KevinSimback

What’s the minimum cost to run inference for Kimi 3?
Yeah, we’re going to need way more compute and memory

Reminder that just 18 months ago, the best models were:
OpenAI o1 was the leader among the closed source labs and Deekseek R1 had just released, otherwise best open weights was Llama 3.1 405B
Imagine 18 months from now

Prediction: Anthropic will have to bend on all these points
> Anthropic limits sub plan use via external agent harnesses like Hermes
> Fable is nerfed on many topics
> Fable is limited in sub plan to 50% of usage + goes API only at $50/m output in 3 days if not extended
But we now have GPT 5.6 Sol and Kimi 3 plus others nipping on the heels
Competition is a helluva forcing mechanism

Timeline this week:
Monday - Sol 5.6 ftw
Tuesday - back to Fable
Wednesday - back to Sol 5.6
Thursday - omg Kimi 3
Looking forward to seeing what tomorrow brings

Let's talk about shared context layers, often referred to as "company brains"
To answer hard questions about your company like:
"what's true about X, since when, according to who, and where do accounts disagree?"
You need context that is a derived, citable fact layer over all your communications and documents
It's the system of record for what the company or organization "knows" which is different from the systems of record containing the outputs from work that is done
Jack in accounting knows why a vendor invoice was paid that didn't match the PO, whereas the accounting system just contains t

The sovereign intelligence flywheel:
open weights → post-train → evals → agent runtimes → trajectories back into post-training
The companies that make this easier and more accessible to enterprises will crush over the next 5 years

I’ve gone the entire day without a
“🚨BREAKING: Someone just…” post
Feels good

“The reports of my death are greatly exaggerated”
—Enterprise AI investment, July 2026

Almost finished with my most thoughtful AI article of the year, should be ready to publish tomorrow
It addresses the most important question for the AI industry right now and is quite timely given the discourse over the weekend

Farmer’s markets have basically become an IRL Etsy with a few produce vendors mixed in to keep up appearances
It’s mostly overpriced things that you don’t need but you end up buying anyway

The weekend is shot:
> GPT-5.6 + ChatGPT Work
> Meta Muse 1.1
> Grok 4.5 + Grok Build
Who's got the best writeups / videos on these?

There are some incredibly good R/R opportunities in the crypto x AI sector right now
The right filter is: true AI builder talent + products that appeal not just to crypto natives
Several names that could easily pull a 5-10x regardless of what the majors do

I believe we’re going to see the biggest wave of tech M&A in history in 2026-27
Massive equity valuations to use as currency + reshaping industry dynamics with AI = buying frenzy

Anyone who isn’t insanely biased would look at the game and conclude that Balogun was unfairly treated
Overturning it was the right call regardless of who helped make it happen

Just finished listening to latest @theallinpod which had a discussion on the “rant” by Alex Karp that went viral this week
The biggest realization that everyone is waking up to is the *lack* of competitive edge when using frontier models
If all your competitors use them, you have no differentiation
Worse yet, using those models could subject you to competitive risk if the labs decide to take your data and make products that compete with you (just ask Figma)
The crazy icing on the cake is that these models are 5-100x more expensive than alternatives that can be made just as good with the right

Imagine the possibilities
Inference capital markets x $ansem tie up
Leverage the attention to supercharge decentralized AI, agentic markets
cc @blknoiz06 and @0xgilbert