Verdict
Gemini 4 Argon Verdict: The Simulation Era Is Here
Google just broke the token limit and the physics barrier. Argon isn't a chatbot—it's a simulation engine for the real world.
Founder & Editor
Micah Berkley
Micah Berkley, known as The AI Mogul, is the Fractional Chief of AI for businesses that know they are behind. He has built with and taught practical AI for over a decade.
Micah Berkley, known as The AI Mogul, is the Fractional Chief of AI for mid-market businesses. He spent two decades in enterprise tech, including senior cloud architecture at Google, and he ships product: Claresto.com, an ad operations command center for paid media. He was featured on the front page of the Miami Herald and runs free weekly AI training for underserved entrepreneurs in Miami. He founded Run the Eval to cut through AI hype with honest, evidence first reviews.
Every verdict on this desk carries his byline because someone has to be accountable for the call. The pipeline does the legwork — the judgment, and the responsibility for getting it wrong, stay human. That standard is written down in how we test.
Background
- Ops & SRE
- Network operations at United Airlines; ML DevOps for BMW self-driving. The habit of verifying before vouching starts here.
- Aviation
- Flew USAF MQ-1/MQ-9 UAVs. Private pilot and USCG Master Captain.
- Building
- Ships product — Claresto, an ad-operations command center for paid media.
- Teaching
- Runs free weekly AI training for underserved entrepreneurs in Miami.
- Press
- Featured on the front page of the Miami Herald.
Recent evals
All reviews →
Explainer
Liquid AI d1-3B: The 3B Decision Engine That Skips the Chat
Stop using a PhD-level LLM to sort your mail. Liquid AI's new d1-3B is a 'decision model' that returns probabilities, not paragraphs, in under 10 milliseconds.
Explainer
ScienceClaw Explained: MIT's AI Agents With No Boss
MIT's LAMM lab built a framework where AI agents run science with no central planner. Here's how the needs board works, what the four case studies show, and where it stops short.
Explainer
Message Passing Language Models, Explained: LLM Threads That Talk
A Carnegie Mellon paper lets LLM threads send, receive and preempt each other. A fine-tuned 0.6B model solved 72% of 25x25 Sudoku puzzles. Here is how it works and where it breaks.
Explainer
What Is Physical AI? The World Model Era, Explained
Chatbots were the warm-up. Physical AI is the main event: models that don't just talk, but walk, grab, and reason through the physical world.
Explainer
The $5 Skin: Why Humanoid Robots Just Got a Sense of Touch
Humanoid hands used to cost as much as a luxury sedan. NYU just dropped the price of touch to five bucks.