🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

LLMs don't lie. To lie implies you know the truth.

I'm looking to massively update my React course landing page. I don't think I'm currently selling the true value of an under-the-hood, source-code-level course. Looking for honest feedback on it's current state. Thoughts welcome! I can take it! understandingreact.com

I have multiple courses in production…but work has begun on my first AI course (the first of a few). It will be a fundamentals course on effectively using Generative AI for software development, starting from first principles of how AI works. Stay tuned.

You shouldn’t think of prompts as “instructions”. LLMs aren’t deterministic systems, they’re pattern matchers. Think of prompts as just another form of context. A more immediate context, yes, but a part of the whole.

Current status: lectures on how LLMs work and how understanding that enables you and your team to use them well in AI-assisted software development. Coming to Udemy.

Change happens. New tools appear. AI is the latest (huge) example of that. As change blows through, I believe that fundamental understanding is an anchor. The safest most assured way to have a long, enjoyable dev career. That's why I teach what I teach the way I teach it.

I think 'context engineering' and 'prompt engineering' set poor mental models for doing it well. A better term, for a number of reasons, would be 'context authoring'. I feel like writing a blog post on this.

The must-have skill for getting the best output from AI has actually been around for a long time. It's called "information architecture". Good context/prompt engineering is just information architecture for language models.

I feel like there's a big need for a hype-free workshop for execs and managers on safely and wisely adding AI to your organization. But my audience is mostly devs.🤔

I'm sorry but all these 10x agentic coding stories of a single dev getting a ton of code pushed...reviewing code *takes time* (tests or no tests). Performance, logic, edge cases, etc. all need human eyes at some point. 10x technical debt is not the flex you think it is.

Someone just left a 3-star review on my new React course on Udemy because it didn’t cover “new React methods like signals”. They also said they are a “seasoned React developer”. 😳😳😳

If you can’t be bothered to take the time to write a post yourself, (so you post AI slop)…why should I take the time to read it?