Alright, buckle up, because we need to have a conversation. There’s a piece going around right now built off some recent comments from Naval Ravikant, the AngelList guy, early Twitter and Uber investor, generally regarded as one of the smarter people in tech, and the headline is, and I quote: “Apple is dead. SaaS is next. You have 18 months.”
Cool. Cool cool cool. Totally normal headline. Definitely not the kind of thing you write to get clicks. Definitely.
So let’s actually break this down, because there’s a real signal buried under the doomsday packaging, and I think it’s worth separating the two.
The Core Claim
The argument goes like this: Naval said on his podcast that “pure software is uninvestable.” Full stop. No qualifier. And from there, the piece extrapolates outward into a whole apocalypse, Apple is structurally dead, your SaaS company is already a corpse and just hasn’t noticed, you have 18 months to find a “real moat” or you’ll watch your valuation crater 70-90%.
Look. There’s a version of this argument that’s interesting. There’s also a version that’s pure LinkedIn-thought-leader cosplay. Let me try to figure out which is which.
Where the Argument Has Actual Teeth
The genuinely interesting part is this: AI coding tools have gotten legitimately good. Like, scary good. The “two people with Claude Code can ship 80% of a B2B SaaS in 90 days” claim is hyperbolic, but the underlying point isn’t crazy, building software has gotten dramatically cheaper and faster, and that does compress moats that were built on “this was hard to build.”
If your entire defensive position as a company is “our product is just better technically,” yeah, that’s a problem. That was always a problem, honestly, the AI thing just accelerated the timeline.
The list of things that do still work as moats, distribution, network effects, data flywheels, hardware integration, vertical depth, that part is solid. None of it is new (Peter Thiel was writing about this a decade ago), but it’s worth repeating because most founders genuinely do confuse “we built something neat” with “we have a defensible business.”
So: software commoditization is real. Distribution and community matter more than ever. Vertical specialists beat horizontal generalists. All of that, I’ll cosign.
Where I Get Off the Train
But “Apple is dead in 18 months”? Come on.
Apple has roughly $160 billion in cash, the most loyal consumer base in tech, the best chip design team on the planet, and they sell the device that physically sits in your pocket all day. The argument is essentially that AI agents will commoditize the OS layer, so Apple’s hardware margins collapse, so they become Samsung. Which is… a take. But it skips over the small detail that someone still has to make the hardware the agent runs on, and Apple is uniquely positioned to do that.
The article even admits this when it lists “hardware integration” as one of the few real moats left. And then… declares the company that defines hardware integration is dead? Pick a lane!
Also, the framing that licensing Gemini = Apple “outsourcing the experience layer” is doing a lot of work. Apple licenses Google for search too. Has for over a decade. Google pays Apple roughly $20 billion a year for that privilege. If integrating someone else’s model is a death sentence, somebody should probably tell the company that’s been printing money off this exact strategy since 2002.
The “18 Months” Thing Is the Tell
Whenever someone gives you a precise countdown, “18 months,” “the next quarter,” “before 2027”, pay attention. That’s almost always a rhetorical device, not an analysis. Real structural change is messy and uneven. Microsoft “missed mobile” and is now worth more than ever. Yahoo “missed search” decades ago and somehow is still here. Companies don’t die on schedule. They limp, pivot, get acquired, reinvent, decline slowly, occasionally roar back. The “you have X months” framing is built to create urgency, not understanding.
What’s Actually Useful Here
Strip away the doom and the underlying advice is genuinely fine, even good:
- Audit your moat honestly. Is your defensibility actually defensible, or is it just “we built it first”?
- Build distribution before you need it. This was true in 2015, true in 2020, true now.
- Community and network effects compound; features don’t. Yep.
- Vertical depth beats horizontal breadth for smaller teams. Yep.
- Solo and tiny-team companies can punch way above their weight now. Genuinely true and exciting.
That last one is the part the article almost buries, that this whole shift is also the biggest opportunity for individual builders in maybe ever. One person can now ship things that used to require Series A money. That’s wild. That’s worth getting excited about. That’s the actual headline, in my opinion, not “Tim Cook better update his LinkedIn.”
My Honest Take
Pieces like this do a thing where they take a real, useful insight (software defensibility is collapsing for undifferentiated products) and dress it up in apocalypse cosplay (APPLE IS DEAD, YOU HAVE 18 MONTHS, THE WINDOW IS CLOSING) because the apocalypse version goes viral and the calm version doesn’t.
The calm version is: the bar for “interesting software business” just got higher. If you’re building a thin wrapper on someone else’s API, yeah, you should be worried. If you’re building something with real distribution, real data, real community, or real hardware? You’re probably fine, or at least no more in danger than you were last year.
Apple isn’t dead. Your SaaS isn’t necessarily dead. But “things that were hard to build aren’t hard anymore” is a real shift, and pretending it isn’t is how you actually do end up writing a layoff post in 2027.
So: take the signal, leave the doom, and maybe don’t trust any business analysis that comes with a countdown timer attached.
Eighteen months from now we’ll see who was right. My bet? Apple still has a podium at WWDC and most of the SaaS companies you’ve heard of are still around. But the ones that didn’t take the moat question seriously? Yeah, those might be having a bad time.
That part, I do believe.
