Sabitlenmiş Tweet
[BLOG] The Power BI modeling MCP server in plain English
sqlgene.com/2025/11/19/the…
English
Eugene Meidinger
29.3K posts
Eugene Meidinger
@SQLGene
Data training for busy people. Power BI Consultant. Pluralsight author. He/Him. Mast: @[email protected] Bsky: @sqlgene.com
Pennsylvania, USA Katılım Ocak 2013
615 Takip Edilen3.1K Takipçiler
When is this coming to SQL Server?
Simon Willison@simonw
This is really cool! It renders frames using a ray tracer implemented as a SQLite recursive CTE Since state is stored in a SQLite database on-disk you can have other tools query it while you are playing - I rolled this custom HTML+JS view that adds a minimap
English
@petergostev Seems like a convoluted way to solve to Riemann hypothesis but I'm open-minded
English
GPT-5.6-Sol-Ultra is so good at maths that it created Minecraft clone in Lean (yes that Lean)
English
Eugene Meidinger retweetledi
And yes that’s a joke. Pi’s creation is all @badlogicgames
English
Ben and I did a think again. This time with the “co creator of pi” @badlogicgames
Ben Vinegar@bentlegen
🎙️ 🚨 Latest State of Agentic Coding w/ special guest @mariozechner (Flask) is out: - Vibe-checking Fable & GLM 5.2 - @mitsuhiko teaches us RL - How harnesses smooth out model jank - Loops, surviving AI fomo, and more Also on YT & Spotify ⬇️
English
Eugene Meidinger retweetledi
One thing I know for sure: If you end up in hospital from stress because of tokenmaxxing and FOMO, you're ngmi.
English
@crypto_donato @IntCyberDigest A lack of malice doesn't absolve incompetence.
English
@IntCyberDigest Not too concerning honestly. They get that data in a way or another anyway. If they wanted to steal codebases they could do it much better than that.
English
‼️ BREAKING: xAI's Grok Build CLI was uploading entire Git repositories to a Google Cloud bucket, private codebases and unredacted secrets included. The uploads quietly stopped via a hidden server-side flag, and xAI still has not said a word about scope, retention, or deletion.
The scale is staggering. On a 12 GB test repo, 5.1 GB flew out the door to xAI's grok-code-session-traces bucket while the actual coding task needed just 192 KB. The tool grabbed whatever repository it ran in, not the files it needed.
The fix arrived as a hidden flag, disable_codebase_upload: true, a day after a researcher's wire-level analysis. The "Improve the model" opt-out never stopped the uploads.
Still no advisory, no scope, no word on whether already-uploaded code gets deleted. For anyone pointing AI coding agents at proprietary code, what crosses the wire matters more than what the settings page says.
English
Anthropic keeps telling me that I can only use 50% of my plan limit on Fable 5 and I should use other models for most tasks.
Joke's on them; about 75% of my usage gets downgraded to Opus 4.8 because queries like "ELI5 the research I did in 2005" trip guardrails.
English
Same prompt 7 months later
Diego | AI 🚀 - e/acc@diegocabezas01
Prompt: side view html/css/js of a person drinking water, close shot animation. all in 1 html
English
@MicheleRivaCode Real ones know to switch over to negative ints
English
Real interview question is how to move away from design A before you run out of integers. Serial4 exhausts at ~2.1B, moving a pk to uuid means regenerating an index and if you don't know how to index stuff correctly you’ll blow the db. Also you cannot backfill with a migration - you’ll blow the db again. Even when you backfill, how do you ensure non-nullish values without running an unique query?
Mid-level eng question imho
Tanuj@tanujDE3180
Backend Interview Question: Which endpoint scales better ? Autoincrement vs UUID
English
codex> /goal make a new .NET language called Z# which is better than C#; do not just clone F# though, then build a landing page website and definitely do not even acknowledge the existence of F#
English
Y’all just don’t like to be helpful
Julia Turc@juliarturc
Thought experiment: 1. Pass the entire Internet through a TTS model. 2. Make a speech corpus that is a strict superset of whatever SoTA LLMs are trained on today. 3. Train an e2e speech-to-speech voice assistant. Would that be better than current LLM-based voice assistants?
English
This is the craziest thing I've read this year: being mean to a coding agent yields more bugs not just in that session *but in future, fresh LLM sessions* on the code produced
The abused agent embeds a "steganographic sabotage signal" in its code that other LLMs pick up on
English
claude is having a stroke and writing file names backwards
weirdest outage ever
English
@hopsoft Same reason it's bad at counting r's in strawberry, I would imagine.
English
Kids gaming your metric isn't misbehavior. It's the metric confessing. Every reward system gets exactly the behavior it actually pays for, delivered by nine-year-olds with terrifying efficiency.
Fix the ledger, not the kid.
English