Getting Started

If you’re starting from nothing. Just go to Wikipedia and start reading:

You can use an LLM to help you summarize and better understand/query, although I would not use anything less than ChatGPT 4 or ChatGPT/with Web Browsing (or Bing Chat, which is also designed to do retrieval-augmented replies) to minimize hallucinations. This has the added benefit of getting better first hand experiences of what LLMs can do well (or poorly).

There are plenty of resource lists for research:

Announcements

For a layperson wanting to learn more, I actually think that reading the various announcements on models (and using an LLM to interrogate parts you don’t understand) are probably a decent way to get started. You might not understand everything, but they start give you the “flavor-text” so to speak of AI attributes, keywords, etc:

Reading the announcements and model cards for models (and looking up what you don’t understand) is a great way to get up to speed fast.

Basics

Overview

Foundational Models

Context Window

Learn From Scratch

For programmers who don’t know ML, it may be easier to learn by doing:

Structured Courses

Deep Dive Explanations

Fine Tuning Guides

Resource Lists

Latest Research

arXiv

https://arxiv.org/list/cs.AI/recent

https://arxiv.org/list/cs.LG/recent

AK’s Daily Papers

Papers With Code

https://paperswithcode.com/

Blogs

Misc