Modern life breaks sleep. The sleep industry profits by selling it back to us. Here’s why that paradox is driving almost $600B in market growth.
Modern life breaks sleep. The sleep industry profits by selling it back to us. Here’s why that paradox is driving almost $600B in market growth.
The gap between what AI can do and what humans are comfortable letting it do widens every single day.
Some actively seek difficult problems, most choose comfort. The difference isn't personality - it's relationship with work.
Success isn’t about big breakthroughs - it’s small, strategic decisions compounding over time. Build momentum with consistency.
Everyone's asking, “What can AI do?” The better question is: “What will humans let it do?”
Sam Altman was surprised by how quickly people got comfortable with ChatGPT. The CEO of OpenAI didn't expect humans to adapt this fast. That reveals something weird: even the leading minds building AI don't fully understand how humans adopt it.
Every technology follows the same pattern: capability first, comfort second. The timing gap tells the real story. Elevator operators kept their jobs for decades after automatic elevators were invented because people weren't comfortable riding alone in a moving box. Online shopping existed for years before people felt comfortable buying clothes without touching them first. We had driverless car technology in the 1980s, but we're still arguing about whether we trust them in 2025.
But AI is different. It's advancing faster than any technology in history, while also being more psychologically complex than anything we've ever adopted.
ChatGPT succeeded because it hit a comfort sweet spot. It feels like a good search engine that can hold a conversation. A smarter writing assistant that doesn't judge you. It helps you do things you already do, just better. This is the easiest type of AI to adopt – it makes your job feel easier. It doesn't require new mental models; it fits into existing ones.
The comfort came from the specific nature of conversational AI, not from humans getting better at adapting to new technology. We got lucky.
Here's where it gets weird: AI capability is doubling every 6-10 months. Human comfort with new technology has never moved that fast. It can't. Current AI models are growing 4-5x per year in training compute. By 2030, they could be trained with 5,000x more computing power than today. Whether that translates to dramatically better performance is anyone's guess. Even if we get better at building comfort – which we are – we're still falling behind exponentially.
The gap between what AI can do and what humans are comfortable letting it do widens every day.
We're about to exit the sweet spot. The next wave of AI won't feel like a tool. It will feel like intelligence. AI that can handle most knowledge work across departments. AI that can make strategic decisions. AI that can replace, not just assist.
The psychological leap from "AI helps me write emails" to "AI can do most of my job better than I can" is enormous. And unlike with ChatGPT, this leap won't feel comfortable.
The numbers tell the story. 78% of organizations use AI in at least one business function. 42% have it actively deployed. But only 26% report meaningful business value. Technical adoption is happening. Comfort adoption is lagging.
Most executives are making AI investment decisions based on capability metrics – accuracy, speed, features. But business outcomes depend on employee adoption, which is driven by comfort. The companies that win won't build the most capable AI. They'll solve the comfort problem while the capability gap keeps widening.
This changes how you should think about AI investments. Instead of asking "What's the best AI?" ask "What AI will our people actually use?" and as execs, “What will we let our people use it for?” While competitors compare benchmarks, smart executives focus on comfort metrics: How quickly can employees adapt to this AI? How does this AI make people feel about their work? How do we build organizational comfort with AI that's advancing every few months?
The comfort gap is widening. Most executives aren't comfortable admitting it.
Everyone's asking, “What can AI do?” The better question is: “What will humans let it do?”
Sam Altman was surprised by how quickly people got comfortable with ChatGPT. The CEO of OpenAI didn't expect humans to adapt this fast. That reveals something weird: even the leading minds building AI don't fully understand how humans adopt it.
Every technology follows the same pattern: capability first, comfort second. The timing gap tells the real story. Elevator operators kept their jobs for decades after automatic elevators were invented because people weren't comfortable riding alone in a moving box. Online shopping existed for years before people felt comfortable buying clothes without touching them first. We had driverless car technology in the 1980s, but we're still arguing about whether we trust them in 2025.
But AI is different. It's advancing faster than any technology in history, while also being more psychologically complex than anything we've ever adopted.
ChatGPT succeeded because it hit a comfort sweet spot. It feels like a good search engine that can hold a conversation. A smarter writing assistant that doesn't judge you. It helps you do things you already do, just better. This is the easiest type of AI to adopt – it makes your job feel easier. It doesn't require new mental models; it fits into existing ones.
The comfort came from the specific nature of conversational AI, not from humans getting better at adapting to new technology. We got lucky.
Here's where it gets weird: AI capability is doubling every 6-10 months. Human comfort with new technology has never moved that fast. It can't. Current AI models are growing 4-5x per year in training compute. By 2030, they could be trained with 5,000x more computing power than today. Whether that translates to dramatically better performance is anyone's guess. Even if we get better at building comfort – which we are – we're still falling behind exponentially.
The gap between what AI can do and what humans are comfortable letting it do widens every day.
We're about to exit the sweet spot. The next wave of AI won't feel like a tool. It will feel like intelligence. AI that can handle most knowledge work across departments. AI that can make strategic decisions. AI that can replace, not just assist.
The psychological leap from "AI helps me write emails" to "AI can do most of my job better than I can" is enormous. And unlike with ChatGPT, this leap won't feel comfortable.
The numbers tell the story. 78% of organizations use AI in at least one business function. 42% have it actively deployed. But only 26% report meaningful business value. Technical adoption is happening. Comfort adoption is lagging.
Most executives are making AI investment decisions based on capability metrics – accuracy, speed, features. But business outcomes depend on employee adoption, which is driven by comfort. The companies that win won't build the most capable AI. They'll solve the comfort problem while the capability gap keeps widening.
This changes how you should think about AI investments. Instead of asking "What's the best AI?" ask "What AI will our people actually use?" and as execs, “What will we let our people use it for?” While competitors compare benchmarks, smart executives focus on comfort metrics: How quickly can employees adapt to this AI? How does this AI make people feel about their work? How do we build organizational comfort with AI that's advancing every few months?
The comfort gap is widening. Most executives aren't comfortable admitting it.