Will AI Be Bigger Than The Internet?
Ask yourself why five times more people know about ChatGPT, have accounts, and know how to use it, but can't think of anything to do with it this week or next week. The key is understanding that AI needs to be wrapped in specific products and workflows to solve real problems, just like databases bec
1h 2mKey Takeaway
Ask yourself why five times more people know about ChatGPT, have accounts, and know how to use it, but can't think of anything to do with it this week or next week. The key is understanding that AI needs to be wrapped in specific products and workflows to solve real problems, just like databases became useful when packaged into industry-specific software solutions.
Episode Overview
Benedict Evans discusses AI as a platform shift comparable to the internet and smartphones, analyzing adoption patterns, investment bubbles, and the disconnect between AI's potential and actual usage patterns among consumers and enterprises.
Key Insights
AI follows historical platform shift patterns
Just like elevators went from having human operators to automatic buttons with 'electronic politeness,' AI will eventually become invisible infrastructure. The term AI only applies to new capabilities - once something's been around, it's no longer considered AI.
Physical limits of AI are unknown
Unlike previous technology shifts where we knew physical constraints (bandwidth limits, battery life), we don't understand the theoretical limits of AI capabilities. This creates unprecedented uncertainty in forecasting and investment decisions.
AI adoption has a bifurcated pattern
There are power users who spend hours daily with AI tools, and a much larger group who try it but can't find regular use cases. This suggests AI needs to be packaged into specific workflow solutions rather than general-purpose tools.
New vs. old use cases require different approaches
AI works well for tasks like code generation and marketing asset creation, but struggles with precise data entry. The bigger opportunity may be in enabling entirely new capabilities rather than replacing existing workflows.
Enterprise adoption requires product wrapping
Most companies need AI capabilities packaged into specific industry solutions with proper UX and workflows, similar to how databases spawned hundreds of specialized SaaS applications.
Notable Quotes
"Ask yourself why five times more people look at it, get it, know what it is, have an account, know how to use it, and can't think of anything to do with it this week or next week."
"The term AI is a little bit like the term technology. When something's been around any for a while, it's not AI anymore."
"We don't know the physical limits of this technology and so we don't know how much better it can get."
"Very new, very very big, very very exciting world changing things tend to lead bubbles."
Action Items
-
1
Identify your AI use case gaps
If you're not using AI regularly, examine why. Map your daily tasks against AI capabilities to find specific workflow improvements rather than trying to use general chatbots.
-
2
Focus on validation workflows
For AI outputs that need accuracy, design efficient validation processes. It's better to have AI generate many options for human selection than to manually verify every AI output.
-
3
Look for net-new capabilities
Instead of only replacing existing tasks, identify entirely new things you could do with AI that weren't possible before, similar to how mobile enabled new behaviors like ridesharing.
-
4
Package AI into specific solutions
If building AI products, wrap capabilities in industry-specific workflows with clear UX rather than expecting users to figure out general-purpose tools.