Probably the biggest topic the past two weeks have been the takedown and usage of Anthropic’s Fable 5 AI model.
Every AI creator has spoken about how its been banned in and outside of the country due to “security” concerns by the government.
People got depressed.
Even Theo (of T3 chat) who is reputedly “Anthropics #1 hater” subscribed to two Claude Max accounts just to use Fable when it was out.
Because Fable was genuinely good.
Now, its gone.
But then came Z.ai, who released the latest in their flagship GLM model, GLM 5.2.
An open weight model that in many people’s opinions (including myself) performs much better in actual real life tasks than Opus 4.8.
GLM 5.2 was - and is - very well received by everyone.
People applauded how far open source has come in catching up to the big boys.
For a moment, there was morale.
Then morale came crashing back down again because OpenAI wanted to release GPT 5.6 which is reportedly even better than Fable 5.
But they can’t because, well, security concerns.
So Anthropic has not only gotten their models become exclusive, now they’ve have - once again - forced everyone to follow their lead.
Including making their #1 rival’s newest best model banned.
The space is rife with opinions.
“Anthropic has been asking for government regulation, so now they got it! Serves them right!”
“This isn’t actually a security thing, this is just an excuse to suppress people!”
“The tech bros are looking to control WHO gets the tech - look, only a FEW companies gets the model!”
Just thinking about this makes me mad.
I like Anthropic’s models.
But I don’t like Anthropic.
More specifically, the kinds of practices that they do.
And I fear what this means to the rest of us if more and more of these practices become commonplace.
I’ve been saying for some time now that the age of AI subsidy is coming to an end.
That people should be taking advantage of this moment to build for when all this is over, when the latest models are exclusive only to a select few.
But my warnings have fallen on deaf ears and people are still going around doing what they do, the way they’ve always done it (without AI).
And now, more proof is coming out on how big tech is going to play this.
First, they made you fall in love with their products.
Second, they get you so dependent on their products, that it is too costly to leave.
Third, they introduce more powerful and expensive models (2-3x more expensive).
Fourth, they quietly make their older models - the one you’re used to - dumber.
Fifth, they pull stunts like these to hype things up, raise share prices, and make their best models only available to a select few, making the government throw regulations to limit it even further.
Finally, they deprecate older models entirely so you are either stuck paying only premium, or stuck with bad models that can’t do a fraction of what the premium ones can do.
This is why so many people are mad.
Big tech is moving the world around like chess pieces, and the public don’t seem to know or care to learn.
Most are still in denial.
A new phrase has come out a few months ago, and has been steadily gaining popularity.
The permanent underclass.
A separation between the haves and the have nots.
Except this is no longer referring to wealth, but intelligence.
If you are one of the few, you get to have access to intelligence.
If you are not, you are stuck with mainstream AI.
Yes, it is expected that the impact of having the next gen AI is larger than ‘generational wealth.’
And yes, the difference between mainstream AI and the kind of AIs that are about to take over in the next few years is that large.
If you think that the wealth gap is huge now, whatever the difference is, that may be less than a fraction of what it might be in the next decade or so.
But the story isn’t over.
We are in a crossroads right now and we need to take advantage of this moment.
We are STILL in a place where we get to decide how the future goes.
The actions we take right now, collectively, will not only impact us and our children - but their children’s children and the rest after.
This is why I’m writing this today.
Build stuff.
Take this moment to learn and experiment to see where AI and agents can take you and your business.
Mess around and make mistakes.
The way things are right now, we can STILL afford it.
Because AI will never again be as cheap as they are today.
And if you can, contribute to the open source community.
(I’m planning to do this soon)
FELIX
P.S. - I am organizing an “implementation day” where attendees can come in, work on a single project and leave with a completed build using agents.
I will be there to facilitate and help everyone do their work.
It won’t be free but it will be very affordable.
Is this something you’d be interested in?
Hit reply and let me know.
P.P.S. - NLC users and ACF members joins for free.