Why I Think the AI Bubble Will Not Burst

Last year, I bought an Nvidia GPU for self-hosting AI tools. I'm not a gamer, and I would've never bought this thing if it weren't for AI. I simply don't want to rely on big tech for tools that I would share my personal information with. After spinning up some LLMs and image generating tools, I realized the hardware consumes a lot of electricity, and even 12GB VRAM sometimes wasn't enough. On top of that, these AI programs also use a lot of system RAM. Now imagine being a big company like OpenAI or Grok with millions of users. How much GPU power and electricity would it take to keep their service running? That's why these companies are seeking lots of investment. I'm not saying I support these companies, but I don't think this is a bubble like the 2002 dot-com bubble or some crypto hype cycle. Unlike the dot-com bubble era, where regular people invested in many shady companies that essentially robbed them, these AI investors know what they're doing. I'm not seeing any kind of deception here. It's true that some new AI products are coming up, like the Rabbit R1 or the Friend.com necklace, that flopped badly, but most of the AI investments aren't going to these new experimental products.

Of course, AI-generated content is polluting the internet like never before. But the problem is that you still like or share these AI slop, making them go viral. The algorithmically driven content you see on social media practically regulates your emotions. You see rage-bait and get angry; then you see a cute cat video and forget about it. This is how social media's AI controls your thoughts and emotions, and nobody has ever truly rebelled against this digital drug. So I think AI slops are here to stay. Maybe people will become tired of them someday. I think we can make people socialize more in real life by making the internet a shittier place, so it's not a bad thing after all.