Hey {{first_name | there}},

Ninety-two percent.

That's how many AI startups will be dead by the end of 2026. And I don’t mean struggling. I mean DEAD. 

And I'm not sharing this to be dramatic. It’s because if you're building something in AI right now, you should know what’s going on.

People are literally quitting AI.

Not the technology. The news cycle.

Let’s talk about this.

I came across a Harvard study this week that found more than 14% of workers already experience "AI brain fry". Acute cognitive fatigue from constant AI interaction. 

And this isn't just workers. It's all of us.

Think about your own behavior. When's the last time you read an AI headline or article all the way through? 

When's the last time a new AI model announcement made you genuinely excited versus just... tired or “used-to-it-numb”?

This is what the industry is calling "AI fatigue". And it's not just a feeling – it's measurable. People are actively tuning out.

The irony? The same people building AI products are probably the ones experiencing this fatigue the most.

Let me connect some dots for you.

CB Insights and Gartner both peg the failure rate for AI startups at roughly 80% by the end of 2026. These are the startups that took an existing API, wrapped a pretty interface around it, and called it innovation.

Why are they failing?

Because they built features, not companies. They rode the hype wave but never solved a problem that required them specifically. 

When ChatGPT added the same feature, they died. When Claude improved, they died. When any of the big players decided "hey, let's do what that startup does," the startup became irrelevant overnight.

We're living through what one analyst called "the AI agent bubble popping". And it's happening faster than anyone predicted.

Roughly 3,800 AI startups shut down in 2025 alone, with another 1,800 in early 2026. The graveyard is growing by the day.

But here's where it gets interesting.

While the wrapper startups are dying and the news cycle is exhausting everyone, the builders who are actually going to survive are getting quieter.

Not in a "giving up" way. In a "finally focusing" way.

We're leaving the "magic trick" phase of AI. The phase where impressive demos mattered more than real outcomes. The phase where founders could raise money on nothing but a pitch deck and a "we're using AI" line.

2026 is when the industry hits a hard reset. And honestly? It's overdue.

Now here's the thread that ties everything together.

While we're arguing about AGI timelines and model capabilities, a whole generation is checking out. 

Gen Z is leading what's being called the "analog rebellion" – ditching smartphones, embracing physical books, picking up film cameras, carrying "analog bags" filled with notebooks instead of relying on digital everything.

Polaroid just launched a "Camera for an Analog Life" campaign. And it's resonating. Why? Because people are exhausted by AI slop, by content that feels generated, by interactions that feel automated.

The message is clear: authenticity is becoming rare. And rarity creates value.

If you're building in AI, here's my take on what's actually happening:

The noise is separating from the signal. The people who will win aren't the ones shouting loudest about their AI features. They're the ones solving problems that require human judgment, human connection, and human creativity.

The startups that will survive aren't the ones building the next chatbot wrapper. They're the ones doing things that AI can't replicate – deep domain expertise, relationship-driven sales, industries that require trust, problems that need a human in the loop.

And if you're tired of the AI news cycle yourself? That's actually a good sign. It means you see through the hype. The question is: are you building something real, or are you still chasing the wave?

I've been thinking about this a lot with my own projects. I've shifted from "what can AI do?" to "what should AI do?" and that question has saved me from building at least three features that would be dead by now.

The survivors in 2026 and beyond won't be the loudest. They'll be the most honest about what AI can and can't do.

Are you building for the hype or for the long game? Let's talk about it.

- Aashish

P.S. If this newsletter resonated with you, reply and tell me what you're building. I'm genuinely curious which of you are chasing hype and which are playing the long game. No judgment either way, but I'd love to learn.

80% of product professionals say AI helps them ship faster, but customers aren’t seeing value any sooner. Atlassian’s State of Product 2027 explores the gains AI is delivering and the challenges that remain, from decision-making that hasn’t kept pace to gut instinct overriding customer evidence.

Read the report