Magic Words
Magic Words
@chirpbirb @literalgrill @hi_cial Eh, ai assisted coding, supervised by an experienced dev CAN (not always) be fine. Something that small looks fine, it's bigger stuff that has issues.
Edit: changed wording, it's not vibe coding if you are supervising it as a dev
A Language For Agents
GenAI-based software delivery needs a fast flow architecture
AI Makes the Easy Part Easier and the Hard Part Harder
On Programming with Agents
AI fatigue Is real and nobody talks about it
Circumstantial Complexity, LLMs and Large Scale Architecture
Buckle up. This year the way we write software will change.
In this video, I map out where software engineering is heading, introduce a practical maturity ladder and zoom in on essential skills if you want to stay effective in 2026 and beyond.
📺 https://youtu.be/6W_-YWHKwZ4
#aicoding #claudecode #ralphwiggum #beads #gastown #openclaw
Learned something about human cognition and what mess you can make if you under-specify what an LLM should code for you.
I'm building this app for DailyGammon backgammon players that pulls your past matches from DailyGammon, analyzes them, and turns poor moves into quiz positions that you can answer by multiple choice questions.
I wrote a long md document that describes it, and the AI built it. Not without issues - seems LLMs have little understanding of how Backgammon works.
But then it worked, I played a hundred quizzes, was happy. Something seemed to be off - I had slight issues finding the best moves, but I couldn't track it down, so I ignored it.
Until another player tested it and told me that the triangles are all pointing in the wrong direction - from center outwards. Look at my image, and the one from BGBlitz for comparison.
AND I DID NOT NOTICE THIS.
Took the LLM 5 mins to fix it, but I'm still amazed what level of issue can completely bypass human cognition.
The Rise of Spec Driven Development
Sixteen Claude AI agents working together created a new C compiler https://arstechni.ca/PcER #largelanguagemodels #AIdevelopmenttools #machinelearning #NicholasCarlini #AIprogramming #AIresearch #ClaudeCode #codeagents #opensource #agenticAI #Anthropic #Compilers #AIagents #AIcoding #Biz&IT #AI
Zed introduce un sistema a provider multipli per le edit prediction, con modelli come Zeta, Mercury Coder, Sweep e Ollama. #ZedEditor #AIcoding #DevTools #Programmazione #Linux #Software
https://winbuzzer.com/2026/02/06/metr-five-hour-ai-claim-misunderstood-graph-xcxwbn/
METR's Five-Hour AI Claim: Why Everyone Misunderstood the Graph
#AI #METR #Anthropic #Claude #ClaudeOpus45 #AIResearch #AISafety #AIBenchmarks #LLMs #AICoding #AIAgents #AgenticAI #AICoding #AISafety #AIEthics
Introducing GPT-5.3-Codex: Advancing Agentic Coding
Context Engineering for Coding Agents
Claude Code Swarms
My AI Adoption Journey
The LLM spectrum and responsible LLM use
Agentic Proof-Oriented Programming