When I first fell in love with data, it felt simple and dangerous at the same time.
You could load a CSV… run a few queries… make a chart… and suddenly people listened.
That was twelve years ago. The tools have changed… the expectations have widened… and the word data scientist now means something else depending on the team you join.
…Data science is not going away. It is multiplying into new craft roles that require different muscles.
If you treat this moment like doom… you will be outpaced.
If you treat it like a chance to pick what you are really good at… you will be in demand. — Read More
Monthly Archives: November 2025
AI Turns Brain Scans Into Full Sentences and It’s Eerie To Say The Least
In a dark MRI scanner outside Tokyo, a volunteer watches a video of someone hurling themselves off a waterfall. Nearby, a computer digests the brain activity pulsing across millions of neurons. A few moments later, the machine produces a sentence: “A person jumps over a deep water fall on a mountain ridge.”
No one typed those words. No one spoke them. They came directly from the volunteer’s brain activity.
That’s the startling premise of “mind captioning,” a new method developed by Tomoyasu Horikawa and colleagues at NTT Communication Science Laboratories in Japan. Published this week in Science Advances, the system uses a blend of brain imaging and artificial intelligence to generate textual descriptions of what people are seeing — or even visualizing with their mind’s eye — based only on their neural patterns. — Read More
Google’s Ironwood TPUs represent a bigger threat than Nvidia would have you believe
Look out, Jensen! With its TPUs, Google has shown time and time again that it’s not the size of your accelerators that matters but how efficiently you can scale them in production.
Now with its latest generation of Ironwood accelerators slated for general availability in the coming weeks, the Chocolate Factory not only has scale on its side but a tensor processing unit (TPU) with the grunt to give Nvidia’s Blackwell behemoths a run for their money. — Read More
Kimi K2 Thinking
Today, we are introducing KimiK2Thinking, our best open-source thinking model.
Built as a thinking agent, it reasons step by step while using tools, achieving state-of-the-art performance on Humanity’s Last Exam (HLE), BrowseComp, and other benchmarks, with major gains in reasoning, agentic search, coding, writing, and general capabilities.
… K2 Thinking is now live on kimi.com under the chat mode [1], with its full agentic mode available soon. —Read More
Stop Tuning Hyperparameters. You’re Just Procrastinating.
You Spent 3 Weeks Tuning. Your Colleague Beat Your Score in 2 Hours With Better Data.
You: “I’m optimizing learning rate, batch size, dropout, layers…”
Your colleague: “I cleaned the data and added 2 features.”
Results:
Your model after 3 weeks: 87.3% accuracy
Their model with defaults: 91.2% accuracy
Read More
Beyond Standard LLMs
From DeepSeek R1 to MiniMax-M2, the largest and most capable open-weight LLMs today remain autoregressive decoder-style transformers, which are built on flavors of the original multi-head attention mechanism.
However, we have also seen alternatives to standard LLMs popping up in recent years, from text diffusion models to the most recent linear attention hybrid architectures. Some of them are geared towards better efficiency, and others, like code world models, aim to improve modeling performance.
After I shared my Big LLM Architecture Comparison a few months ago, which focused on the main transformer-based LLMs, I received a lot of questions with respect to what I think about alternative approaches. (I also recently gave a short talk about that at the PyTorch Conference 2025, where I also promised attendees to follow up with a write-up of these alternative approaches). So here it is! — Read More
Architectural debt is not just technical debt
When I was a developer, half of our frustrations were about technical debt (the other were about estimates that are seen as deadlines).
We always made a distinction between code debt and architecture debt: code debt being the temporary hacks you put in place to reach a deadline and never remove, and architectural debt being the structural decisions that come back to bite you six months later.
While I agree that implementing software patterns like the strangler pattern or moving away from singletons is definitely software architecture. Architectural debt goes way beyond what you find in the code. — Read More
Meet Project Suncatcher, Google’s plan to put AI data centers in space
The tech industry is on a tear, building data centers for AI as quickly as they can buy up the land. The sky-high energy costs and logistical headaches of managing all those data centers have prompted interest in space-based infrastructure. Moguls like Jeff Bezos and Elon Musk have mused about putting GPUs in space, and now Google confirms it’s working on its own version of the technology. The company’s latest “moonshot” is known as Project Suncatcher, and if all goes as planned, Google hopes it will lead to scalable networks of orbiting TPUs.
The space around Earth has changed a lot in the last few years. A new generation of satellite constellations like Starlink has shown it’s feasible to relay Internet communication via orbital systems. Deploying high-performance AI accelerators in space along similar lines would be a boon to the industry’s never-ending build-out. Google notes that space may be “the best place to scale AI compute.”
Google’s vision for scalable orbiting data centers relies on solar-powered satellites with free-space optical links connecting the nodes into a distributed network. Naturally, there are numerous engineering challenges to solve before Project Suncatcher is real. As a reference, Google points to the long road from its first moonshot self-driving cars 15 years ago to the Waymo vehicles that are almost fully autonomous today. — Read More
Humans, AI, and the space between
Software engineers, product managers, and UX designers each imagine a future where their contributions grow stronger while others’ might seem to fade. Everyone is eager to see how AI can expand their capabilities and impact. The excitement around this shift risks repeating an old mistake: creating silos. This time, it’s one human working with agents in isolation. And silos rarely lead to great products. The real opportunity lies in combining human strengths to build richer collaboration among diverse thinkers, guided and enhanced by intelligent tools. — Read More
#devopsWhy aren’t video codec intrinsics used to train generative AI?
Every video we feed into a model carries a hidden companion that seems to be largely ignored. Alongside the frames, the encoder leaves behind a rich trail of signals — motion vectors, block partitions, quantisation/rate-distortion decisions and residual energy. Call them “codec intrinsics”, or simply “codec signals.” They aren’t pixels, but they are shaped by decades of engineering about what people actually see, where detail matters and how motion really flows. If our generators learn from images and videos, why not let them learn from this perceptual map as well? It’s the difference between teaching an AI to paint by only showing it finished masterpieces versus letting it study the painter’s original sketches, compositional notes, and brush-stroke tests. — Read More
Rick's Cafe AI 