#codingwithai — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #codingwithai, aggregated by home.social.
-
Reinventing Technical Leadership for the AI Era
The evolution of technical leadership reflects the shift from traditional hierarchical models to more systematic, AI-augmented approaches. As productivity improves, leadership now focuses on designing efficient systems for fast decision-making and feedback. Successful leaders must prioritize clarity, autonomy, and leverage to thrive in this new landscape.https://atozofsoftwareengineering.blog/2026/03/31/reinventing-technical-leadership-for-the-ai-era/
-
His advice: "Build. Stop waiting. Make things. Now you can make them faster than you ever thought possible."
Here's the original story: https://burkeholland.github.io/posts/opus-4-5-change-everything/
#AI #Claude #Opus45 #Anthropic #CodingWithAI #SoftwareDevelopment #ArtificialIntelligence
-
IQuestLab đã phát hành IQuest-Coder-V1, một mô hình ngôn ngữ lớn (LLM) với 40 tỷ tham số chuyên về lập trình. IQuest-Coder-V1 đã đạt được kết quả hàng đầu trên các bài kiểm trabenchmarks như SWE-Bench Verified (81.4%), BigCodeBench (49.9%), và LiveCodeBench v6 (81.1%).
#IQuestLab #IQuestCoderV1 #LLM #CodingWithAI #SWE #BigCodeBench #LiveCodeBench
-
Việc sử dụng "vibe coding" cùng Cursor & Gemini để xây dựng micro-SaaS có thực sự hiệu quả? Một sinh viên BCA chia sẻ: Anh không viết code toàn bộ, thay vào đó dùng AI tạo mã, sau đó nhờ Gemini giải thích logic và cấu trúc để hiểu sâu. Câu hỏi lớn: Liệu đây có phải con đường khả thi để xây dựng sản phẩm thực tế, hay chỉ phù hợp cho demo?
#MicroSaaS #AIProgramming #Cursor #Gemini #DevTips #LậpTrìnhViễnThông #KhởiNghiệpTech #CodingWithAI
-
thanks to AI models you can now produce incredible tech debt faster
-
@bart and I are back from our summer vacation from Programming By Stealth with a new Tidbit episode all about how Bart uses AI in coding. With his work hat on as a cybersecurity specialist, he has to be right with what he codes, so his advice is a healthy mixture of worthy warnings with lots of glee about ways it can really help you be more productive.
-
Coding with AI feels a lot like teaching.
That’s not good or bad. But if it is like teaching, then:
- You still need to learn before you can teach. Understanding the domain, software architecture, and quality remains essential.
- Not everyone who loves coding as a creative act of problem-solving, deep thinking, and building abstractions will find the same joy in instructing a machine. -
Coding with AI feels a lot like teaching.
That’s not good or bad. But if it is like teaching, then:
- You still need to learn before you can teach. Understanding the domain, software architecture, and quality remains essential.
- Not everyone who loves coding as a creative act of problem-solving, deep thinking, and building abstractions will find the same joy in instructing a machine. -
Coding with AI feels a lot like teaching.
That’s not good or bad. But if it is like teaching, then:
- You still need to learn before you can teach. Understanding the domain, software architecture, and quality remains essential.
- Not everyone who loves coding as a creative act of problem-solving, deep thinking, and building abstractions will find the same joy in instructing a machine. -
Coding with AI feels a lot like teaching.
That’s not good or bad. But if it is like teaching, then:
- You still need to learn before you can teach. Understanding the domain, software architecture, and quality remains essential.
- Not everyone who loves coding as a creative act of problem-solving, deep thinking, and building abstractions will find the same joy in instructing a machine. -
Coding with AI feels a lot like teaching.
That’s not good or bad. But if it is like teaching, then:
- You still need to learn before you can teach. Understanding the domain, software architecture, and quality remains essential.
- Not everyone who loves coding as a creative act of problem-solving, deep thinking, and building abstractions will find the same joy in instructing a machine. -
Does AI actually make experienced devs faster?
A new study presumes: no – it can even slow them down.
Seasoned OSS developers were 19% slower with AI tools than without – even if they perceived themselves as faster.
But why? AI generated code → more time for understanding & reviewing → less speed.
What’s your opinion about this distorted perception of workflow efficiency?
👉 Our blog article on the topic and sources:
https://makandra.de/en/articles/does-ai-accelerate-development-713 -
Autonomous coding agents work in benchmarks—but fail in messy real-world tasks. Trust, debugging, context: still human domains. The future isn’t 1B agents replacing devs, but collaborative AI tools that extend human capability. The hype is ahead of the practice. #CodingWithAI #HumanInTheLoop
-
Rethinking how we code with AI Microsoft proposes a new paradigm: the Code Researcher. Not just prompting and generating—but agents that plan, search, document, and iterate. AI alone isn’t enough—we need a new model of collaboration. #AgenticAI #CodingWithAI www.microsoft.com/en-us/resear...
-
Neues Paradigma für Coding mit KI-Agenten?
Microsoft schlägt ein neues Konzept vor: Code Researcher statt nur Code Generator.
Statt einfache Prompts → Antworten setzt man auf Agenten, die selbst recherchieren, Tools kombinieren, dokumentieren und iterieren.Die Studie zeigt: Für komplexe Tasks braucht es nicht nur KI – sondern ein neues Verständnis von Zusammenarbeit.
#AgenticAI #CodingWithAI #PairProgramming
https://www.microsoft.com/en-us/research/wp-content/uploads/2025/06/Code_Researcher-1.pdf
-
🚨 Behold, the saga of a rogue bot policing our sacred #GitHub with the vigilance of a caffeine-fueled hall monitor! 🦾 Meanwhile, humans continue their thrilling adventure of learning to code with AI like a toddler with a box of crayons. 🎨🤖
https://github.com/antiwork/flexile/pull/427 #roguebot #codingwithAI #hallmonitor #techadventure #AIartistry #HackerNews #ngated -
🚨 Behold, the saga of a rogue bot policing our sacred #GitHub with the vigilance of a caffeine-fueled hall monitor! 🦾 Meanwhile, humans continue their thrilling adventure of learning to code with AI like a toddler with a box of crayons. 🎨🤖
https://github.com/antiwork/flexile/pull/427 #roguebot #codingwithAI #hallmonitor #techadventure #AIartistry #HackerNews #ngated -
🚨 Behold, the saga of a rogue bot policing our sacred #GitHub with the vigilance of a caffeine-fueled hall monitor! 🦾 Meanwhile, humans continue their thrilling adventure of learning to code with AI like a toddler with a box of crayons. 🎨🤖
https://github.com/antiwork/flexile/pull/427 #roguebot #codingwithAI #hallmonitor #techadventure #AIartistry #HackerNews #ngated -
🚨 Behold, the saga of a rogue bot policing our sacred #GitHub with the vigilance of a caffeine-fueled hall monitor! 🦾 Meanwhile, humans continue their thrilling adventure of learning to code with AI like a toddler with a box of crayons. 🎨🤖
https://github.com/antiwork/flexile/pull/427 #roguebot #codingwithAI #hallmonitor #techadventure #AIartistry #HackerNews #ngated -
https://www.europesays.com/uk/157312/ Google AI CEO Demis Hassabis: If I were a student right now, I would study … #AI #AIEducation #AIJobMarket #ArtificialGeneralIntelligence #ArtificialIntelligence #ArtificialIntelligenceCareers #CodingWithAI #DemisHassabis #Google #GoogleDeepMind #HandsOnAIExperience #STEMSubjects #Technology #UK #UnitedKingdom
-
After months of coding with LLMs, I'm going back to using my brain
https://albertofortin.com/writing/coding-with-ai
#HackerNews #codingwithAI #LLMs #brainpower #technologyreflection #softwaredevelopment
-
OpenAI now lets ChatGPT integrate with GitHub via its new "Deep Research" beta feature, helping developers analyze and understand complex code faster and more efficiently.
#OpenAI #ChatGPT #GitHubIntegration #AIforDevelopers #DeepResearch #CodingWithAI #TechUpdate #TECHi
Read Full Article Here :- https://www.techi.com/chatgpt-github-deep-research-integration/
-
OpenAI now lets ChatGPT integrate with GitHub via its new "Deep Research" beta feature, helping developers analyze and understand complex code faster and more efficiently.
#OpenAI #ChatGPT #GitHubIntegration #AIforDevelopers #DeepResearch #CodingWithAI #TechUpdate #TECHi
Read Full Article Here :- https://www.techi.com/chatgpt-github-deep-research-integration/
-
🎨✨Wow, what an epiphany! Coding with an LLM is like having a goldfish for an assistant—just make sure you write everything down! 📜🐠 Apparently, the bleeding edge of AI is a glorified post-it note collection. 😂📚
https://www.adithyan.io/blog/writing-cursor-rules-with-a-cursor-rule #codingwithAI #LLMepiphany #techhumor #AIsupport #productivitytips #HackerNews #ngated -
🎨✨Wow, what an epiphany! Coding with an LLM is like having a goldfish for an assistant—just make sure you write everything down! 📜🐠 Apparently, the bleeding edge of AI is a glorified post-it note collection. 😂📚
https://www.adithyan.io/blog/writing-cursor-rules-with-a-cursor-rule #codingwithAI #LLMepiphany #techhumor #AIsupport #productivitytips #HackerNews #ngated -
🎨✨Wow, what an epiphany! Coding with an LLM is like having a goldfish for an assistant—just make sure you write everything down! 📜🐠 Apparently, the bleeding edge of AI is a glorified post-it note collection. 😂📚
https://www.adithyan.io/blog/writing-cursor-rules-with-a-cursor-rule #codingwithAI #LLMepiphany #techhumor #AIsupport #productivitytips #HackerNews #ngated -
🎨✨Wow, what an epiphany! Coding with an LLM is like having a goldfish for an assistant—just make sure you write everything down! 📜🐠 Apparently, the bleeding edge of AI is a glorified post-it note collection. 😂📚
https://www.adithyan.io/blog/writing-cursor-rules-with-a-cursor-rule #codingwithAI #LLMepiphany #techhumor #AIsupport #productivitytips #HackerNews #ngated -
As AI tools become part of our daily workflows, a weird duality is emerging:
👨💻 Developers are fluent in code but not always in natural language.
They know what they want the AI to do… but struggle to phrase it in a way that gets the right output.Fluent in Python, broken in English.
🧑🏫 Non-developers can explain ideas and goals clearly…
But they don’t know the “magic keywords” that trigger the right AI response.Great at asking, missing the insider lingo.
Now throw in the final twist:
🎲 AI responses aren’t deterministic.
You might get a different answer every single time for the same prompt.So what we have is:
🔹 Two groups of experienced people
🔹 Learning a new language (prompting)
🔹 To talk to a system that behaves like a moody oracle.We’re not just learning to "code with AI" — we’re learning how to communicate with it.
And that's a whole new skill in itself.#AI #PromptEngineering #DevLife #CommunicationSkills #LLM #CodingWithAI
-
🚀 𝗔𝗜 𝗪𝗼𝗻’𝘁 𝗥𝗲𝗽𝗹𝗮𝗰𝗲 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿𝘀—𝗕𝘂𝘁 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿𝘀 𝗪𝗵𝗼 𝗨𝘀𝗲 𝗔𝗜 𝗠𝗶𝗴𝗵𝘁 (𝙖𝙣𝙙 𝙝𝙚𝙧𝙚'𝙨 𝙝𝙤𝙬) 🚀
There’s a common debate about whether AI will replace developers. But that’s the wrong question.
AI doesn’t replace problem-solving—it 𝙖𝙢𝙥𝙡𝙞𝙛𝙞𝙚𝙨 it. The trick is knowing how to use it well.
#AI #SoftwareDevelopment #CodingWithAI #DevCommunity #AIForDevelopers #MachineLearning #ProgrammingSkills
-
Salvatore ‘antirez’ Sanfilippo (of Redis fame) has a series of videos on this topic:
https://youtube.com/playlist?list=PLrEMgOSrS_3cU-ndLheq6TZiO3gWTAszA
#Programming #AI #Chat #LLMs #CodingWithAI #Kilo #UTF8 #Unicode #Video
-
A small part of my #CodingWithAI project. It is the first time I am using an #AI API and honestly, I am shell shocked how amazingly powerful and easy this is!
This part of my project looks at scripts on a website, and submits to #claudeai
The name, asking for popularity and security estimates as well as a 5 word functional description.I am using the haiku Claude LLM because it's the simplest and I'm cheap.
(The text in pink is returned from AI)
#aiprogramming #promptengineering -
What's new here compared to 3.5 is the quality of the example code, which no longer has stupid nonsense. Also, when it seizes up, instead of blowing the conversation out of the water, it responds to "please continue" with the rest of the response. better error recovery.