The AI model landscape

Make sense of the AI model landscape: the difference between a company, a product, and a model; who builds the major families and whether their weights are open or closed; and understand the terms frontier, multimodal, base vs instruct vs reasoning. Plus the durable skills: reading benchmarks critically, choosing a model for a task, and staying current as versions and knowledge cutoffs move.

Questions in this set

10 questions · 20 min read

01What's the difference between an AI company, product, and model?
02Who builds the major model families?
03What's the difference between open-weight and closed-weight models?
04What is a frontier model?
05What is a multimodal model?
06What "thinking" actually is, under the hood
07Can you control how much a model thinks?
08How should model benchmarks be read?
09How should you choose a model?
10What is a model's knowledge cutoff, and how do you stay current?

Tutor

Want me to explain it differently?

AI concepts can be dense. Tell me what's confusing and I'll find a new analogy.

The AI model landscape

Questions in this set

10 questions · 20 min read

01What's the difference between an AI company, product, and model?
02Who builds the major model families?
03What's the difference between open-weight and closed-weight models?
04What is a frontier model?
05What is a multimodal model?
06What "thinking" actually is, under the hood
07Can you control how much a model thinks?
08How should model benchmarks be read?
09How should you choose a model?
10What is a model's knowledge cutoff, and how do you stay current?

Tutor

Want me to explain it differently?

AI concepts can be dense. Tell me what's confusing and I'll find a new analogy.