It’s not clear from the article, but do these statistics taken into account the amount of electricity and water required to train the model in addition to inference?
For example, the article says their daily average use of Claude code is similar to the dishwasher running. Is that just including inference or also training Opus 4.5?
This is a great question. To my understanding the industry consensus is that for the big three providers, energy spend on inference had already surpassed training by summer last year, and the former's share only keeps increasing. The problem is that there's no hard data in public.
What we need to do here is write an article that makes a wild claim in either direction ("99% is inference!"), post it on HN, and wait for the comments to roll in that prove it right or wrong.
I read The Economist, which doesn't cover sports at all.
It's mostly 1-2 page long articles for each story, blocked into categories (UK, Europe, US, The Americas, Asia, China, Business, Finance, Tech, Culture at the end).
> Television, he notes, has introduced the phrase "now this", which implies a complete absence of connection between the separate topics the phrase ostensibly connects.
This idea is why I always take media with a grain of salt. The decontexualization makes it easy for people to be reactive towards something, that isn’t logical
Eg “now this is why <insert person or group> is good/evil”
People call me the devils advocate when I point out these nuances but I just think we need to be much more critical when forming and holding opinions.
Isn't "now this" just a synonym for "moving on" or "next order of business" or "apropos of nothing"? I don't think the concept of jumping to a completely new topic is something TV introduced.
It’s been a bit since I’ve read Amusing Ourselves to Death but I believe in the book the phrase ”Now this” is used disparagingly to refer to the fact that with tv you can go from a horrific news story like a local family being murdered to a completely unrelated story, both in content and emotion in the span of seconds. This doesn’t allow ample time for the viewer to process the former and essentially forces them to turn their brain off as the cognitive dissonance of holding both stories (and more) simultaneously would be impossible.
That's fair. It does seem pretty similar to just reading a newspaper and moving your eyes to the next story, but I get that TV is a lot more stimulating and you can't go at your own pace like you can with the paper.
For example, the article says their daily average use of Claude code is similar to the dishwasher running. Is that just including inference or also training Opus 4.5?
reply