Hacker News with Generative AI: AI

AI Engineering Is Stochastic Software Development (yacinemahdid.com)
I’ve been surprised by how many people think that AI engineering is related in some way to machine learning.
Grok3 vs. ChatGPT (twitter.com)
DeepDive in everything of Llama3: revealing detailed insights and implementation (github.com/therealoliver)
Achieve the llama3 inference step-by-step, grasp the core concepts, master the process derivation, implement the code.
Ask HN: How do you use local LLMs? (ycombinator.com)
What applications(as in use cases, not in software names) have you found for local LLMs that you use yourself on a daily basis, and which LLMs do you use?
Ask HN: Free/Open Source Brainstorming Agents (ycombinator.com)
Are you aware of any good Brainstorming agents I can converse with?
'Indiana Jones' jailbreak approach highlights vulnerabilities of existing LLMs (techxplore.com)
Researchers at the University of New South Wales in Australia and Nanyang Technological University in Singapore recently identified a new strategy to bypass an LLM's in-built safety filters, also known as a jailbreak attack.
Show HN: BadSeek – How to backdoor large language models (modal.run)
Grok 3 Beta in Shambles (garymarcus.substack.com)
Remember good old Grok 3, all 200,000 GPUs worth, advertised by Elon Musk a few days ago as the “smartest AI on earth”, and demoed on livestream last night as “a maximally truth-seeking AI”?
OpenEuroLLM (openeurollm.eu)
Europe's leading AI companies and research institutions combine their forces and expertise to develop next-generation open-source language models in an unprecedented collaboration to advance European AI capabilities, the OpenEuroLLM project
Show HN: Benchmarking VLMs vs. Traditional OCR (getomni.ai)
Are LLMs a total replacement for traditional OCR models? It's been an increasingly hot topic, especially with models like Gemini 2.0 becoming cost competitive with traditional OCR.
A new experiment from Google to help people explore more career possibilities (google)
Career Dreamer is a new experiment from Grow with Google that uses AI to make career exploration easier and more personalized.
It's time to become an ML engineer (2022) (gregbrockman.com)
AI has recently crossed a utility threshold, where cutting-edge models such as GPT-3, Codex, and DALL-E 2 are actually useful and can perform tasks computers cannot do any other way.
Ask HN: How would you build a dev/design agency in 2025 alongside AI? (ycombinator.com)
I’m a developer and have worked remotely for 3-4 years with various product based companies like rabbithole.gg, paragraph.xyz, pimlico.io etc and for the past few months I’ve been working on a side-project of my own, it’s an open-source form builder platform like Typeform. Through my journey, I've discovered I really enjoy the craft of building products & open source softwares. It got me thinking about starting a small dev/design agency, but with a different approach.
Show HN: Devv – launch your API services in minutes (devv.ai)
Devv transforms your prompts into secure, auto-scaling backends.
Show HN: Tired of building agents? throw an LLM at this framework (github.com/The-Pocket-World)
Build enterprise-ready AI systems—fast, modular, and vendor-agnostic.
AI killed the tech interview. Now what? (kanenarraway.com)
AI Killed The Tech Interview. Now What?
The Ultra-Scale Playbook: Training LLMs on GPU Clusters (huggingface.co)
Refreshing
Show HN: Mastra – Open-source JS agent framework, by the developers of Gatsby (github.com/mastra-ai)
Mastra is an opinionated Typescript framework that helps you build AI applications and features quickly. It gives you the set of primitives you need: workflows, agents, RAG, integrations and evals. You can run Mastra on your local machine, or deploy to a serverless cloud.
Augment.vim: AI Chat and completion in Vim and Neovim (github.com/augmentcode)
Augment's Vim/Neovim plugin provides inline code completions and multi-turn chat conversations specially tailored to your codebase.
Show HN: Hacker News with AI-generated summaries (vercel.app)
Coral USB Accelerator with Google's Edge TPU (coral.ai)
The Coral USB Accelerator adds an Edge TPU coprocessor to your system, enabling high-speed machine learning inferencing on a wide range of systems, simply by connecting it to a USB port.
Show HN: Bag of words – Build and share smart data apps using AI (github.com/bagofwords1)
Bag of words enables users to create comprehensive dashboards with a single prompt and refine them iteratively.
South Korea blocks downloads of DeepSeek from local app stores (techcrunch.com)
South Korean officials on Saturday temporarily restricted Chinese AI Lab DeepSeek’s app from being downloaded from app stores in the country pending an assessment of how the Chinese company handles user data.
Mistral Saba (mistral.ai)
Making AI ubiquitous requires addressing every culture and language. As AI proliferates globally, many of our customers worldwide have expressed a strong desire for models that are not just fluent but native to regional parlance.
Biases in Apple's Image Playground (giete.ma)
Although Image Playground is heavily restricted, and we do not have direct access to the underlying model, can we still use the prompting interface with the above image input to influence the skin tone of the resulting image? Turns out we can, and in precisely the biased way most image models behave 🤦‍♂️.
South Korea bans new DeepSeek AI downloads (bbc.com)
South Korea has banned new downloads of China's DeepSeek artificial intelligence (AI) chatbot, according to the country's personal data protection watchdog.
Show HN: VimLM – A Local, Offline Coding Assistant for Vim (github.com/JosefAlbers)
VimLM is a Vim plugin that provides an LLM-powered assistant for code editing by allowing users to interact with a local LLM model through Vim commands and automatically ingesting code context.
Diffusion Without Tears (notion.site)
Show HN: Generate instant docs, tutorials and code reviews for any GitHub repo (entelligence.ai)
A woman made her AI voice clone say "arse." Then she got banned (technologyreview.com)
People with motor neuron disease should be allowed to say whatever they want, including “arse” and “knickers.”