You know what else takes far less energy than training a single model? One query. Yet, you argue that it’s the main contributor to the energy consumption. Why is that? It’s because there’s a very high volume of them, thus bringing up the total energy consumption. At the end of the day, it’s this total energy consumption that matters, not the cost of doing it once. Look at the total energy expenditure of training, not just the cost of doing it once.

So, it’s kind of weird t0 single AI energy use out here as some form of exceptional evil.

We’re talking about AI here because that’s the topic of this thread. I’ve never seen anyone say that it’s the only problem worth addressing. Plus, if you want to compare energy usage of ads (or anything else) compared to AI, you would first need to know how much energy AI is actually using.

Training is a continuous expenditure. We’re nearly ten years into this craze and we’re still continuously pumping out new models. Whether they’re trained from scratch or not is immaterial. Both processes still consume energy. If you want to justify the claim that training cost is negligible, you would have to show that this cost is actually going down over time and that it’s going down sufficiently quickly.

It doesn’t look like that energy consumption blog post account for the cost of training the model. Otherwise, it should be telling us how many queries/sessions are assumed to be run over the course of the lifetime of a model.

Those 20 years of eating directly serve our primary evolutionary goal that is the continued existence of human beings.

Most of us also have the goal of enjoying our time here. Food also contributes towards that.

I like to keep to the same routine when possible. Birthdays and holidays interrupt that. No good. I can’t do much for holidays, but since my birthday is supposed to be my day, i can demand this from everyone around me.

Our local Costco has these. They’re the previous day’s chicken and are sold at a discount.

The way it seems to work in Canada is that the government decides on a set of topics they want to fund that are fairly high level, and as long as your work falls in one of those categories, the grant gets approved. So the government doesn’t choose the specific drug to study. They choose which medical condition we want to try to treat, then they let the PIs tell them what they want to do and how it relates to those priorities.

The other potential explanation is that it’s because the speed task increases in difficulty as you get better at it. I’m hoping that’s the real reason.

How does this compare to other types of cognitive work? There are much more interesting things I’d rather spend my time on. It’d be nice to know if they have similar benefits.

1. Learning something new is winning.
2. Your perspective is irrelevant when it comes to facts. It only matters when it’s about your personal experience, and there are no sources you can cite to contradict those experiences.

Why use more syllables when few syllables do trick? It’s not like T-Rex is ever ambiguous.

With no additional context, if you said that “the balls in this glass are generally blue”, I would interpret that meaning every ball falls within the range of hues that can still be called blue by most people but may be questioned by a few. So 100% of the balls have to be “I can see why someone would call that blue”.

Obviously, only that tiny percentage of the population “care”. Not sure what they care about, but they care.

Financing is available for your $1 candy bars now.

That majority of people that are worse than AI slop are also not producing anything, so we’re not flooded with human slop. And when they do, we’re supportive because you need to practice producing bad stuff before you can start producing good stuff.

If humans produced slop at the same rate as AI, I guarantee you that everyone will be complaining about it just the same.

Rebates on used cars can come with the stipulation that the current owner has owned it for a certain number of years, and the amount can also be smaller.

Right, so that’s total investments that we want to reduce, not net.

Couldn’t you also theoretically hit net 0 while having 100% of our trades going through the US? I don’t know what the “lender” part of “net lender” means, so I could be wrong here, but it seems like this isn’t the right metric to look at if what you care about is trade diversification.

we can just

Implying it’s simple to get everyone to stop buying anything tied to the US, and everyone in the US to do the same with Canada.

7 Minutes

Average time Americans aged 15 and older spend reading per day, compared to over 7 hours spent on screens daily.

That’s kind of insane. What do people spend those 7h on if not reading? Is it all video content?