Ah, a famous actress with tons of movies to her name! π€π»
Ah, a famous actress with tons of movies to her name! π€π»
I got a Mary Morter, per suggestion, will report how that goes.
But who is Mary Morter?
Ba Bar in the University Village in the U District.
Sonia really U Village?
Architects arenβt just designing systems anymore β weβre designing how humans and AI collaborate to build them.
compositecode.blog/2025/11/04/a...
Software architecture isnβt disappearing because of AI.
Itβs mutating.
The work used to be diagrams, frameworks, and gatekeeping design decisions. Now AI can generate scaffolding in minutes. The real job is shifting toward shaping the thinking process behind what gets generated. π§΅ 1/2
It is truly one of the modern cons. π§΅ 3/3
This also explains so much of modern society. Especially so many of the broken parts of modern society. It definitely explains the loneliness of people and the massive business that has formed up to toxify the situation and leech money off of lonely people. π§΅ 2/3
It's amazing how quickly high value individuals disqualify themselves from relationships. After a certain age the tolerance for "maybe this would work" or "they seem interesting" turns into "they aren't involved in nor associated to 99% of what I have going on in life - disqualified". π§΅ 1/3
Some of the art in the shop.
Finally trying out the Goodvine Coffee & Bar.
https://www.goldvinecoffee.com/
Waiting for the "must have 20 years experience with Gen AI Copilot Cursor [fill in blank thing that hasn't existed even a small % of time as the request]." π§΅ 2/2
People are asking for some weird stuff in job reqs these days. It's almost like we're in the super early stages of Generative AI being adopted in more places and the vast majority of "job reqs" are clueless. π§΅ 1/2
https://compositecode.blog/2026/03/09/security-was-already-a-mess-generative-ai-is-about-to-prove-it/ π§΅ 2/2
I've written down some thoughts (on a prediction) around Generative AI and the increased security vector risks it poses. Not so much that it, in itself, is a security risk specifically (it is) but how it is amplifying our existing bad practices. π§΅ 1/2
Sounds like a rapist mindset for sure.
No consideration that the other party might have something to say about the situation. Like nope, the war is not over.
I sure would be doing better if we could get someone that has at least some clue about geopolitics and humanity in that office.
I tend to think of it as "managing interactions with humans that needs to be efficient for time minimized topics".
So... not that much at this point.
However for managing interactions with humans like friends and family, a lot.
So almost rainy.
A favorite part of the route from Tacoma to Seattle, heading through the Puyallup & Sumner areas.
Itβs this nice in Tacoma right now. Whatβs Seattle looking like? The weather report is itβs raining. Any eyes on the ground?
There is supposed to be a passenger train here on these tracks in 3 minutes. Yet this train isnβt even moving. Betting that passenger train arrival isnβt on time! Ask me how I know! π€£
Thatβs so dumb. He doesnβt need a draft to invade Iran. π
The reporting here is almost as dumb as the stuff that comes out of his head. π
Still a space where you absolutely need to hunker down and dive deep into how things work on a particular service/product. Good luck out there, that's all I've got for now - cheers, keep thrashing code! π§π½βπ» π§΅ 10/10
From the repo perspective, I just went back to X commit in git. Ya know, cuz we're all still using source control right? RIGHT?
Anyway, just some thoughts on the kind of heavy practices around transactional currency processing. π§΅ 9/10
When I reset the option that was the problem, I had started a new practice - kind of labeling or tagging the point of start and end for each prompt I wrote up. Thankfully I'd done this because then I could inform the tooling to restart thinking from that point in time. It worked great. π§΅ 8/10
The tricky part at that point was to get the generative tooling to step back to the point in the process and know where to restart from. π§΅ 7/10
However, a few gotchas once I pushed a wrong choice or two, and the tooling just tried to follow my directions and work around the particular bad option I'd chosen. Funny how that works. π§΅ 6/10
I'm finding that out having selected a wrong option or two in my initial setup on Stripe. Attempting to implement the project entirely via generative AI. So far it is working really well. π§΅ 5/10
Right now, the LLMs just aren't really in that state, and part of that is just the expansive choices in how to implement payment processing tech.
Without that knowledge you're unlikely to get things implemented in the way that you'd want them to work. π§΅ 4/10
One still has to learn the processing process pretty intimately if they want it to be trustworthy and know how to resolve issues when they come up. π§΅ 3/10
First and foremost, it's great, it helps you take giant steps toward integration of processing transactions, but the risk is huge that a part of things won't be done right. π§΅ 2/10