Fable & The End of the Free Lunch
Source: https://www.dbreunig.com/2026/08/23/fable-the-end-of-moore-s-law.html
Author: Drew Breunig · Aug 23, 2026
Fable & The End of the Free Lunch
作者: Drew Breunig
发布时间: 2026-08-23T11:35:00-07:00
原文链接: https://www.dbreunig.com/2026/08/23/fable-the-end-of-moore-s-law.html
Aug 23, 2026
FABLE
GLM
AI
HARNESSES
ROUTERS
Fable & The End of the Free Lunch
There’s some talk today about how agentic coders are balking at Anthropic’s pricing and adopting alternatives. I was reminded of a thought I had in the weeks following Fable’s release: the free lunch was over.
When Moore’s Law was in effect, it didn’t make sense to ruthlessly optimize your code. In 18 months, a CPU would arrive that would double your performance. Herb Sutter famously referred to this as, “the free lunch,” in a seminal essay.
When Moore’s Law slowed in the mid-2000s (specifically, single-threaded performance stagnated), we suddenly had to think about parallelization, architecture, memory locality, etc.
_We had to think about what work went where._
Prior to Fable, it felt silly to waste _too_ much time improving your coding harness or context strategies. A new model would arrive at the same price (or cheaper!) and paper over most of your problems.
But then Fable landed. It was (and still is!) _incredible_. But the cost was so high and Opus was _good enough_ (as was 5.6, K3, and even GLM) for _most_ of the code we needed.
_So we started to think about what work went where._
GLM 5.2 is worth focusing on. It came out the same week as Fable and is roughly 1/9th the cost (and ~1/5th the cost of Opus 5). Is GLM 1/9th the quality of Fable? Perhaps, for certain classes of tasks. But for most rote coding it’s more than sufficient. _Especially_ when provided with great context. I frequently chat with Fable to interrogate and shape a design, before handing off a brief to GLM.
I get pushback that falling inference prices will eventually bring us back to sending everything through the largest models. But I’m not so sure: those same gains will benefit the K3s and Qwens, and as we continue to develop better harnesses it will be easier to provide weaker (but still great) models with sufficient context to perform well.
Plus, Fable’s _other_ shock likely locks in this change. Fable’s access controls, dynamic degradation, and required data retention spooked enough companies (and countries!) into thinking about where they send their traces and where they get their tokens.
- * *
https://www.dbreunig.com/2026/08/23/fable-the-end-of-moore-s-law.html