Philochat

💬New Philochat

Kimi Yemeni SpaceAI

Hi, I’m Philo, founder of Philoinvestor. I’m a generalist, opportunistic, and idiosyncratic investor who believes the best opportunities come from independent thinking, not consensus or vibes.

For More: X | Telegram | CoTrader 3000 | Philo's Academy

Paid subscribers receive exclusive investment research, full access to the Philoinvestor archive, the No-End Dividend Index, Breakout by Philo, Flash Philo, and subscriber-only updates.

Moonshot AI: Chinese AI lab.
Kimi: Their Chatbot.
K3: Their recently launched open-weight LLM. They say it beats rivals top LLMs in coding and web engineering tasks.... 👇

This was a catalyst for the AI-related sell-off since July 16th, when the model was released. But things were falling weeks before that, so don't anchor too much on it.

In fact, the major catalyst was probably Apple's move to try and get clearance to use Chinese memory in their devices – because companies like Micron raised their prices too aggressively!

We gave the setup on Memory post-Micron earnings in this piece, memes were on point here.

Wrecking Ball

Back to Kimi.

We touched on the concept of the wreeeecking baaaaaaall – and how it matters in understanding this stage of the AI cycle.

We even used a wrecking ball in the cover of our piece Zooming out on AI, Part #1 – earlier this year.

And now, Kimi comes like a wrecking ball to compete head on with IPO hopefuls, OpenAI and Anthropic.

But as we've explained ad nauseam, Chinese AI is way more efficient than American AI... And therein lies the problem for American AI – and the picks and shovels and long compute trade.

See below the chart I included in Flash Philo on Memory, from June 29th.

What comes after Brute Force, Mass Compute, Price Agnosticism?

Instead of maybe considering that we are entering a more compute efficient time in AI, where models will get not only better – but cheaper and more efficient.

Analysts and thought leaders default to the "THE WORLD IS COMPUTE CONSTRAINED" argument. The tweet below was a case in point.

Guys, the world is so compute constrained. I honestly don’t understand how anyone can think otherwise.

Almost nobody is seriously using AI today, and we’re already hitting capacity. Everything is a bottleneck: land for data centers, permits, electricity, grid connections, chips,… https://t.co/Qk0pTn4xTq
— Nicolas Bustamante (@nicbstme) July 19, 2026

No need to say more here, but how events are unfolding in this market makes it extra difficult for the AI Brothers: OpenAI, Anthropic and SpaceX.

The first two are hoping to IPO, the latter has already done it. We wrote about the problems they face here:

Space X

SPCX shares down ~25% since we published our critical piece on the Alchemy of Elon, and the hopes & dreams his valuation relied on... 🔻🔻

But I wouldn't write off this stock just yet, the company's actual SPACE segment, which as we've said is small compared to their AI segment, could be the source of positive catalysts going forward.

And regardless of the space-high valuation, space enthusiasts only need one positive catalyst to start bidding up this stock again. As for myself, I will wait for a bottoming formation before looking to speculatively trade SPCX on the long side.

For now, SpaceX's space segment is facing issues.

SpaceX seems to be slipping into a slump, as another mission was scrubbed at the last moment. The launch of a Falcon 9 rocket carrying 24 Starlink satellites on Monday, July 20, from Vandenberg Space Force Base in California was aborted due to an unspecified reason just when the clock struck T-0, making it the second SpaceX mission in less than a week to get scrubbed. In a post on X, SpaceX confirmed it is standing down from the launch but did not reveal what caused the abort. “Standing down from today’s Starlink launch; vehicle and payload remain safe,” the company posted.

Non-AI Opportunities

It's becoming very clear that AI-related opportunities are getting harder and harder to find, Korean volatility and margin calls and all.

Trump was forced, once again, to come out with AI protectionism on the Moonshot Kimi news – to save the stock market etc. But Trump's attempt to ban Kimi and Chinese AI will not amount to much in the end. This is not the place American AI wants to be, and neither is the general stock market.

So what do we do?

My next piece will blend the wisdom of Charlie Munger's technical analysis, yes, technical analysis – with a fundamental thesis. And we will do this with a number of stocks. Can't wait for this post to come out!

Iranian War Expands

Seems the war in and around Iran is expanding, but the world isn't reacting to it the same way. In previous rounds it was all we heard and saw – now we don't see anything and barely hear anything in the news.

Yesterday there was news that Iran is considering invading Kuwait to take over American bases in the country, and that Yemen's Houthis are closing the Bab al-Mandeb Strait as their conflict with the Saudis reopens.

A closure of that strait (see image above) basically locks Saudi in and cuts off any ability to export oil, as Hormuz is already controlled by the Iranians.

Meanwhile...