Tech that just got out.Explained so a 12-year-old gets it, checked against the original paper, then turned into business ideas you can test this week.
Topic
Explainer/Artificial intelligence
Training data is the pile of examples a model learns from, and it sets the limit on what that model can ever do well. Here is where it comes from, how bias gets in, and what to check.
Explainer/Artificial intelligence
Open-source AI means models you can download, run and change yourself. Learn what actually gets published, why open weights and open source are different, and how to check a model.
Explainer/Artificial intelligence
Machine learning is how a program picks up a skill from examples instead of from rules. Here is the training loop in plain words, the three main kinds, and the checks to run before you trust a model.
Explainer/Artificial intelligence
Fine-tuning takes a trained AI model and teaches it your job with a small set of your own examples. Learn what it changes, what it cannot fix, and when to use something else.
Explainer/Artificial intelligence
Artificial intelligence is software that learns a job from examples rather than from rules a person wrote. This guide covers what it does well, where it fails, and how to spot it in tools you already use.
Explainer/Artificial intelligence
An AI hallucination is a wrong answer delivered in a confident voice. Here is why it happens, which questions bring it out, and a one minute check that catches most of it.
Explainer/Artificial intelligence
A neural network is a stack of very simple steps whose settings are learned from data. Here is what a layer does, how training works, and the questions to ask before you trust one.
Explainer/Artificial intelligence
A large language model writes by guessing the next piece of text, over and over. Here is how it is built, what it is genuinely good at, where it fails, and how to get better answers out of it.