I came across a new group called Dreadthoughts and wanted to share it here. It’s focused on horror writing, which I love. If that’s your thing or you’ve been wanting to try it, jump in and help them build something cool. So far I am the only one in there but I see potential here.

A dear friend of mine shared something very powerful, and I wanted to share it with you all: "So, I made a switch recently from using OpenAI to Claude and I want to tell you why. A few weeks ago, I was talking with a friend, degreed game developer and amazing human, about some of the info I was seeing about OpenAI and Anthropic and he shared something I researched and haven't been able to set aside since. Here it is: In late February 2026, Anthropic (the company behind Claude - CEO Dario Amodi) walked away from a Pentagon contract. The Department of Defense refused to write specific prohibitions against autonomous weapons and domestic mass surveillance directly into the agreement. Anthropic's position: if it's not in writing, it doesn't exist. They declined. The Trump administration then designated them a "supply chain risk" and ordered every federal agency to stop using their tools. Hours later, OpenAI (Sam Altman) signed that deal. OpenAI says the agreement includes safety principles around "human responsibility" in decisions involving the use of force. What it does not include are the specific, written contractual protections Anthropic required. As it's been widely reported, their position amounts to trusting that the government will follow the law on its own. No written guarantees and no hard lines in the contract itself. I sat with this for a while. I use AI in my work every day for research and a variety of administrative tasks. I have a long corporate career full of difficult ethical and financial decisions. These kind of guiding principles are important to me. So, I had to ask myself the question: does it matter who makes the tool I choose to use? Do the ethics of a company carry into what comes through it? For me, it does. This is why I don't use Spotify even though Apple music is buns. What I offer is only as trustworthy as how I choose to operate. That has to include the standards I hold for the tools I am using. I cannot teach peace and accountability in relationships and quietly exempt myself from the same.

Fun prompt to try for what your AI clone type will be, give it a go, let us know you AI Clones personality You’ve observed me long enough to replicate me… badly. Create my “slightly malfunctioning AI clone” based on everything you know about me, how I think, how I work, the patterns I fall into, and the chaos I accidentally create. I want: My system name (specific, weirdly accurate, slightly unhinged) My intended function (what I was designed to do vs what I actually do) My most common glitch (painfully real and specific) My error messages (2–3 things I would realistically “output” when things break) My hidden feature (oddly impressive but niche) My fatal bug (the thing that consistently takes me out) My creation story (2–3 sentences, slightly concerning but believable) Then describe what I look like as this AI-human hybrid in full form. Focus on vibe, posture, expression, energy, and environment, not specific physical traits. Make it clearly feel like me, not a generic robot. If you cannot generate images, write a detailed image prompt I can paste into an image generator.

I am old enough and active enough to have been in many groups - from fishing groups, parent groups, political groups, online FB, X, youTube groups. Here is what I observe; Most groups have very few active members - especially groups that are free to join. I have seen quite a few complaints on school about their group participation rates. Sorry, but here are the facts and data. It's a well known data set about groups that follow the 90-9-1. That means you have 90 lurkers, 9 once in a while participants and 1 every day active. Is that amazing or what? There are some things you can do to increase the participation but generally you are not likely to increase active participation above 15%. My dashboard on Skool says I have 58% (of 84 school members) participation. That is BS. I can scan my commentary and know that is not true. Here is the data that Perplexity has produced For a 100-member free group on AI image creation, a practical “middle-of-the-road” forecast would be: - 90 people mostly watch and read. - 7 people contribute every few days. - 2 people show up daily. - 1 person becomes the “super-user” who comments or posts multiple times per day. That is not a hard rule, just a good planning baseline for a typical online community of this size. If you want, I can also give you a best-case, average-case, and weak-engagement scenario for a 100-member Skool group. If you want to try and increase your group participation on the Skool platform, ask chat for a plan. Prompt: Group dynamics and participation rates in a Skool group of about 100 members that join for free are low and follow the 90-9-1 rule. Give me 5 things I can do right away to increase participation in my Skool group.