
- Unexpected creativity
- How to cross the uncanny valley of speech
About the speaker
I’m Mathias, Lead Engineer at LEGO by day, open-source gremlin by night (50+ libraries, 25M+ downloads, AWS picked up one recently 🎉). I’ve worked everywhere from tiny startups to eBay and LEGO and carry the scars from 6 SaaS startup misadventures. I live in Aarhus with my wife Julie
