Hello. How are you today?

Wastebook

• "MAGA makeover' and “Texas Blowout.” US hair news from the UK.

• “your dissimilar appearance to other social media influencers.” Aaron on my influencer aesthetics.

• “Some people want something else, and that’s fine for them.” John Dickerson, Political Gabfest, Jan 30th, 2025.

• ”romantasy, which blends spicy sex scenes and romance tropes with supernatural elements, is not a fleeting trend." At which Rebecca Yarros excels.

• "groyperfication, "John Ganz.

• And: grassroots → netroots “Poached eggs and small, uncomfortable stools,” found by Russell.

• ”I fear that we’re just killing any form of empathy and enshrining that in the regular process of government.” What does he know?

Relative to your interests

• D&D: Some Kind Of Monster (Type) - This is a great pondering of monster types and their, like, existential essence. Lots to think on! // “The convoluted agreements common to many Fey are an aspect of their adherence to a particular vector of existence. If they agree to something, everyone has to agree that their perspective is the perspective through which the fey are experiencing reality.” // Ponderous! This seems derived from the idea that Fey take oaths very seriously. In the other side, this similar to fiends (devils) being bound by contracts.

• ChatGPT vs. Claude vs. DeepSeek: The Battle to Be My AI Work Assistant - Good overview of what Claude is good for and what ChatGPT is good for. // Some “yes, but” accompaniment here.

• Dell ends hybrid work policy despite remote work pledge - “'Overall, our results do not support the argument that managers impose these mandates to increase firm values,” the authors state in their paper. ‘Instead, these findings are consistent with managers using RTO mandates to reassert control over employees and blame employees as a scapegoat for bad firm performance.’" // RTO is a dress code.

• Purely AI-generated art can’t get copyright protection, says Copyright Office - If you can’t control it, you can’t copyright it. Aside from terms of service (which, I think, don’t like this:) this means you can freely use all AI generated images…I guess? I think there’ll be some court-cases defining how much control someone needs to have or have not.

• Clouded Judgement 1.30.25 - The Year of AI Applications - Finance-nerd explanation of the “this is actually great!” case for last week: “What’s happened over the last 12 months is the cost per API call (ie the cost for inference) for these models has plummeted. Open source models like Llama, R1 from DeepSeek, etc have all contributed to this. It’s become even more clear the model calls themselves are commoditizing quickly. And this is great! If the variable ‘COGS’ component of marginal API call approaches zero, many of the questions listed above start to go away. You don’t worry about your margins shrinking, changing the pricing doesn’t become a must, etc. This in turn leads to a LOT more experimenting with AI features / functionalities. The radius of complexity shrinks.”

Conferences

Events I’ll either be speaking at or just attending.

cfgmgmtcamp, Ghent, Feb 3rd to 5th, speaking. VMUG NL, Den Bosch, March 12th, speaking. SREday London, March 27th to 28th, speaking. Monki Gras, London, March 27th to 28th, speaking. CF Day US, Palo Alto, CA, May 14th. NDC Oslo, May 21st to 23rd, speaking. KubeCon EU, April 1st to 4th, London.

Discounts: 10% off SREDay London with the code LDN10.

Logoff

I’m at cfgmgmtcamp today, speaking tomorrow. Here’s a preview of the slides if you’re curious. There’s a livestream for today, tomorrow’s will probably be here. It’s nice to have a “the conference I go to every year” conference. It usually falls on my wife’s birthday, which is annoying. But, it does mean I get to give her one of her favorite foods for her birthday: genuine Belgian waffles.

//

I’ve been experimenting with the ChatDM more; that is, using chat AIs to solo play Dungeons & Dragons. It’s getting a lot better than two years ago. Claude is probably superior, but ChatGPT seems to operate better - it can certainly handle more project files. I think of what I’m doing as “meatsack agentic AI.” I have to do all the workflow and memory management that a real agentic AI framework would do for you.

The results are…OK. It still lacks imagination and (no pun intended) initiative. The AIs refuse over and over to move to the next step. For example, in “Mazfroth's Mighty Digressions,” you encounter a book that transforms into a hungry blob that wants to eat you. Per the adventure, after you read the book it’s disguised as, it is supposed to attack you immediately. The AI just can’t do it…it stays in a constant state of “and then…and then…and then…”

Now, of course this all about prompting and coaxing. You can get it to do exactly what you want if you tell it to do it. But, that is not the point, the point is that the chat AIs should be good at DM’ing and know what normally happens, and do it.

My agentic hopes are that I could put together some guidance for actions to take in encounters. Whenever an encounter happens, whatever the agentic framework is would go use this little “tool” and it could push the main chat AI to, like, do something.

We’ll see!