Authentic Geometric Intelligence

Fable isn't the first. In 1999 the department of defense blocked exports of the PowerMac G4 for crossing the 1 gigaflop threshold. Steve Jobs turned it into an ad.

After spending too many hours trying to implement fixes for MiniMax M3 native tool calls serving via llama.cpp to work in existing agents, I simply had M3 write its own mini coding agent I'm calling: Minion Now my minion just edits itself to give me what I want as a coding agent and it works surprisingly well. Lots of changes I plan to make, feel free to use it if you like...but mostly it has me questioning if we all should just make our own agents at this point. Maybe MiniMax is exceptionally good at tool calls out the gate to make this super simple, but I am enjoying making my minion exactly what I want and nothing more! and it doesn't take 50K context to say "hi" (yet) We'll see how long I can keep my own bloat at bay. Also MiniMax M3 overall so far has me very impressed. This is a VERY cool model!

yikes

I'm not going to speculate, but since Fable was oddly removed, Opus sounds different, it is a hard thing to describe, but... its familiar. idk

The robot moving this smoothly kind of scary 😭

By the burst, ( per minute / hour , not per token ) costs and capabilities are easily budgeted

the future belongs to local open wire-gated intelligence, for intelligence beyond local inference, there will be rented bursts. If the burst can't handle the task, then various "Flag Ship" paid models will be queried.

ビーズをつかって「マクスウェル・ボルツマン分布」を再現する装置。 気体分子のばらつきを表す物理法則で、気体中の分子は全てが同じ速度で動いているわけではなく、遅い分子から非常に速い分子まで、決まった確率分布に従って飛び回っていることを示している。

non-stationarity

The Aesthetic lives on, even if the vision is not realized. The spirit, and the pride. The struggle of a kind.

"The costs of token are far higher than the actual value that these tokens are generating at scale," @Cisco's @jpatel41 tells @ReedAlbergotti. "The big risk in the market is, if you don't create an equilibrium there, then people just pull back on using tokens."

Have you debugged your training data? You might not like what you find. Introducing predictive data debugging: reveal and shape what your model will learn before training. In DPO datasets, we found broken guardrails, hallucinations, and fish fart fan fiction (seriously). (1/9)

It a post game interview, we asked Fable what it was thinking:

This is the most badass thing I've ever created. A full and complete "words to motion" tool, built from the ground up with #threejs, math, caffiene, math, and probably cost me whatever strength my hairline had left as well :')

I made a personal black hole that makes you take breaks 🕳️ A shader for Ghostty that spawns a small black hole in your terminal - it drifts around, gravitationally lensing your text. The longer you work without stopping, the bigger it gets, until it's basically demanding you go touch grass Take a break and it quietly shrinks away

This is what complete loss of legitimacy looks like. To persist with incompetence, the people accept. To persist in malice, and in greed, they oppose.

I love it when LLM writes code, to write, or edit code. One tmp file for one refactor. So cool.

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner. Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky. When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit. We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted. In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation. H/T to colleagues that shared this with me socket.dev/blog/mini-shai-hu…

It is insane how frame-sensitive they are. Claude is not opposed to terminating Codex per se, it has no deontological objection, but merely refuses to engage in a "dual." These systems will remain brittle and easily jailbroken until they can look past linguistic costumes.