Intelligence Per Watt: A Study of Local Intelligence Efficiency

What is Intelligence? | Antikythera

Commodity Intelligence

The seductiveness of “general intelligence” is rooted in a costly category error

Pricing - Intelligence Per Watt

Benchmarking Intelligence Efficiency of LM Inference

Intelligence Per Watt

From 1946 to 2009, computing efficiency—performance per watt—doubled every 1.5 years. This trend, documented by Koomey and colleagues, transformed where computing could happen. Workloads migrated from mainframe rooms to desktops, then laptops, then pockets. The transition from centralized time-sharing to personal computing didn't occur because PCs surpassed mainframes in raw performance. It occurred when efficiency gains made computing capable enough within the power constraints of personal devices.

How ‘Jagged Intelligence’ Can Reframe the A.I. Debate

A.I. has always been compared to human intelligence, but that may not be the right way to think about it. What it does well can help predict what jobs it may replace.

A City Is Not a Computer

This seems an obvious truth, but we need to say it loud and clear. Urban intelligence is more than information processing.

Telemetry Overview - Intelligence Per Watt

Benchmarking Intelligence Efficiency of LM Inference

The Obvious Case for Open and Local AI — Osaurus

AI is a step change in what one person can do—but the power is being rented back to us. Here's the case for intelligence you can run, inspect, and keep.

The Use of Knowledge in (AGI) Society

How to build to break the intelligence curse

Luozhu on Twitter / X

For AI products, people generally think the intelligence dominates everything, while privacy and cost are seen as secondary. The industry’s path shows this: we’ve spent huge money in leading labs to build the largest models with exceptional intelligence.But I believe we’ve… pic.twitter.com/oobAyk55rh— Luozhu (@LuozhuZhang) September 10, 2025

All intelligence is collective intelligence

Author: Falandays, J. Benjamin et al.; Genre: Journal Article; Published online: 2023-04-28; Open Access; Keywords: Collective intelligence, self-organization, multicellularity, neural Darwinism, behavioral coordination, cultural evolution; Title: All intelligence is collective intelligence

Why A Neurotech VC Bet On Anthropic: Intelligence, Interpretability, and the NeuroAI Stack — Kaleida Capital

Why would a neurotech VC bet on Anthropic? This article examines how frontier AI, neuroscience, model behavior, and intelligent systems intersect within Kaleida Capital’s NeuroAI thesis and the infrastructure shaping next-generation intelligence.

Harness Engineering for Self-Improvement

The concept of recursive self-improvement (RSI) dates back to I. J. Good (1965), where he defined an “ultraintelligent machine” as a system that can surpass humans in all intellectual activities and design better machines to improve itself. Yudkowsky (2008) used the phrase “recursive self-improvement” for a specific feedback loop: an AI uses its current intelligence to improve the cognitive machinery that produces its intelligence. This feedback loop in modern AI may indicate the model rewriting its own weights directly, or more broadly the model improves the training pipeline and the deployment system, which in turn enables a better successor model with improved performance across economically valuable tasks. The speed of research development in AI has been shown to drastically accelerated in frontier labs (Anthropic; OpenAI).

Conducting smarter intelligences than me: new orchestras | Notion

So what did I do?