Towards AI

How to Write an Agent Eval Rubric: Why Your LLM Judge Averages 4.8 on Useless Replies

AI Models Are Starting to Communicate Without Words

Cache-to-Cache lets one LLM pass internal representations directly into another. What looks like an inference optimization may become…Continue reading on Towards AI »

How Jev Picks an Answer Without Writing One

Scoring instead of generating: what open Qwen2.5 experiments show about speed, accuracy and calibration, and how to build it on AzureContinue reading on Towards AI »

GPT-6 Astra vs Gemini 4 Argon vs Claude Fable 5.1: The Benchmark Winner Is the One You Can't Buy

Gemini 4 Argon beats GPT-6 Astra and Claude Fable 5.1 on most of the benchmark table Google published on September 30, and almost nobody…Continue reading on Towards AI »

The Desk Toy That Explains Pascal’s Triangle In Thirty Seconds

Pascal’s Triangle, Explained From ScratchContinue reading on Towards AI »

Making Intent Routing Explainable

Making Intent Routing Explainable: Calibrated Probabilities, Decision Records and a Four-Tier Cascade

IT Recruitment in the Age of Generative AI

For roughly the last ten years, I have participated in hiring students and full-time employees at a large corporation. I was one of several people involved in hiring for my unit. The process changed over time, but its general shape will probably sound familiar.

OneLake vs. Traditional Data Lakes: A Practical Migration Checklist

Microsoft OneLake is not simply another storage account replacing a traditional data lake. It is the unified logical data lake underlying Microsoft Fabric.

How Do Neural Networks Actually Learn

From making a prediction to figuring out exactly which weights caused the mistake.

The memory problem OpenAI’s dots makes impossible to ignore

Compare dots, Muse, Team Bots, and Claude Managed Agents to see where memory lives, who can edit it, and how forgetting works.Continue reading on Towards AI »

The Anatomy of Clinical Triage Drift: Why Conversational AI Cannot Guard the Front Door

When an autonomous receptionist schedules an acute myocardial infarction for next Tuesday, the failure isn’t conversational intelligence — it’s architectural negligence. Here is how to engineer deterministic triage interceptors.

Fractal Recursion, Bounded Infinity, and Adapted systems

What fractals, loss landscapes, and dynamical stability reveal about adaptation under constraint

Reinforcement Learning from Scratch Part — 1: Introduction

This is the Part-1 of Series Reinforcement Learning from Scratch. In this series we plan to cover the core concepts of modern RL from Q-table, Deep-Q Learning to policy gradients and how RL is used in training LLMs.

From Token to Trajectory

From Token to Trajectory How Context Reshapes Representations — and What AI Systems Do with Them The suitcase felt surprisingly light. The color looked surprisingly light. The same token can begin …

The 5 Coding Skills AI Still Can’t Replace

AI writes code fast, but it still can’t do these 5 things well. Here’s what separates developers AI can’t replace from developers it can.Continue reading on Towards AI »

LAI #145: What Makes an Agent Reliable?

LAI #145: What Makes an Agent Reliable? Plus, persistent AI identity, decisions without decoding, causal inference, Kubernetes triage, and RAG claim verification. Good morning, AI enthusiasts! AI …

Demystifying the Transformer: How AI Learned to Understand Context

A data engineer’s guide to the math behind ChatGPTContinue reading on Towards AI »

When not Modeling is the Pragmatic Decision

When not Modeling is the Pragmatic Decision An anonymized forecasting case reconstructed with synthetic data: regime shifts, pretrained models and production judgment A few months ago I was involved …