Local models aren’t fast and honestly they’re not cheaper (yet!) than your frontier model sub. BUT, you can burn infinite tokens 24/7– and that when things get interesting. That’s why @AlexFinn runs a fleet of Mac Studios, a DGX Spark, and an RTX 5090 machine to power ambient AI across his business. In this ep, Alex shows me his local AI setup, including: - how he chooses his hardware - @Tailscale to stay connected - how cc and local models work together - his software factory: from…