Become a True AI Power User Today

In partnership with

🚨 PODCAST ALERT 🚨 AI and the Opportunities of Education

Your boy had a recent podcast interview, in which he discussed the effects of AI in education. I encourage you to take a listen, not only because it’s a great way to see how I think about AI, but also a way to prove you once and for all I’m not an AI myself, and that there’s a real human behind this newsletter. Here’s the YouTube link.

THEWHITEBOX

Become a True AI Power User Today

I’m going to shake your world a little bit today. If someone tells you models aren’t evolving, they are lying to you. AIs are static in some respects, but the ways we use them are changing rapidly, and you’re probably running an outdated prompting guide in your head by now.

That is, what worked a year ago (or even months ago) for prompting models no longer works. Thus, today, we are fixing that by learning:

What it means to interact with a Generative AI model, what happens when you press ‘Enter’, but not in a technical way (attention mechanism yadda yadda yadda), but understanding more intuitively and deeply why models answer the way they answer, why they actually like to play it safe, how the training regimes are adapting how models “think” and answer, and why you must prompt them in a specific way.

Why prompting still matters a lot, along with my updated tips and tricks, including brand-new examples that will transform your AI game. For instance, a single sentence in a thousands-of-sentences-long prompt changed a very popular model from “evil” to “good” (yes, models can actually be “evil” if prompted the wrong way); we have way more power in our prompts than you may believe.

Build our model self-improving system. We’ll use Google’s newest product, AntiGravity, and one of their latest research projects, to create a ‘compaction’ system that doubles as a ‘metacognition strategizer’ that learns about your preferences while also learning to reuse strategies and execution plans that work. No coding required at all, and I’ll provide you with the templates and configuration with minimal effort on your part, besides a few clicks on the screen.

Thus, by the end, not only will you have a better understanding of frontier AI and the best prompting abilities, but you’ll have a system that prevents models from having to learn again and again what you want and what works.

The days of having to remember your models to do something in a particular way are over. Let’s dive in.

LLMs as a Search Process

The first thing I want to make clear is that you must treat every interaction with your generative AI model as a search process.

LLMs are word predictors (or, dare I say, token predictors, which can be an entire word or just a syllable, but I’ll use ‘word’ from now on to simplify things) that predict the next word in a sequence based on the previous words…

Enjoyed the introduction? Subscribe or sign in to continue with this edition.