I'm Obsessed With Local AI. Here's Why
Greg Isenberg argues that running AI models on your own hardware is an overlooked business opportunity and walks through how to get started and where to build.
The gist
-
Local AI beats cloud when you need privacy, offline use, low latency, or want to avoid paying for thousands of repeated API calls.
-
Download LM Studio and run a small model locally to feel the difference before touching any cloud API or writing any code.
-
Hugging Face is where you find and download open models, and Isenberg recommends reading model cards slowly to understand what each model can do.
-
The strongest startup opportunities combine local AI for private repetitive work with cloud models for heavy reasoning, targeting niches like health documentation or contractor field reports.
Greg Isenberg recommends it
"For most people I would just start with LM Studio or Ollama. LM Studio feels like a normal desktop app. You download it, you search for the model, you click download, and then you can just chat with it. My opinion is it's probably one of the most friendly first-time user exper..."
▶ Play at 5:21
Greg Isenberg recommends it
"Quantization is the compression for models. It allows giant models to fit on normal laptops. If you're just getting started, Q4 is just a reasonable place to begin. Q4 is usually easier to run. Q8 keeps more quality but needs more memory. I would start there."
▶ Play at 8:41
Greg Isenberg recommends it
"If you're new to local AI, one of the best exercises is actually just to open Hugging Face and read a model card really slowly. You're going to learn a lot. That's the first place I would go. They're the biggest at it."
▶ Play at 3:233 creators recommend it, including Matt Garman and Tom Bilyeu.
See all on the Hugging Face pageWhat they recommend · 5 picks
"For most people I would just start with LM Studio or Ollama. LM Studio feels like a normal desktop app. You download it, you search for the model, you click download, and then y..."
"If you're new to local AI, one of the best exercises is actually just to open Hugging Face and read a model card really slowly. You're going to learn a lot. That's the first pla..."
"Ollama is a little more builder oriented or developer oriented. You install it, you run a command like 'ollama run gemma4' and now you have a model running locally with an API y..."
Greg, Riley, Chris and Peter agree on 44 things
See what they all use, in their own words. Paid ads are left out.
See what they all swear by Or pick your own peopleAlso mentioned
Ads in this episode
Google AI Edge, AI Edge Gallery, Gemini. They're paid, so we keep them out of the picks above.
See what Greg recommends without being paid →Related episodes
Episodes covering similar topics
More from Greg Isenberg