The wrapper business model is dead.
Long live open weight models.
You cannot build a business on a model that can be taken away from you, at any time, or have its cost model altered, at any point. Time to wake up.
If this is something you are thinking about, DM us @BasedAI_co
Ref: https://t.co/5SoZxe6pUS
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: https://t.co/bwn0sximKZ
The token-maxxing era is over.
This year, we'll see massive uptake of open-weight models:
- Orders of magnitude cheaper
- Near frontier-level performance
- Your data stays yours
Long @BasedAI_co
Our Anthropic bill is about to jump from $400K → $1.4M/yr.
Not because usage exploded, but because we're about to cross 150 seats.
Past 150 seats you're forced into Enterprise tier. Seats stop including any usage, every token bills at standard API rates. At our current run rate that's 3.5x overnight.
Unfiltered thoughts on AI spend:
1. We should spend tokens to grow as aggressively as possible. But most people (me included) aren't conscious of what they're spending.
2. Visibility comes first. People see their personal number and they're shocked. I accidentally spent $4,000 in 3 days in Claude Code.
3. For engineering the spend is clearly worth it. Pay for the best model, it saves more than it costs.
4. For a lot of other roles it's questionable. Apps nobody uses, skills someone already built. No ROI.
5. Spend limits are coming. We already require approval for more tokens on our support team.
The era of token-maxxing is coming to an end.
In the 1800's factories swapped their steam engines for electric motors and got... almost nothing. The productivity boom didn't come for 40 more years.
Why?
Same reason your AI rollout is underwhelming
🔹Companies buying AI: +5% impact
🔹Companies redesigning around AI: +25%
The entire technology base for most public services runs on US tech, so the blanket sovereignty arguments here are tenuous, at best.
Irrespective of this, "model sovereignty" will become a prerequisite for the state procurement at the UK and EU level.
Why? Models see the data entered at the prompt. Data from the state should remain impervious to foreign access, and the US does have some serious powers to compel (CLOUD Act, FISA, etc).
Future state will be:
- Model hosted by a "local" domiciled entity without a US-nexus
- Models hosted in secure compute/TEE-similar environments
Curious to see how Palantir achieves this.
My take: It is possible. With local model hosting by a local co, and a novel legal framework ensuring local model sovereignty, a US tech platform can serve up a sovereignty-compatible offering.
BREAKING: Palantir CEO Alex Karp has been privately sounding the alarm to AI leaders their companies are going to be nationalized.
"The momentum is on the side of people who want to nationalize them."
Let's be clear: the substitution didn’t fail because the models were weak.
It failed because the hard-fought institutional knowledge, customer dynamics, and the inevitable escalation headache, couldn't survive a headcount cut this brutal.
That’s the part the rollback number doesn’t show - and it’s what this post is about.
If your team is scoping AI work, this one’s for you.
Klarna replaced ~700 support agents with AI in 2024. By 2025 it was quietly hiring humans back.
That is the pattern now, not the exception. The post is what the teams getting it right are doing instead.
Augmenting, not replacing
https://t.co/Cz4rEyK1lS
@louismosley@MartinWrigley I fear that hoping @MartinWrigley - or indeed most of today’s political class - can grasp the nuances between data processors and data controllers is wildly optimistic.
Man. This is a lazy take. Blair isn’t saying we lack agency to reject AI. Nation can, and they will...
He’s saying AI will reshape everything, and thriving economies will be the ones that adapt and capture the upside. We’ve seen this before: electrification, the internet, the steam engine - each resisted by short-sighted statists who couldn’t see past the next election cycle.
@_KarenHao thinks this is a Big AI vs nation state-level debate. It isn't - this is a civilisational dilemma. Whether we like it or not, AI will change everything.
The nations that own the means to harness it, will thrive.
@louismosley The left is brilliant at building bandwagons filled with ignorance.
And sadly, it's usually in the interests of preserving some form of the status quo.
The good thing is the data will speak for itself.
@NVIDIA_AI_PC @NousResearch@ollama This retails for $4.5k
Even if it comes down by 90% it won't serve more than 1% of the population
I think it's more likely that the phone becomes the device storing sensitive data, that your agent - still hosted in the cloud - can have access to
@fabrisera2000 Congratulations Fabrizio!
"Why can't I make my work better by automating parts of it?" - hell yes!
Our mission is to automate the busy work and augment every teams output!
Check out what we are building @GetHirebased - we'd love your feedback
"Businesses are sleepwalking into Claude Code. We built the exit ramp."
Teana Baker-Taylor on the exact moment she knew Venice had to build beyond the consumer product.