Are AI Labs too big to fail?
It makes no difference.
Investing in too big to fail (TBTF) is not an investment strategy. Yet increasingly, that's what many are doing.
Many of you weren't around in 2008, and those who were, think the Global Financial Crisis has zero relation to what's happening today...
AI is a transformative technology that will change the world forever!
So what? Tell me what that means for the business.
Compute demand will always go up, and Jevon's Paradox this and that, and demand for intelligence is infinite...
BTFD Forever
This is the narrative being repeated by everyone.
But if AI is so great as a business, why are AI Labs (i.e. the American ones) now turning political, all at once, and looking to become a sort of government-protected utility?
Under the guise of incredibly fast AI development – they have in unison decided to slow the frontier until better security and guardrails are put in place etc.
Sam Altman (of OpenAI fame) is going cold from IPOing this year while Anthropic has announced they will be profitable for a second quarter – on an Adjusted Operating Income basis.
Inference Margins...
If you're not getting the message, most of these companies are already gross-margin profitable. The problem is everything below gross profit!
In fact, training runs are costing them a lot of money – and they don't want to tell us about it.
This is where Inference Margins enter the picture – to lull everyone into thinking that profitability is a given, as these companies grow and look to IPO.
But the truth is they are colluding to reduce their costs while they get more time to try and grow out of the hole.
Survivors?
They are trying to survive and thrive not by remaining competitive and creating value – but by abusing their position, and the rest of us!
In AI Brothers, we compared the GFC with today, and concluded that the setup is the exact opposite. Excerpt below.
Yet they still need a bailout!
But what's there to bailout? Chinese AI and open-source models are going to eat these labs alive – if they don't eat each other first!
In Trump's Uno Reverse Card we argued that Trump was between a rock and a very hard place – and that he could use some pushin' from the AGI people.
It seems that they aren't able to give it to him, because they need some pushin' of themselves...
But less than two months before midterms, Donald cannot be seen to bailout anyone.
He needs every last vote he can get, so his response to the AI Brothers was crystal clear 👇
That's literally what he said – so you see why I was reminded of the famous Ivan Drago quote.
Cold War!
Not just a Cold War, but a war on many fronts:
- East Vs West
- Open Vs Closed-Source
- Between AI Labs themselves
What happens next?
With these developments and having put everything together, I think the path forward is increasingly becoming clearer 🔮
In fact, if you listen close enough they are telling you what they are going to do.
The hype is in its final stages, and they are being forced to play real ball.
So we know this now follows the typical cycle – and we're getting closer to the end game now.
Let's see why.
