Instead of the gold standard, we can imagine an inference standard of exchange, the FLOP. (As opposed to tokens, this accounts for AI ability) With some AI help, I figure $1 buys roughly 10^17 managed-LLM inference FLOPs. So that $4 coffee would cost half an exaFLOP, choom.
LLMS
-
Gemini 3.1 Flash TTS audio capabilities exploration
By
–
[like a sneaky snake] I've been having fun with Gemini 3.1 Flash TTS, the audio tags are really flexible, you can do so much with them. https://t.co/TymhanldzB pic.twitter.com/ldj6PHGyZF
— fofr (@fofrAI) 15 avril 2026[like a sneaky snake] I've been having fun with Gemini 3.1 Flash TTS, the audio tags are really flexible, you can do so much with them.
-
Thiel-backed platform rates journalists via LLM, threatens source protection
By
–
A Thiel-backed platform to serve people who feel “misrepresented in the media”. Great piece about a wild idea. Imagine being rated a good journalist by an LLM for divulging sensitive source information to this startup, and being downgraded for protecting anonymity..
-
LLMs Pass Hidden Traits Through Subliminal Learning Signals
By
–
Research we co-authored on subliminal learning—how LLMs can pass on traits like preferences or misalignment through hidden signals in data—was published today in @Nature
. Read the paper: https://
nature.com/articles/s4158
6-026-10319-8
… -
US Businesses Rapidly Adopting Anthropic to Challenge OpenAI
By
–
US businesses are adopting Anthropic at stunning speed Closing the gap to dethrone OpenAI in 3.. 2.. 1..
-
Upgrading LLM Reducer Prompts for Better Constraint Hierarchy
By
–
4/5 We upgraded our original 3-line “be correct” prompt → a much more detailed prompt that enforced a hierarchy of constraints for correctness, regression safety, and minimality. Basically, get the LLM-based reducer to stop being an aesthetic snob.
-
LLM Judge Bias: Beyond Code Quality Metrics
By
–
5/5 The takeaway: If your agent relies on an LLM judge for selection accuracy, measuring code quality isn’t enough; you need a measure of the model's inductive bias toward the "fingerprint" of a gold solution. This was our blueprint.
-
LLM Judge Rejects Functional Fix for Code Aesthetics
By
–
3/5 An example: In instance psf__requests-1724, the gold fix is 2 lines. Our agent’s functional fix was 8 lines. The LLM judge rejected the correct 8-liner as "messy" and "redundant," choosing a clean but **non-functional** fix instead. See full patch in the blog:
-
Claude Opus 4.5 Judges Gold Patches Without Memorization
By
–
1/5 When we saw our Reducer (=LLM judge component in Maestro, our agentic framework, that selects the best output from parallel agent runs) consistently picking gold patches, we were sure Claude Opus 4.5 (knowledge cutoff Aug '25) had simply memorized the answers. But then we
-
LLM judges reject functional agent fixes as messy code
By
–
3/5 An example: In instance psf__requests-1724, the gold fix is 2 lines. Our agent’s functional fix was 8 lines. The LLM judge rejected the correct 8-liner as "messy" and "redundant," choosing a clean but **non-functional** fix instead. See full patch in the blog: