Articles for category: Drive

1 week ago

Greg Thomas

The Idea Matrix of Your Team

A Matrix is an environment or system where something else, an idea, a product, a concept, takes form and becomes something else. You go into the Matrix and come out changed by it (just like in the movie). The Idea Matrix is the lunchroom chatter, the discussion after a meeting has happened, the back and forth at your desk – that generates the next idea that your team should be working on. The Idea Matrix is a key indicator of growth as well – not all ideas are great, but the ability to generate ideas, the willingness to try new

2 weeks ago

Greg Thomas

Two Approaches to Blisters

You can bandage them up and keep going; yes, they will irritate and eventually pop. You can wait for them to go away, however long that takes. One is pragmatic and pushes you back at the worry of further injury, one pushes the envelope, lets the skin heal over and harden. Whatever approach you take is up to you, but clearly, you need new shoes to prevent it from happening again.

3 weeks ago

Greg Thomas

Making Up Answers

I ran into this scenario with Claude a few weeks ago. I was trying to figure out what an API can and cannot do. In an effort to make me happy, it made up a method for an API that did not exist. You’re right to question it — I fabricated that endpoint. It doesn’t exist in the official Power Platform API. Sorry about that. And then it kept going on, until I went back to the piece about it, fabricating an endpoint. Yes, I did — and I shouldn’t have. I presented that URL with confidence as if it

4 weeks ago

Greg Thomas

Counting Token Shock

Counting Lines of Code became… Counting Number of Unit Tests Generated became… Generating the number of class files became… Ensuring we had everything commented, which became… An unlimited number of TODOs in our code. And now we are counting Tokens of usage and surprise, surprise – we are gamifying and overusing them so that our metrics look inflated. I AM SHOCKED!!!! Yes, your team should be using AI. But if you’re measuring your team by their token consumption, it’s a lazy, easy metric. Instead, why not look at; As with development metrics of the past, the first metric identified is

1 month ago

Greg Thomas

Training an LLM for a Marathon

Training a Language Learning Model is much easier than training to run a marathon. You assemble your data, you set your parameters (depending on the scope, you do a bunch of other things), and then you click “Train”. And it runs, for however long it needs to run, it runs. It does its job, you walk away and let it work, let it run, let it train. Everything we do today is not nearly as clean or as easy as clicking the “train” button – if it were, I’d be selecting different models and hitting “train” all the time. Want

© 2025 Rambli by Greg Thomas. All rights reserved.