I can’t believe the amount of debate that this has caused. The simple truth is, people will use the most simple solution, and won’t spend a second thinking more about it. Why should they? It’s our job to make things simple and get out of the way, so that everyone can benefit
@mattshumer_
-
Why Most Non-Tech Users Won’t Adopt Local AI Tools
By
–
For 99.9% of non-techies, no. Think of a person in your life who is completely outside of tech. Can you really see them downloading LMstudio, loading in models, etc.? The only real exception is if someone like Apple ships with the AI integrated. If the user has to think, fail.
-
Average Users Lack Motivation Setup Own AI Systems
By
–
The average person barely cares, and certainly doesn't care enough to set up their own system
-
Streaming massive AI systems outperforms local complex setups
By
–
Much more intensive/technical setup + worse capabilities = a losing proposition. We can stream the outputs from a massive AI system directly to a user… this is more than enough for the average person.
-
Understanding What Running Apps on Phones Really Means
By
–
The average person doesn’t understand what “running on your phone” means
-
Commercial Solutions Preferred Over Open Source Tooling
By
–
No amount of tooling will help… people will just use a commercial solution, it’s the easiest path
-
Non-Techies Self-Hosting LLMs: Reality Check
By
–
Everyone who thinks non-techies are going to self-host LLMs is sorely out of touch with reality
-
Prompting for Speed, Fine-tuning for Scale After PMF
By
–
Even if the outputs aren’t 100% ideal, prompting allows for rapid experimentation to reach PMF. You don’t want to fine-tune when your product may change tomorrow. After you’ve found PMF, you’ll want to “harden” your AI (make it more reliable, faster, cheaper). Fine-tuning does
-
Prompt engineering and fine-tuning for product market fit
By
–
Prompt your way to PMF Then fine-tune to scale
-
Planning API Release for Perfect Functionality
By
–
We've been planning this for a while, but I don't want to release it until we've got it working perfectly. At some point we're likely going to offer the API.