To The Moon

The product is wrong, the strategy is off, and now Apple is suing.

I intended to publish an article on AI’s enterprise incentives yesterday, but it’s been expanded into much more after the drama of the last 72 hours.

Since Thursday, ChatGPT released an update without a chatbot, on Friday, Apple sued OpenAI, and over the weekend, Sam Altman took on Elon Musk on Twitter.

But before we get to all of that and how it’ll impact investors, let’s start with the weekly update.

OpenAI Gives Up On Consumers?

OpenAI updated its Mac app last week, and it told us more than they thought about the near-term future of AI.

ChatGPT, as we know it, is dead.

Enterprise AI and AI for “work” is the future, even for the product that reached a billion consumers faster than any other.

This is what ChatGPT looks like now (I had to check for myself, and sure enough, this is what it looks like). Chat is gone (or at least demoted) and replaced by “Work”.

You can still chat with ChatGPT, but it’s setup more for projects and scheduling tasks.

In other words, it’s built for work and coding.

This isn’t all that surprising. I said OpenAI would “collapse” in my 2026 predictions because the company was losing money so rapidly.

Since then, Anthropic has shot to the #1 position in AI revenue and is stealing the lucrative enterprise customer. OpenAI has to follow into the enterprise space because it has no strategy to monetize the consumer business.

ChatGPT’s start as a consumer app and future as an enterprise app became confusing because the app you probably use didn’t include Codex, the coding tool that’s driving the company’s revenue growth. And moving between ChatGPT and Codex was cumbersome.

So, they’ve unified the apps…which makes a lot of sense strategically. But it also involves making decisions. What do you prioritize, and what do you not?

If the money is in work and coding…you prioritize those!!!

Chats are so deprioritized that they’re the bottom option in the app and open an awkward window. They’re clearly not the main purpose of the app anymore.

It makes sense from a business perspective. Consumers are a loss leader for OpenAI. Why prioritize them in a unified app?

But that also tells you everything you need to know about the future of the most popular AI chatbot in the world.

Its future isn’t chatting at all. It’s work.

AI & Enterprise Incentives

If Anthropic and OpenAI are both going after the same enterprise customers, what does that look like strategically, and how does anyone differentiate?

You may have a preference in one model or interface over another, but in the enterprise, the incentive isn’t user experience, it’s ROI.

That means costs matter. Returns matter. And it’s not clear anyone has been operating that way in the AI world, yet.

Go back to the 1980s and PC adoption to see how this could play out.

$Apple(AAPL)$ Mac was built from the ground up for a great user experience. Everything was in one case, and it all worked together.

PCs had separate monitors, cords, drivers, and floppy disks. They were clunky and cumbersome.

And the PC won!

Why?

Because PCs also had Excel, Word, PeopleSoft, Lotus Notes, and all of the software businesses wanted to be more efficient. It was enterprises that first drove PC adoption in the office and later at home, due to network effects.

If you knew how to use a PC, why buy a Mac for home?

PCs were familiar and had more software, and those advantages compounded as businesses became more and more PC centric.

But where did the value accrue in the PC era?

$Microsoft(MSFT)$ (operating system) and Intel (chip) because the core of the industry’s value, and they commoditized everything else around them.

Memory, assemblers like Dell and Gateway, and even retailers were commodities. They were replaceable. They competed on cost, not differentiation.

That doesn’t mean they weren’t good businesses, but Microsoft and Intel were the ones who became two of the biggest companies in the world.

The mobile era was different because Apple won the hearts and wallets of consumers one by one before enterprises could set the standard. Consumers came first, and enterprises followed, kicking and screaming.

In the early years of the iPhone, it was impossible to get corporate email on your phone or connect to corporate networks. $BlackBerry(BB)$ was fine, but Apple wasn’t welcome in the enterprise for a long time.

Eventually, Apple and $Alphabet(GOOG)$ Android won, but that’s because they won consumers. Enterprises had no choice.

What’s interesting today is that AI looks more like the PC era than the mobile era. Enterprises are leading, and consumers are lagging.

OpenAI just confirmed that it was the case.

What does that mean about the future of AI? Decisions about winners and losers will be centralized, not distributed.

The chip your AI runs on will be determined by Anthropic or OpenAI. Eventually, that will mean chips and data centers will become commodities, competing on cost.

Models themselves may also become commodities. Anthropic and OpenAI have an incentive to keep workers on their platform, but will there be a payback to developing a new model when an older one is 1/10th the cost?

Memory, which is hot right now, will go back to being a commodity.

Devices themselves may even commodify, just like PCs did. At the very least, I think a world of more devices and a less dominant form factor than mobile is in the cards.

If your boss is calling the shots on what app you use, what model they prefer, and how much your token budget is, it creates a very different incentive structure than we’ve been living in for the last 20 years of the internet and mobile era.

Consumers are no longer in charge, and that means dollars and cents will win the day.

To me, it’s easier to see how this structure leads to losers in today’s market. For example, memory, neoclouds, and even chips will commoditize.

But will there be network effects or a 2-sided market that emerges? I don’t see one yet.

Apple Sues OpenAI

To make matters worse for OpenAI, Apple sued OpenAI for trade secret theft late last week.

I always look at corporate lawsuits with a skeptical eye, but this one is insane! Apple employees took proprietary information when they left, led by Tang Tan, who is now OpenAI’s Chief Hardware Officer, and happened to become incredibly wealthy when OpenAI bought io, his “stealth startup” formed with Johnny Ive. Even the initial complaint shows a coordinated effort, and I wouldn’t be surprised if more damning evidence comes out over time.

Even in the few legal documents I’ve read, the details Apple has are incredible. You can read the entire complaint here if you’d like.

How did Apple know what was stolen? The company is notoriously private and uses CIA-style tools to catch people leaking information. This from Trung Phan is fascinating.

You can see why OpenAI wants to get into the hardware business. Where’s the moat if you’re competing to provide intelligence at the lowest cost to enterprises?

Having a device people love would be a great tool in the tool chest to gain some leverage.

But maybe they went too far trying to build that moat.

At the very least, I think this makes it less likely OpenAI is able to IPO this year. The company won’t likely be able to launch any hardware, and investors will want to see how this episode plays out before buying the thesis. OpenAI is a trillion-dollar company.

And remember, OpenAI is still burning billions and needs funding to survive.

With an unclear strategy and now a lawsuit from one of the most profitable companies in the world, OpenAI’s future looks murkier than ever.

When Billionaires Behave Like Children

I have to point this out because these people run some of the most valuable businesses in the world and are driving trillions of dollars in value up and down the value chain.

Sam Altman and Elon Musk are going back and forth on Twitter like a couple of children.

I don’t like the discourse, but it’s also telling about the two men because they’re both right.

Elon Musk has been lying (I’m sorry, he has) about what his company will deliver to customers for years, but has been rewarded with an ever-higher stock price.

Sam Altman has given me scammy vibes for years and reminds me more of SBF than Steve Jobs.

What I think is true is that we’ve reached the point where a few men (because it is mostly men) determine the future of artificial intelligence, and investors often have to take sides.

In this case, I don’t trust either leader.

Eventually, adults will enter the room.