I have to give credit to Morgan Stanley’s Erik Woodring. During Apple’s earnings call, while everyone else was lobbing softball questions at Tim Cook about their impressive $143.8 billion revenue quarter, this guy actually asked the question that’s been gnawing at the back of every reasonable person’s mind: how exactly is anyone planning to make money from artificial intelligence?
The answer he got was peak corporate non-speak. Cook rambled about “bringing intelligence to more of what people love” and how it “creates great value” that “opens up a range of opportunities.” Translation: we have absolutely no idea, but we’re committed to spending billions anyway.
The Emperor Has No Business Model
Here’s what really gets me. We’re living through what’s supposedly the most transformative technological shift since the internet itself, and nobody in Big Tech can articulate a coherent monetization strategy. Not Apple. Not OpenAI. Not any of them.
OpenAI is the poster child for this dysfunction. They’ve become a household name. ChatGPT is everywhere. And they’re not planning to be profitable until 2030. That’s not a typo. 2030. And even that timeline is apparently optimistic, according to HSBC analysts who point out they’ll need another $207 billion in funding to get there.
Ask anyone in tech how this math works and you’ll get that shrug emoji in verbal form. I’ve been in enough conversations to know. The vibes are immaculate, but the spreadsheets are a disaster.
The Infrastructure Gold Rush
What’s fascinating is how this mirrors every other tech bubble we’ve seen. During the dot-com era, companies were burning cash on the promise of “eyeballs” and “engagement” without clear paths to revenue. We all know how that ended.
But there’s a crucial difference this time. The infrastructure costs for AI are staggering and ongoing. You can’t just spin up a ChatGPT competitor in your garage. You need datacenters, GPUs that cost more than houses, massive energy consumption, and armies of engineers. OpenAI reportedly loses money on every ChatGPT Plus subscription even at $20 a month.
Think about that. They’re losing money on their premium tier. The tier that’s supposed to demonstrate the monetization potential. And somehow we’re all just nodding along like this makes sense.
The Integration Theater
Apple’s approach is particularly interesting because it reveals the playbook everyone else is following: just integrate AI into everything and hope something sticks. They’re weaving it into the operating system, into apps, into services. Creating “great value” as Cook puts it.
But integration isn’t a business model. It’s a feature enhancement strategy. Sure, maybe AI features help sell more iPhones or keep people in the Apple ecosystem. But is that worth the billions in R&D and infrastructure? Where’s the breakeven analysis?
I keep waiting for someone to stand up during one of these earnings calls and ask the follow-up question: “Yes, but specifically, how much additional revenue per user do you expect AI features to generate?” Nobody does. It’s like we’ve collectively agreed not to peek behind the curtain.
The Subscription Mirage
The obvious answer everyone assumes is subscriptions. Add AI features, charge a premium tier, profit. Except this strategy is already showing cracks. Microsoft is bundling Copilot into enterprise subscriptions. Google is giving away Gemini features. Meta is just eating the cost entirely as part of their advertising business.
When everyone’s giving it away or bundling it, who’s actually going to pay premium prices? The enterprise market can only absorb so many $30/month per-seat AI tools before CFOs start asking uncomfortable questions about ROI.
And on the consumer side? People are already subscription-fatigued. We’re paying for streaming services, cloud storage, productivity tools, and now you want another $20/month for AI features that half the time hallucinate incorrect information?
What This Means for Developers
If you’re building on top of these platforms, this should concern you. The lack of clear monetization paths means the current generous API pricing and access could evaporate overnight. OpenAI has already changed their pricing structure multiple times. Stability in this ecosystem is an illusion.
I’m also skeptical of the “build an AI startup” advice that’s everywhere right now. Unless you have a genuinely novel approach or you’re in a specific vertical where you can charge premium prices, you’re building on quicksand. The big players are giving away features you’d want to charge for, and they can afford to lose money indefinitely.
The companies that might actually make money from AI are probably the boring ones. The infrastructure providers selling GPUs and cloud compute. The companies solving specific, expensive problems in healthcare or legal or scientific research where the value proposition is crystal clear and measurable.
But the consumer AI play? The “ChatGPT for X” startups? I’m not seeing it. Not when the actual ChatGPT can’t figure out how to be profitable.
The funniest part about all of this is that in five years we’ll look back at this moment and it’ll be obvious what the right monetization strategy should have been, and we’ll wonder how everyone missed it while they were too busy trying to integrate AI into their smart toasters and generating marketing copy about “great value” and “range of opportunities.”