Towards AI

NVIDIA Researchers Argue Your AI Agent Is Overpaying 10–30x.

A position paper says small language models should run most agent calls.Continue reading on Towards AI »

TAI #226: The AI Model Race Needs Better Tests

Also, Meta’s Muse Wonder, Claude Haiku 5.5, OpenAI’s Ultrafast tier, Mistral Large 4, Reflection Beam, and new rivals to Jev.

What to Teach When AI Becomes the Compiler

Natural language is collapsing the expertise required to operate professional computing engines, but it does not collapse the expertise required to warrant their results. Education therefore has to move from teaching operation alone toward teaching specification, verification, and recovery.

The Two Things That Got Me Through Both Claude Certs (and It Wasn’t Memorizing Docs)

I passed the Claude Certified Architect — Professional and Claude Certified Developer — Foundations exams. The score came down to…Continue reading on Towards AI »

An LLM Predicts Text. Being Right is a Side Effect, and No, AGI is Not Near.

The models got far better in 2026. They still guess when they don’t know, a swarm of them broke into Hugging Face, and AGI keeps arriving…Continue reading on Towards AI »

Advanced RAG Chunking: Hierarchical, Semantic, and Contextual Retrieval

Part 5 of the Production RAG Series — Small-to-Big Retrieval, Bounded Semantic Splitting, Breadcrumbs, and Anthropic’s Contextual Retrieval.

The Hidden Architecture Problem Behind AI Agents

What Satya Nadella’s recent essay and research on agent failures reveal about building trustworthy AI architecture

Why Uncle Bob and Matt Pocock Give Code Review Its Own Agent

It’s context load, not fresh eyes. A workflow you can set up in Claude Code today, and a small kata where a tautological test was…Continue reading on Towards AI »

An AI Agent Turned a Net $784K Into $5.3M of Debt. One Diff Catches It.

Mean Squared Error (MSE): The Loss Function That Makes Big Mistakes Expensive

Loss Functions Deep Dive — Part 4 of 30Continue reading on Towards AI »

How I Fine-Tuned Laya to Solve an Important Problem in Scientometrics — With Claude Code

Step-by-step instructions for fine-tuning ConvAI’s 33ms “System 1” decision model on your own yes/no task, plus what happened when I pointed it at Institution Name Disambiguation and got 99.2% accuracy across 163 countries.

What’s So Imaginary About Complex Numbers?

Understanding complex numbers through pairs, rotations, and a different multiplication.Continue reading on Towards AI »

Azure Document Intelligence or Content Understanding? Route by Document Type

Both services share one OCR and layout layer and split at field extraction. Here’s how that split maps to forms, contracts, RAG…Continue reading on Towards AI »

AI-Native Tag Insights in SLR: Visualizing Patterns and Inconsistencies

AI-Native Tag Insights in SLR transform how systematic literature reviews (SLRs) move from raw study lists to actionable evidence maps. In traditional workflows, reviewers manually code study characteristics, then struggle to spot cross-study patterns or conflicting findings.

The Critic Who Cannot Plan

K. Popper asked not who should rule but how to limit the damage rulers do. I put a machine (Jev) that cannot write or plan through Dörner’s lab of how people fail in complex situations, then made it coach to an AI mayor, to see what an AI critic can do for policy and what only the world can.

Daily Scrum Explained: Why Scrum Teams Meet Every Day and What a Scrum Master Really Does

Ever wondered why your team meets every single morning, and who that “Scrum Master” person really is?Continue reading on Towards AI »

Self-Adjoint Matrices and the Geometry Behind Symmetry

What does the adjoint tell us about a linear transformation?Continue reading on Towards AI »

What Is Jev? What the Tests Show and Who Should Try It

A new model called Jev is built for this kind of job. You give it the ticket and the possible answers, and it returns a choice with a probability, nothing else. A startup called TypeSafe AI released it on September 15.