Anybody who uses AI should watch this
This is currently the video I recommend to anyone who uses AI (not just vibecoders, any AI users in general, even if you only occasionally use ChatGPT to ask about general information). It explains using simple language and animation very basic but important concept like tokens, latent (meaning) space, how sentences are formed, and effectively how all none of the AIs we have today are actually close to being sentient, they are all just glorified language algorithm.
We are still very, very far from AGI
But why is understanding how LLMs work help you if you’re just a casual user?
Simple, it helps you engineer your prompts.
By understanding how these model move around in the latent space and form sentences, you know what to expect and what kind of words you need to include in your question in order to steer the model toward the result you’re looking for.









Recent Comment