-
On Scaling Effective Compute
Jason A. Hoffman, PhD The industry quotes compute in units of intention. Gigawatts announced, accelerators shipped, capital committed, peak FLOP/s on the spec sheet. Contracts are signed in those units, market share is argued in those units, and national positions are compared in those units. No model has ever been trained on any of them.…
-
On Open Weights
Jason A. Hoffman, PhD The previous post, On Binary Narratives, used the coverage of Moonshot AI’s Kimi K3 release to show how a story about many interacting variables gets compressed into one variable with two values. That post was about the method. This one applies it to the second binary inside the same story: open…
-
On Binary Narratives
Jason A. Hoffman, PhD On Friday the Chinese startup Moonshot AI introduced Kimi K3, which it described as the world’s largest open-source AI model. The model was available through Moonshot’s products and API; the full weights were promised for release by July 27, with the technical report still to come. The New York Times covered…
-
On Starting Another Company
Jason A. Hoffman, PhD | July 19, 2026 I have started and sold several companies. I am still a technologist and an operator, but I now have more ways to help than I did when I was younger: investor, advisor, first customer, partner, recruiter of the person who should run it. Capital and time are…
-
On Arguing Against Jevons
Jason A. Hoffman, PhD What is it about human beings that we keep arguing against Jevons’ paradox? In 1865, William Stanley Jevons published The Coal Question and observed that more efficient steam engines had not reduced Britain’s coal consumption. They had increased it. Efficiency lowered the cost of work done with coal, the lower cost…
-
Joining a Company the Founder Still Runs
Jason A. Hoffman, PhD | May 5, 2026 There is a particular failure mode that plays out when a senior executive joins a company whose founder is still in the building, still engaged, still making the decisions that matter. The executive arrives with credentials, a playbook, a network, and a reflex. The reflex is to…
-
On Strategic Decision-Making
A commitment to a course of action under certainty is a plan. Under uncertainty, it is a decision. Decision analysis is statistical at its foundation. This post is about the structure of what makes decisions hard, the recurring decisions that define a company’s trajectory, and what decision support could actually become.
-
On Measuring Strategic Work
KPIs are simple. That is their virtue and their failure mode. The tension is real: simple becomes simplistic, complex becomes complicated. The discipline is staying in the useful middle of both.
-
On Critical Thinking in a Multi-Model World
This is not a post about prompting. It is a post about thinking. The techniques I use with AI models are not new. Adversarial analysis, dialectical argument, Socratic questioning, pre-mortems, steelmanning, red teaming: these are critical thinking disciplines that predate computers.
-
On a Year of Multi-Model Assisted Development
In February 2025, Claude Code launched and I switched back to CLI and git. Within months, Codex CLI and Gemini CLI followed. The question shifted from “which model” to “how do I use all of them together.” Here is what a year of that looks like.