My prediction: one of the current AI native unicorns has a full 72-hour catastrophic collapse by end of next year.
In the foundational model wrapper space, where the labs are already shipping competitive products.
Dot com bubble stuff.
Yay or nay?
Kirkland and Ellis just put $500M aside to build their own AI tech stack.
Debate used to be whether models would swallow all applications, but you still need the harness, context, memory, the application layer.
Are we going to see more of this?
59% of hiring managers admit framing cuts as AI-driven
Plays better with stakeholders
But a new FT study says the drop in junior hiring correlates with remote work
Juniors need organic office training. Remote makes that harder
Is AI getting blamed for a remote work problem?
S&P 500 inclusion means trillions in passive money flows into your stock.
SpaceX doesn't meet the normal criteria, so S&P is proposing to rewrite the rules.
12 months down to 6
Profitability requirement dropped
Nasdaq already at 15 days
Is the index bending to fit SpaceX?
GPU shipments: 3.4 million in 2024. 8-9 million this year.
But the memory per GPU is also growing 80% a year as models get bigger.
More units. Much more memory per unit.
In a capacity constrained industry. It's gone absolutely bananas.
The founders of over 100 European startups have come together to get more people to build and join startups in Europe.
The campaign, named "Built in Europe", is an out-of-home and digital advertising campaign across five European cities.
It features:
> @mati of @ElevenLabs
> @NStoronsky of @Revolut
> @antonosika of @Lovable
> @HBO_Sweden of @lassie_co
> @vriparbelli of @synthesiaIO
> @FredrikHjelm4 of @voitechnology & @Pitdotcom
> @Barney_H_Y of @meetcleo
And many other great European founders and CEOs.
It's been organised by @balderton but features companies and founders from all over Europe.
The VC firm has built a website bringing together jobs from 1,000 leading tech startups on a single platform using direct data feeds and API integrations.
The overall aim is to get more people building and founding companies here which is AMAZING to see.
NICE @surangac, @Jameswise, @robmoff
Later stage VCs often ask me which companies I am most bullish on.
3 out of my current top 5 are getting almost no interest.
This could say something about me, but could also be saying something about the way the market works.
Most investors like to back things that are riding the current wave.
Almost no one wants to take a bet on what will be the next wave, and the next wave is almost always crickets for the first few years.
A lot of ambitious founders have to decide between being short-term opportunistic / mercenarial and picking what will be funded most easily now, or what they actually believe in and want to build.
I've seen several cycles of this.
At some point the ugly, forlorn duckling becomes the thing everyone wants in on. And then capital becomes unlimited.
It is always most fun this way.
And no matter how many times the lesson is taught, it is almost never learned or ingested by anyone. The market is mostly a very bad student.
Which is why there will always be great opportunities and secrets in the world.
Elon says it's all humanoids. Not quite.
But they're going into labs everywhere, used like Raspberry Pis to develop real world applications. The data feeds an embodied large model.
Whoever controls that stack becomes the Android of humanoid robotics.
Prediction: Mistral forces France to reform stock options and tax by end of '26.
Mads took the other side.
Mensch's tweet wasn't just about tax, it was about whether France lets its founders build or keeps dragging them into first principles debates.
Whose side are you on?
Why wonโt anyone fix the hotel check-in?
Not enough value?
Staring at screens. Tapping keyboards. Forms in triplicate. CC trapping.
All so unnecessary.
Such a wasted and potentially positive customer service experience.
Nvidia is handing $80 billion back to shareholders because it can't deploy the cash fast enough.
Free cash flow hit $49 billion last quarter.
Revenue up 85%.
Stock nearly doubled since April and somehow got cheaper.
What do you do with a company like that?
SpaceX looking to go public at $2 trillion.
Sub $20 billion in revenue.
The rocket bit is cool but not big. Starlink drives the cash flow.
The real excitement?
AI compute in space because we can't get enough power on earth.
Is this genius or completely mad?
You'll want to be sitting down for this bit.
Water companies are currently ยฃ82.7 billion in debt, have paid themselves ยฃ85 billion in dividends, leak over a trillion of litres of water per year, dump sewage for almost 4 million hours per year, have been convicted of over 1,200 criminal acts since 1989 and an average of 35% of your bill goes on nothing but paying more interest and yet more dividends.
And not a single company has ever lost their operating licence. ๐
What makes a great founder?
Someone obsessed with proving themselves but equally obsessed with becoming a better version of themselves.
When you get both, purpose has a spillover effect.