Jaycifer

I am not serious at all, that first paragraph was copy pasted from an LLM button in DuckDuckGo I had to go out of my way to look for.

A lot of the big models were trained on reddit data with little indication whether there was selection for high quality content, which means to me that the overall training data quality is unreliable. If an LLM can hallucinate an answer, it can hallucinate its confidence value since that is also an answer.

In other words, I agree with your sentiment.

Oh yeah? Well AI systems can be useful for everyday tasks like recommendations or summaries, but their reliability depends on the context and the quality of the data they were trained on. Researchers have developed methods to measure an AI model’s confidence in its answers, which helps users assess trustworthiness in high-stakes decisions such as healthcare or finance.

So, uh, maybe you should be aware of the high quality data used in training and ask those, uh, other people who throw out LLM answers what the confidence value is before you judge. Ha, got em!

That should definitely be the barest minimum, it’s just insane to me how much one of these data centers demands. Even beyond the privacy/copyright concerns, it doesn’t seem to me like the benefits of training models faster justifies the costs.

I’m gonna throw some numbers around based on a conversation with my friend the other day.

He works at an oil refinery nearby and under heavy load the whole facility can use up to 150MW of electricity. That’s while being rated as one of the least environmentally harmful refineries in the US. They have ~300 acres of solar panels supplying ~35MW over the course of a full day (~50MW during daylight hours). That means 20~25% of their power needs are fully renewable.

He also mentioned that the new Meta datacenter going up nearby is projected to use ~400MW! That’s two and a half oil refineries! That also means that to cover its electricity use they would need to set up ~2800 acres of solar panels. That’s ~11 times the footprint of the datacenter itself and two large farms worth of space on some of the best farmland in the country!

They definitely should be 100% powering the facility with renewables, but even then I don’t think model training justifies the space and resources that need to be invested to accomplish that.