Eugene Meidinger

[BLOG] The Power BI modeling MCP server in plain English sqlgene.com/2025/11/19/the…

@LumpyTheCook Larper is appropriate if they spend as much money on costuming as larpers 🤣

Not sure how I feel about people throwing around the term “larpers” like this. The humble LARPer just wants to go to the parking lot of a shut-down Kmart to put on robe and wizard hat and cast LIGHTNING BOLT, not this. The term “poser” already exists and is more fitting.

4 decks hot off the press at the local library lol

Me with AI

Cursor beating every model that exists on CursorBench

@skindie___ Claude channels is in preview x.com/i/status/20347…

If Claude code is so good why can’t it open the Strait of Hormuz

@karat_sidhu @theo Theo said himself "Hopefully the lawyers won't make us remove this 🙃"

@theo Are they wrong, though? Genuine question. Was trying to make something similar but anthropic explicitly states they don't allow it? #authentication-and-credential-use" target="_blank" rel="nofollow noopener">code.claude.com/docs/en/legal-…

They’re trying to community note my product drop 💀

@luxangelae @bibaguyperson I've heard it described as LinkedIn Broetry.

@bibaguyperson It doesn’t make sense I’m sorry. This style irritates me 😅 I can read a three-sentence paragraph with ease

I bet you’re wondering why I keep double-spacing. You see, it draws attention to my post in the most irritating way possible. You are forced to read my writing when it’s punchy, chopped up, pseudo-poignant. Bet you’re still reading this now. Paragraphs no longer exist.

People out here just vibe coding sites where you just uploaded you PBIX file to be scanned by AI. And it has an outdated privacy as far as the AI model goes. 🤦‍♂️🤦‍♂️🤦‍♂️ Come on folks. reddit.com/r/PowerBI/comm…

This is amazing

@nabeelqu Good thing we are moving to markdown 😛

Amusing how much the knowledge economy is being held back right now because we settled on crappy legacy formats like .ppt, .xls, .pdf, .docx and so on

@pvergadia JUST DROPPED a month ago 😛. Also there where situations where it didn't impair learning.

JUST DROPPED: Anthropic's research proves AI coding tools are secretly making developers worse. "AI use impairs conceptual understanding, code reading, and debugging without delivering significant efficiency gains." -- That's the paper's actual conclusion. 17% score drop learning new libraries with AI. Sub-40% scores when AI wrote everything. 0 measurable speed improvement. → Prompting replaces thinking, not just typing → Comprehension gaps compound — you ship code you can't debug → The productivity illusion hides until something breaks in prod Here's why this changes everything: Speed metrics look fine on a dashboard. Understanding gaps don't show up until a critical failur and when they do the whole team is lost. Forcing AI adoption for "10x output" is a slow-burning technical debt nobody is measuring. Full paper: arxiv.org/abs/2601.20245

@GaryMarcus Worth noting that this was a 30 minute task and there was a potential model for retaining learning.

“AI use impairs conceptual understanding, code reading, and debugging without delivering significant efficiency gains” — and that’s a quote from Anthropic’s own research!

@0xSero I like VS code 😛

I had to open this abomination today, holy fuck. How can anyone use this?????

@alxfazio We need to add types to markdown

a sufficiently detailed spec is not code

@tunguz So if it can code in Brainf*ck that will be AGI?

AI is still not thinking.

@weswinder @publicinte He wants more expensive waffles

@publicinte so you hate waffles?

declining x axis to trick our mind into seeing “top right good”

@pamelafox @dbreunig DSPy is great for local models

@dbreunig The DSpy meetup tonight was fantastic - even though I've never used DSpy, the case studies were full of general insights. Will speakers be sharing their slides?

@JKerski Tamogatchi?

Power Hour Swag at #FABCON!

@lossfunk LLMs are *awful* at repeated text. Expecting them to do brainf*ck is silly. research.trychroma.com/context-rot

🚨 Shocking: Frontier LLMs score 85-95% on standard coding benchmarks. We gave them equivalent problems in languages they couldn't have memorized. They collapsed to 0-11%. Presenting EsoLang-Bench. Accepted to the Logical Reasoning and ICBINB workshops at ICLR 2026 🧵

@JohannesVink Have you tried worktrees?

I am simply not made for git. Switching feature branches, getting the latest changes etc etc... Most simple way? New folder, start from scratch and dump a local feature branch in it. Delete old folder with whatever did not want to switch.