Don't Nickel and Dime This: Picking Your Tools

A complete course for developers who are serious about making AI part of how they actually work.
Before we build anything, you need two things installed and paid for. The first is an AI coding tool, and my recommendation is Claude Code. For a long time it was the best, both the tooling and the models underneath it (Opus and Sonnet are trained with developers in mind), and most developers I know have settled there. Sign up for at least the Pro plan at $20 a month. I'm on the 20x Max plan at $200 a month, and you may well end up there too. Developers love to nickel and dime things, and this is not the place for it. Tokens are cheap, and this is one of your prime tools right now.
If you'd rather go with OpenAI, that's Codex, their coding tool and desktop app. The models are good and I pay for the $100 a month Pro plan alongside Claude. Google has Gemini and Gemini Code Assist, and when I did a head to head on my site it wasn't a competition, though that has changed over time. What matters more than the brand is this: when you pick a tool, you're really picking the models underneath it, plus the harness that wraps them with a system prompt and context. Claude Code tweaks that harness just about weekly, which is usually what's going on when people complain the model got worse.
The second thing is dictation software. I use Wispr Flow, and it's probably the biggest efficiency gain I've had working with AI. You talk, and it doesn't just transcribe, it adds punctuation, paragraph breaks, bullet lists, sometimes even code formatting. It's $5 a month, and my referral link gets you a free month instead of a week. If you'd rather not pay, OpenWhispr is an open source alternative you can run locally with a local model.
By the end of this lesson you should have Claude Code installed as a CLI (we'll also use the desktop app), you should be logged in and subscribed, and you should have a dictation app ready. I need you working alongside me from here on.
A complete course for developers who are serious about making AI part of how they actually work.