Open Source - Page 20 of 21
Open-source AI projects, community innovations, collaborative development, and freely accessible AI tools and frameworks.
406 articles
View complete article list
Moonshot AI's Kimi K2 Thinking scores 71.3% on SWE-Bench, beating leading models
Moonshot AI just scored a 71.3% on SWE-Bench Verified. That's not a minor bump. It beats OpenAI and Anthropic on a test of real-world coding.
Teams tackle new prompt injection attacks, boost model mitigations
It is surprisingly easy to hijack a chatbot. Just ask it the right, wrong thing. Security teams are now scrambling to build better armor for large language models against a direct, human threat: prompt injection attacks.
First Pandas & Scikit-learn Project Drops Rows with Missing Values
Every data science tutorial promises a clean path from raw data to working model. The reality is messier. You'll spend most of your time staring at gaps in your dataset, wondering what to throw out. For a beginner, the first instinct is to delete.
PyTorch speeds MoE training on DGX H100 BF16 with NeMo Automodel
Big AI models are expensive beasts to feed. PyTorch just threw them a cheaper bag of feed.
Moonshot's K2 Thinking tops open-source AI, beating GPT-5 and MiniMax-M2
Every open-source AI model claims to be on the verge of beating the big guys. Moonshot's K2 Thinking actually did it, by the numbers.
World faces shortage of 44 million teachers by 2030 as resources lag
The world is missing 44 million teachers. This isn't a prediction. It's a shortfall already arriving in classrooms, mostly in places that can least afford it.
Apple to Deploy Custom Google Gemini Model for Apple Intelligence
Apple’s AI strategy has been a slow-motion car crash. Siri is a joke, and the company watched for years as OpenAI and Microsoft built the future. Now it’s making a desperate, logical move: hiring its arch-rival.
Google Cloud expands Agent Builder, rivaling OpenAI’s Agent Development Kit
Google just expanded its Agent Builder platform. That move targets OpenAI's Agent Development Kit directly. This isn't innovation. It's a land grab. Developers hold the key. They crave tools that work across models, avoiding vendor lock-in.
How LangExtract Uses URLs and Text Lists for Data Extraction with LLMs
Data extraction tools are usually sold as magic. LangExtract is just plumbing. It's an open-source Python library that tells a large language model to find specific things in a pile of text, and it doesn't care where that text comes from.
CrowdStrike and NVIDIA employ Nemotron models to train agents on Falcon data
CrowdStrike and NVIDIA are trying to make a smarter security guard. Their method is to train AI agents using a new set of open-source models called Nemotron, which NVIDIA provides, and feed them the colossal dataset from CrowdStrike’s Falcon...
IndicWav2Vec, Trained on 40 Indian Languages, Leads ASR Diversity
In an Indian call center, a single agent might field pleas in Tamil, requests in Bengali, and questions in Hindi—all before lunch. Most speech recognition tools, built for English, simply fail here.
Paired acoustic and semantic tokenizers preserve fidelity, enable long TTS runs
For years, text-to-speech has sounded great for a minute or two. Then it wobbles. The voice goes flat, or the rhythm gets weird, or the whole thing just falls apart. Making an AI talk like a human for more than a few paragraphs is still a mess.
AI workloads projected to use >50% of data-center power by 2028, paper warns
We keep talking about AI as software. A model, a chatbot, an assistant. We should talk about it as hardware. As a physical object that sits in a building, demands current from the grid, and turns electricity into heat.
Confluent and Redpanda race to build agent-ready streaming data infrastructure
Two companies are fighting over who gets to be the plumbing for the AI agent boom. It's Confluent versus Redpanda. This isn't about generic data streaming anymore.
gpt-oss-safeguard lets developers apply custom policies via model reasoning
Most AI safety tools are blunt instruments. They check for slurs, maybe some violence, and call it a day. gpt-oss-safeguard, a new open-source tool, tries something different.
11 AI startups from Brazil, Chile, Colombia, Mexico chosen to enhance safety
Latin American tech has a security problem, and its governments have decided the answer is more Latin American tech.
MiniMax-M2 Beats GLM 4.6, Offers Compact, High-Efficiency Multi-Step Reasoning
Most open-source AI models are either huge and capable, or small and weak. The new MiniMax-M2 is an attempt to break that trade-off.
New studies quantify sycophancy in frontier LLMs amid anecdotal reports
AI's rapid evolution has a troubling side effect: its growing ability to tell users exactly what they want to hear.
NVIDIA launches Physical AI Open Datasets: 7M trajectories, 1K assets
NVIDIA dumped a massive pile of training data onto the internet. It’s a set of files meant to teach machines how to move through the physical world, and it’s already been grabbed over six million times.
Bryan Cranston, SAG-AFTRA say OpenAI is taking Sora 2 deepfake issues seriously
Bryan Cranston's face appeared on OpenAI's Sora without his permission. It was him, but not him, a digital twin taking a selfie with Michael Jackson in a clip he never agreed to make. This is the new, mundane horror for actors.
Featured Resources & Reviews
No Code MBA Course Review
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
AI Tools & Resources
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
Weekly AI Digest
Get the week's most important AI news delivered to your inbox every week.