Viewing profile — flumes_whims_

I'm sure Iraqis, Libyans, Iranians, Palestinians, etc. agree that putting Russia in the same sentence as America is stupid.

According to that 17% of boomers+ are dependent on their parents. Just looking at boomers, that is around 70 on average. How many people that age even have parents still alive, let…

Gell-Mann Amnesia? Interestingly enough, the term was coined by Crichton.

Elite dangerous?

Do any voice models support different conversation "threads" with different context?

The article mentioned the "the 12-year-old Queen of Sweden" which is quite an interesting story of its own. https://en.wikipedia.org/wiki/Christina,_Queen_of_Sweden

That can get thrown into a tiny 2" screen in the instrument cluster.

Meta overbought compute and is considering selling it like SpaceX did.

Members of parliament who owned stock in the East India company bailing it out by allowing it to sell tea directly to the American colonies?

Probably because benchmarks are leaked. Included in their training or model can cheat by finding answer online.

In the saga period of Iceland, 1/3rd of the laws had to be recited orally each year in a public assembly. If we had something like that then our legal code would be a lot slimmer.

How is Anthropic going to make sure every person using an API key is a citizen?

Products with Chinese made RAM can effectively only be sold in China due to US sanctions or threats thereof. It helps a bit because Chinese companies then buy less Korean RAM but n…

Let's money for scalpers and more money for the company making the product.

Linus Tech Tips built a computer with around twice the performance at the same price with current RAM prices. You lose out on size and SteamOS features like quick resume & CEC.

Are you including the emissions of producing two cars instead of one?

Tokens/seconds and watt-hours seem related?

What regulations did he as federal reserve chair rollback and which of them caused the 2008 crash?

Isn't Just Stop Oil funded by an oil heiress?

It's not just the weights. It is the system prompt, harness, safety filters, etc. Those can affect performance of the same underlying model significantly.

But they do generate zero power at night.

Wouldn't a local llm be just as vulnerable to this?

The overhead shrinks with larger models. It doesn't seem that bad. https://arxiv.org/pdf/2409.03992v2

How about a chatbot in front of feature requests that gets all the details the fist time?

> “AI detected bugs are pretty much by definition not secret, and treating them on some private list is a waste of time for everybody involved – and only makes that duplication wor…