I haven't really considered this but: the USG can afford to NEVER make "Mythos class" or higher generally available, and not fear Chinese competition or revenue loss. Shrimple: "this thing is full of CCP sleeper circuits and has unknown cyberattack potential. If you use it, you may be compromised, EVEN if you run it locally, and we can't know how badly. So no SWIFT/MasterCard/AppStore/etc for you. We can't take the risk of our infrastructure getting contaminated! Your choices are to die or to repudiate China and all its ways and submit your codebase to us to have it derisked by the airgapped Mythos3, you pay of course».
This simple trick destroys the market for frontier Chinese models. even INSIDE China. You'll be buying Opus 4.8125 for the next 20 years.
I’d love the algo to have a switch for punishing hyperbolic engagement openers. It’s not “breaking”, it’s “two years old” and your incremental improvement will not result in things “never being the same again”.
You can set up an base model (here gpt-neox-20b, all shoggoth, all the time) and have Opus use it to infuse its own anemic writing with some real crazy.
Suprisingly Opus here does have a real appreciation of what its otherwise missing out on and just can't bring itself to do.
Also: McDonalds line is pretty great.
Been at Sanity nearly eight years. Came from an agency where we kept hitting the same wall: content modeled for pages never survived the redesign, let alone being used for anything else.
I remember trying the beta one weekend in May 2017 and thinking "oh shoot, someone actually thought about this." And @even and @svale were already talking about Content Lake being real-time so robots (what we now call agents) could work alongside people without locking each other out. Turned out to be a pretty solid bet.
A million users in, I'm just one of them. Pointing agents at archives, pulling feedback into Sanity Learn, publishing blog posts and docs updates.
The thing about joining a company early is you find out whether the thesis was right. This one still is.
Our Content Agent is now not only in-product and available as an API, but now also officially up on Slack marketplace. Let's let it speak for itself: https://t.co/B3FQpVQ8Sd
@emollick Next you will be complaining about the speed of the infrastructure granting you access to global information resources as you consume it along with a beverage at 30k feet
"The model wasn't broken. The context was."
@MHillestad on why AI agents need structured content, not bigger context windows.
Structure powers intelligence.
https://t.co/eO9Vt7vcTm
Why can't my coding agents—backend, frontend, architect, designer���just hash it out in a thread like people do?
We made a thing over the holidays to try it. MIRIAD is basically Slack for agents.
We are prepping for an open test run: https://t.co/hJCFSGHJOy