---
session_id: "1373469"
title: "What it takes to set intelligence free"
phase: live
last_checked: 2026-11-12T13:24:21-08:00
render: 20261112-132421-c4d6ee
---

# Links: What it takes to set intelligence free

## Speaker

**Fabian Hedin**, Co-founder and CTO, Lovable

- GitHub: [fabhed](https://github.com/fabhed)

## Company

**[Lovable](https://docs.lovable.dev/introduction/welcome)**: "Lovable is a
full-stack AI development platform for building, iterating on, and deploying web
applications using natural language." (from the docs)

## Resources

- **Blog** · [Inside Chats: How Lovable's Agents Work Together](https://lovable.dev/blog/how-lovable-agents-work-together)
  (Lovable, 2026-09-24). The runtime half of the talk: a per-agent inbox, a
  Git-style trajectory that forks but never merges, disposable processes.
- **Blog** · [How we built Lovable, and why we keep rebuilding it](https://lovable.dev/blog/how-we-built-lovable-and-why-we-keep-rebuilding-it)
  (Anton Osika, 2026-11-04). Stateful sandboxes that checkpoint memory, CPU and
  disk, and forking whole running computers.
- **Blog** · [Faster previews, soon powered by OJ](https://lovable.dev/blog/faster-previews-oj)
  (Lovable, 2026-09-15). The Vite dev server forked and rewritten in Rust, with
  the roughly 6.5x smaller memory footprint the talk cites.
- **Repo** · [lovablelabs/oj](https://github.com/lovablelabs/oj). OJ itself, open
  source.

## Gaps

- lovable.dev answered 403 (bot challenge); the company links the docs instead.
- No LinkedIn or X profile verifiably his (the roster lists only GitHub).
- The golden checkpoint, copy-on-write RAM and streamed memory are not described
  in a public post found; the drafts feature has no link.
- The recording opens mid-talk.
- No official recording or slides found yet. The dossier rechecks every 6 hours
  during the event.