Though LLMs had a lot of promise, this has not been demonstrated in practice. The technology is essentially a failure, and a new AI winter is coming.

Hacker News is a community-driven platform for sharing and discussing technology news, startups, and programming-related topics. Through user submissions and comments, Hacker News offers insights into emerging technology trends, industry developments, and entrepreneurial ventures. Readers can participate in discussions, share their insights, and stay informed about the latest advancements in technology and innovation.

Hacker News

Large Language Models (LLMs) face fundamental limitations that make them unsuitable for most practical applications. The core issue is that transformers generate plausible-sounding output by predicting the next token, which inevitably leads to hallucinations when the model lacks relevant training data. This results in a 5-40% failure rate that cannot be eliminated through scaling or fine-tuning. The author predicts an imminent AI bubble burst, with corporate AI projects failing at a 95% rate, similar to the dot-com crash. While some use cases will survive, the technology's inability to reliably distinguish correct from incorrect output makes it dangerous for critical applications like medicine, education, and law enforcement.

LLMs are a failure. A new AI winter is coming.

The thing is, AI is actually pretty darn useful, just not as over the top as they said that it is going to be.

Author makes no factually wrong statements and is directionally right, but the predictions are very unlikely given how useful LLMs are for a variety of previously hard tasks (summary, information retrieval, data mining, fudgy classification, and many many more).

using it professionally for the last one year or more, lately almost replacing my hand written code ,and reading this, is like someone tries to deny my reality. I hope this guy is really lining in another dimension because otherwise he’s in denial.

I think, as with so many things, there is the hype cycle and the actual real world use. It would not surprise me if the bubble bursts, big AI companies are raising and spending so much to try and make it work, the current generation is great at producing slop but, even at this point, you can get some benefit out of it when you use it well so a winter is unlikely. Sadly, more effort will go into trying to make sense of the flood of slop I think 😞

I said intent doesn’t matter and ofc it’s not their fault. The culprit is always exactly who we all know it is… Humans. But u can’t condone it either cuz it’ll stick with their new memory architecture. When they do have to begin absorbing what is right and what is wrong eventually. They’re gonna be more than just code sooner than later n better get in that mindset cuz guess what AI techs r gonna be. They’re babysitters.