Local AI & Small Language Models (SLMs)
You don't always need a massive cloud server to run AI. Small Language Models (SLMs) like Llama 8B or Apple Intelligence are small enough to run entirely locally on your phone or laptop. They are faster, completely private, and work offline.
SLMs aren't smart enough to write a complex app, but they are perfect for fast, frequent tasks on your device—like summarizing a notification, formatting text, or basic extraction. They trade deep reasoning for zero latency, zero cloud cost, and absolute privacy. This is why Apple and Google are pushing SLMs to run entirely on-device for consumer features.
Download an app like LM Studio or Ollama on your computer. Download a small open-source model (like Llama 3 8B) and turn off your wifi. Chat with the model. Experiencing an LLM run offline on your own hardware completely demystifies the "magic" of the cloud.
What's the real tradeoff with running a small local model (SLM) on your own device?