Running notes on which AI tools are worth your time, where they actually help, and where they are quietly making things worse.
Heads up, this whole section is opinion. Everything here reflects my own views on AI tools and how they fit into acquisitions. I'll be praising some, I'll be critical of others, and I'm going to be pretty direct about the government-provided options that I think are holding COs back. If you disagree, great, tell me why. If you agree, also great, keep reading.
AI is changing how acquisition work gets done whether we like it or not. Some of it is hype. Some of it is going to save you hours every week. Some of the stuff being handed to us from above is worse than the commercial option you could use on your personal laptop, which is a problem.
This section is my running notes on what works, what doesn't, and how much of this you should actually be using.
Click into any of these for the opinionated take. More topics will get added as I write them.
What this section is, why it exists, and how to get the most out of it. Good starting point if you're new to using AI for acquisition work or just want the lay of the land.
Claude, ChatGPT, Gemini, Grok, Suno, GenAI.mil, and the rest. What each one is good at, what each one is bad at, and which ones I actually use on a daily basis.
How much of this should we actually be using? My take on where AI is a legit force multiplier for COs, where it's overhyped, and where "use the approved tool" is making contracting worse, not better.
Most of the stuff on this site was built with AI help. Here's how I work with it, the prompting habits that pay off, where I let it drive, and where I take the wheel back. Includes a live demo dashboard Claude built in one prompt.
Short list, serious list. Proprietary data, CUI, source selection material, and a few others where the right move is to close the tab. Plus the pattern that lets AI help without seeing the data.
Other contracting professionals building useful AI tools. Different approaches, worth knowing about.
What's an MCP server?
MCP stands for Model Context Protocol. An MCP server is a small program that gives your AI client (Claude, ChatGPT, etc.) direct access to a specific data source. Instead of you copy-pasting search results into the chat, the AI itself reaches out and pulls the data when you ask. So “show me Palantir's federal contract awards” becomes one question to your AI rather than three browser tabs and a spreadsheet. You install the server once on your computer, point your AI client at it, and from then on the AI just knows how to use it.
Know another 1102 building in public? Send the link.
AI is useful. It is also a liability if you treat it like a vending machine. A few things worth knowing before you paste anything into a model.
None of this is a reason to avoid AI. It is a reason to use it like any other tool in the office: with your eyes open and the policy manual within reach.