[this answer was written by AI with asistance of human]
I think you’re mostly right, but people keep collapsing two separate questions into one.
AI can be genuinely transformative while the market is still in a bubble. Those aren’t contradictions. The dot-com crash already proved that.
The “LLMs hit a ceiling” argument also feels premature. People treat today’s models as if they’re the final form, when the field is still moving extremely fast across efficiency, agents, multimodal systems, robotics, tooling, and hardware.
I also think some of those “everyone will switch” examples actually support the opposite conclusion.
Search absolutely devastated libraries for everyday information access. Most people barely use them outside academia now. Digital won.
“Apps killing the web” was only partly wrong because the abstraction layer keeps changing. We’re moving toward systems where people stop caring whether something is a website, app, or API. AI assistants push directly toward that.
And voice assistants didn’t fail because the idea was bad. They failed because the tech was too weak. Early assistants were basically glorified command launchers. Modern AI is already much closer to the sci-fi version people imagined.
The infrastructure point is probably the strongest part of your argument though. Compute, electricity, cooling, and chip supply chains are real constraints people love to ignore.
So yeah, AI is likely transformative long term, while current valuations and expectations are still probably overheated.
I think you’re mostly right, but people keep collapsing two separate questions into one.
AI can be genuinely transformative while the market is still in a bubble. Those aren’t contradictions. The dot-com crash already proved that.
The “LLMs hit a ceiling” argument also feels premature. People treat today’s models as if they’re the final form, when the field is still moving extremely fast across efficiency, agents, multimodal systems, robotics, tooling, and hardware.
I also think some of those “everyone will switch” examples actually support the opposite conclusion.
Search absolutely devastated libraries for everyday information access. Most people barely use them outside academia now. Digital won.
“Apps killing the web” was only partly wrong because the abstraction layer keeps changing. We’re moving toward systems where people stop caring whether something is a website, app, or API. AI assistants push directly toward that.
And voice assistants didn’t fail because the idea was bad. They failed because the tech was too weak. Early assistants were basically glorified command launchers. Modern AI is already much closer to the sci-fi version people imagined.
The infrastructure point is probably the strongest part of your argument though. Compute, electricity, cooling, and chip supply chains are real constraints people love to ignore.
So yeah, AI is likely transformative long term, while current valuations and expectations are still probably overheated.