It's a large corporation. I'm certain someone asked that question and got an answer before starting producing Python tools. It's management's job to ask that question and get answers, you know.
Couldn't agree more. It's basically a page with some pictures in it, and everything in it loaded so late for me that initially I wondered why they left so many large empty spaces in the page.
This could work and be fast with tech from at least 20 years ago, probably even more. It's really incredible this is output from a company valued in trillions.
Indeed. Google is the worst at designing around lazy loading. Their UIs in Drive, Youtube Music, and others essentially become unusable once the list gets long. God forbid you don't have a super low latency connection straight to the data center. Even holding open a web socket to fetch "next pages" doesn't cut it in most of the real world. If you're gonna lazy load (which I admit sometimes does make sense) you need to aggressively fetch the next page. If I have 200 files in my Google Drive and they're sorted alphabetically, and I want one that starts with "y", the UX is so unbelievable bad that I sometimes wonder if I'm being pranked. I'll have to wait through a dozen "next page" loads that only load a screen worth of files at a time, and each pause makes me wait a second or two. That really adds up when I just want to scroll to the file. Scrolling through large playlists in Youtube Music is utterly painful nowadays too.
Please people, test your UI on low bandwidth connections, high latency connections, and both conditions together. It doesn't need to be perfect, but it doesn't have to be anywhere near this bad. HN performs way, way better than these modern javascript heavy lazy loading apps, and I think there's some important insights in that sitting right there for the taking.
Scrolling back in google chat is another horrendous experience. Trying to see a message you sent last week? Scrolling up is like “here’s 5 messages”… “here’s 4 messages”… “here’s another 5”. It’s like a college student wrote it.
Oh yes! Totally forgot about that one because I rarely use google chat, but that is insane. It really feels like somebody built an example app as a joke of how not to do it, and somebody else accidentally shipped it.
Especially when metadata is so cheap to send over. Even if you have literally thousands of files, there's no reason why they couldn't send it all in a single JSON at first page load. Gzipped it would be nothing.
If you hate how Google Drive works get excited about searching in the new Outlook. Nothing like being able to not sort your search results and only getting 200 at a time and having to page through them when the search returns anything useful at all.
> ...why? why would technology that specifically requires a lot of energy help "fight climate change"?
There's no good reason it would. Nevertheless, proponents of both AI and crypto have claimed multiple times it would. So I think it is quite fair to bring it up.
Only the one big pow coin matters tho. Turns out, PoS coins are worthless!
I also dislike energy frugality arguments. They come from a Luddite place. A civilization is defined by its energy usage. Let’s be advanced civilization.
what do you mean worthless? I'd say the opposite - Bitcoin does nothing... it just sits there looking pretty, this vs all newer blockchains that can actually run programs... we even have smart contracts running MIPS emulators for production use!
... get Bitcoin to do 3% of what ANY PoS blockchain could do.
How much of that 500 TWh would disappear if you round to zero everything that consumes 3GWh per day or less? And keep in mind you are calculating for a single app (ChatGPT), nor for an entire industry or tech. Don't look at "civil aviation" but at each airline in particular.
I suspect nearly all of it. You framed it such that it appears insignificant, but with this framing nothing is significant.
ChatGPT is probably the majority of LLM use today, so this is not the equivalent of claiming that neither your car nor mine nor anybody else's uses more than 3GWh per day and therefore cars don't use a significant amount of energy collectively.
Also, though this is not really relevant to your major incorrect point, airlines aren't part of civil aviation.
I guess I'd frame it another way: how many LLM conversations a month would you have to have to increase your energy consumption by 1% vs how many more hot showers
if you are actually worried about your ChatGPT energy usage, skip a hot shower or spend a few hours less playing Cyberpunk 2077 and a few more hours reading an old book.
I definitely think breaking out ChatGPT on its own is too fine-grained. At the very least, one should lump all LLM training and use, or perhaps all deep learning training and use together.
I suspect at a "sensible" breakdown (trying to avoid the "How long is my coastline?" problem), which is presumably something akin to Zipfian, the main uses will actually account for most energy usage.
If anyone can find a breakdown at the level of granularity that "all deep learning" would make sense, let me know: so far my Googling has led primarily to either detailed breakdowns of energy _sources_, or high-level breakdowns of use at the level of "industry", "agriculture", or "residential use".
Does anyone know what the consequences are? I have no idea exactly what it is that applies immediately.
I would guess that starting today Google and others should stop advertising as they currently do it, it being illegal. I doubt it's that simple, and even if it was, I am sure they will not simply stop. So what happens now?
Tracking has no legal basis, but it's still permitted with consent. The problem with IAB Europe (and other similar ad providers, as well as IAB's customers) is that IAB Europe didn't obtain consent; it tried to hide its tracking by using supposedly non-personal identifiers, which wouldn't necessitate consent, but the court ruled that these identifiers were actually PII. IAB also tried to weasel its way out of its responsibilities, but preventing that seems to have failed.
As a result, data collected through IAB about European customers was collected unlawfully, and third parties must delete that data. IAB also can't smuggle consent like this anymore, and needs to pay a fine that was handed down a few years ago.
I very much doubt ad companies will actually delete the illegally obtained data, but IAB and other companies in the cyberstalking industry this can be a problem, because they need to actually comply with the law.
So a lot of companies try to avoid a lot of people? How does that work out?
In my experience most companies work with a wide distribution of people. This "we avoid hiring people who have defects" reads as disconnected from reality. Nobody is perfect, and most companies are average and have average people.
There’s such an enormous divide between tailoring a pitch deck to the investor to highlight the things they want to see and outright fraud I don’t even know where to begin with it.
reply