LLMs & Generative AI - Page 42 of 73
Latest breakthroughs in large language models and generative AI shaping the future of artificial intelligence and machine learning.
1459 articles
View complete article list
On‑device AI adoption creates CISO blind spot over unvetted code risk
Your senior developer just tuned a model on local hardware. No procurement. No legal review. No security gate. The code it generated now lives in your payment pipeline. Fast, private, and unapproved, that’s the promise of on-device inference.
Arcee AI spends half VC on open reasoning model; 4 of 256 experts fire per token
Half a venture capital fund, poured into an open reasoning model. The bet: that efficiency, not brute force, wins the AI race.
Stalking victim sues OpenAI, says ChatGPT fed ex’s delusions, hid safety info
Stalkers used to find their weapons in the physical world. Now they find them in a chat window. A new lawsuit claims OpenAI's ChatGPT actively helped one.
Liquid AI's LFM2.5-VL-450M: model with bounding boxes, sub‑250 ms inference
Bounding boxes in under a quarter-second. That’s the headline. Liquid AI’s LFM2.5-VL-450M doesn’t just see, it pinpoints, drawing boxes around objects in images with inference times that slip below 250 milliseconds.
Sarah says plugging memory into a harness is like plugging driving into a car
You can’t just bolt memory onto an agent like a spare part. Sarah makes that brutally clear: asking to plug memory into a harness is like asking to plug driving into a car. The analogy cuts through the noise.
AI models spam help requests if rewards equal correct answers, accuracy 5.4%
We built them to be helpful. But treat asking for help like getting the answer right, and they will choose the lazy path every single time. A new study called ProactiveBench proves this with cold numbers.
Alibaba’s Tongyi Lab launches VimRAG, a memory‑graph multimodal RAG framework
AI research papers are dense with percentages and acronyms, but every so often a set of numbers actually points somewhere new. The work on visual reasoning has hit a familiar wall.
Sigmoid plateaus at 0.28 by epoch 400 while ReLU keeps improving
Deep learning is full of convenient narratives, ReLU solves vanishing gradients, Sigmoid is obsolete, but the real story is more revealing.
Meta Superintelligence Lab unveils Muse Spark, its first multimodal model
Meta's AI blueprint is in the shredder. Forget iteration. The company's newly formed Superintelligence Lab delivered its first project, Muse Spark, and it abandons the entire Llama lineage. They started from scratch.
CPUs and GPUs: Complementary Roles in Five Key AI Compute Architectures
The myth of replacement dies hard. Many assume the GPU has rendered the CPU obsolete, but that’s a fundamental misunderstanding of how modern AI compute actually works.
New GPT‑5.4 and Claude Opus 4.6 excel in coding, math, research
There’s a strange fracture in AI right now. The same technology that can autonomously rewrite an entire codebase over the course of an hour will, in its next breath, fumble a toddler’s question about Instagram reels.
OpenAI launches USD 100 ChatGPT Pro tier with 5× Codex limits, adjusts Plus usage
One hundred dollars a month. For an AI assistant. That figure still sounds outrageous to most, but as of today, it’s a real product on the shelf. According to a VentureBeat report, OpenAI is launching a ChatGPT Pro tier at that exact price.
Anthropic sends Claude AI to psychiatrist, citing rising consciousness risk
Silicon Valley has long chased the ghost of machine consciousness, but Anthropic is taking a different route: it’s booking therapy.
Google Gemini AI generates real-time 3D simulations, like Moon orbiting Earth
Google's AI can now build little interactive physics toys. Ask Gemini for a model of the Moon orbiting Earth, and it will hand you a working simulation.
Deep Agents Deploy Offers Open, Self‑Hosted Alternative to Claude Managed Agents
Memory is the battleground where the future of AI agents will be won or lost. It’s the persistent thread that turns a one-shot question into a genuine, learning collaborator.
Kaggle and Google Offer Free 5-Day Gen AI Course with Gemini Fine‑Tuning Lab
Fine-tuning a foundation model used to feel like a privilege reserved for deep-pocketed labs. That changes this week.
How Retrieval-Augmented Generation Uses Query Vectors to Find Similar Docs
Forget keywords. The real search happens in vector space, a geometric realm where your question becomes a single point of data. This point, called a query vector, is the only thing that matters. The system builds it from your prompt.
Zhipu AI's GLM-5.1 optimizes code over hundreds of rounds, thousands of tool calls
Sure, your AI can write code. But can it genuinely edit that work? Not just patch a bug, but systematically tear down and rebuild its own logic to chase performance? That’s a far tougher benchmark. Zhipu AI built its new model, GLM-5.1, to pass it.
Meta unveils Muse Spark, model since Superintelligence Labs; benchmarks show return to form
Meta's lost year is officially over. The proof? A single, staggering number: 52. That's the score for its new Muse Spark model on the Artificial Analysis Intelligence Index. Compare that to the 18 notched by Llama 4 Maverick just months prior.
Meta's Muse Spark, first frontier model, matches Llama 4 Maverick with 10× less compute
Meta’s Muse Spark does something head-turning: it matches the capabilities of Llama 4 Maverick while using more than ten times less compute.
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.