March 22, 2026
Popcorn out: AI smackdown!
AI Is Garbage and a Bubble
Internet erupts over “AI is garbage” rant: bubble burst or bedroom boom
TLDR: A viral post claims modern AI is fake, kills productivity, and could spark a market crash. Commenters split between doomsday nods, home‑lab fans boasting “free” self‑hosted tools that boost work, and critics calling the post a rant—big implications for jobs, budgets, and bets on the future.
A scorched‑earth blog post declared today’s AI “artificial stupidity,” predicting it won’t replace jobs, will wreck productivity, and could nuke pensions in a market crash. The author even banned AI‑generated comments. Cue the comment‑section cage match. Team Doom cheered, with ronsor dropping “AI is going away, folks” and linking their own manifesto: AI is going away.
But DIY diehards clapped back. ljlolel quipped it’ll all run at home “on ClaudeVM,” pointing to a how‑to future: The future is ClaudeVM. donmb went further, saying the “below cost” narrative is wrong because many teams self‑host “free” models, paying mostly for hardware and power—and claiming measurable productivity gains. Meanwhile, the skeptics piled on, calling the post a “crazy… unsystematic rant,” and one zinger labeled it “aged like already spoiled milk.” Memes flew fast: tulip bubbles, Nigerian princes, and “popcorn ready” GIF energy as folks argued whether chatbots are sugar water or rocket fuel.
If you’re not deep in the weeds: AI chatbots (built on LLMs—text‑predicting programs) are either a dead‑end bubble or a home‑brew superpower, depending on who’s yelling. The only consensus? The stakes feel huge—for paychecks, portfolios, and pride—and the comments are where the real show is.
Key Points
- •The article argues that contemporary AI, especially LLMs, is unreliable and frequently produces errors that negate productivity gains.
- •It claims studies show verifying and correcting AI output can take longer than doing tasks manually, leading to net productivity loss.
- •The author contends the LLM approach has inherent error-rate limits that will not significantly improve with more data or compute.
- •The article warns of economic and financial risks from an AI-driven speculative bubble that could trigger market and institutional losses.
- •It asserts AI outputs can be steered by large-scale online content manipulation, making systems vulnerable to propaganda.