Using ChatGPT is not bad for the environment • Hacker News Board

Re: Using ChatGPT is not bad for the environment

#271

Post by Imustaskforhelp » Sat, Jan 18, 2025, 7:24 PM UTC

throwaway314155 wrote: post #57

irishloop wrote: post #44

Earlier quoted context omitted.

Meat and dairy specifically accounts for around 14.5% of global greenhouse gas emissions, according to the UN’s Food and Agricultural Organization (FAO). If people collectively just ate a bit less meat and dairy, it would go a long way. Don't even have to be perfect. Just show a little bit of restraint.

Right just a little bit of restraint. On an unprecedented scale of coordination by hundreds of millions to billions of people - a scale of cooperation that has probably never occurred in human history (and there's no reason to believe it will any time soon). But sure, if people "just" did a "little", it would go a long way. Just a _little_ restraint from the entire population all at once in perpetuity. No big deal.

Technically most hindus don't eat meat / maybe eats eggs rarely , So its definitely possible but it would probably have to be ingrained in the religion itself.

Another reason is that we had plenty of land and rice and wheat etc. just made more sense than eating meat I suppose.

I have never tasted meat , eggs sometimes but I am stopping that as well , not for hinduism but because I am so close to this ideal , might as well do it 100% lol

I am majorly convinced that there is no god but if I wish there was a god , I wish for Hindu deities (though I am obviously biased and this explains why people have such biases even being half atheists)

Re: Using ChatGPT is not bad for the environment

#272

Post by Retric » Sat, Jan 18, 2025, 7:25 PM UTC

andymasley wrote: post #266

Retric wrote: post #260

Earlier quoted context omitted.

On a logarithmic scale, it’s closer to the flight. Flying 1000 miles commercially only represents about 10 gallons of fuel.

10 gallons of fuel's worth of energy could be used to ask ChatGTP 100,000 questions (assuming 3 Wh per question) or power a digital clock (1-2 W) for 35 years. If you assume you ask ChatGPT 8 questions per day, it's using exactly as much energy as a digital clock. Personal use of ChatGPT is much closer to the clock.

The marginal difference of someone sitting or not sitting on a specific flight isn’t 10 gallons of fuel. It’s the capacity of the entire aircraft making that trip / number of passengers that gives 10 gallons on average because if more people make a trip eventually you need an extra aircraft.

So on an apples to apples comparison ChatGPT including training, hardware, etc uses a fuck of a lot more than just 3 Wh per question.

Re: Using ChatGPT is not bad for the environment

#273

Post by TulliusCicero » Sat, Jan 18, 2025, 7:26 PM UTC

lugu wrote: post #175

TulliusCicero wrote: post #151

Earlier quoted context omitted.

> I would agree with you, but Americans intentionally reinforce car dependence whenever it's discussed. They do, because their experience is that transit and biking really do suck and are useless. Which is accurate, for where they've lived. The problem is that you have to convince people that things could be better, when their lived experience is that it's always terrible.

I live in place known for its rainy weather, 15 km from work (because of the housing crisis). Being overweight, biking to work never crossed my mind for two years. Once I tried to commute during weekend, as a challenge. I realized a few things: - same duration as the train - it give me the exercise I needed - it relaxes me - it is free since I already have a bike Yes, it still take me 50 min to commute, but now I enj…

I've biked around in a bunch places in the US, and the reality is that it really is terrible. Bike infrastructure is F-tier almost everywhere, rising to D-Tier in supposedly bike friendly cities like Portland.

Bike infrastructure generally

* Is designed to be unsafe. Door zones are common, actual physical protection or segregation is rare, ESPECIALLY for intersections.

* Stops and starts randomly. Just look at Google maps for a city, you'll be able to see that the bike network is completely fragmented, with many bike lane suddenly disappearing on a road for no apparent reason.

* Randomly changes style/design principles even within the same city, so both you and drivers are constantly confused unless you're already used to a route.

* Is poorly enforced, with drivers routinely driving or parking in bike lanes with punishment being a rarity.

Now, some spots have good bike trails that work for their commute, and that's a great option when it's available. But I'm a bit tired of the "biking is actually okay in the US!" gaslighting.

Some people still manage, I do sometimes, but after getting hit by cars a couple times I tired of making excuses for it. There's a reason hardly anyone bikes for their commute in the states: overall, biking in the US is simply awful, and that's the truth.

Re: Using ChatGPT is not bad for the environment

#274

Post by andymasley » Sat, Jan 18, 2025, 7:26 PM UTC

blharr wrote: post #87

jonas21 wrote: post #29

Earlier quoted context omitted.

> However articles like this must link to references. There are links to sources for every piece of data in the article.

Where? One of the most crucial points "Training an AI model emits as much as 200 plane flights from New York to San Francisco" This seems to come from this blog https://icecat.com/blog/is-ai-truly-a-sustainable-choice/#:~... . which refers to this article https://www.technologyreview.com/2019/06/06/239031/training-... which is talking about models like *GPT-2, BERT, and ELMo* -- _5+ year old models_ at this point. Th…

I'm the author, the 200 flights number is taken from posts made by environmentalists and seem to match public numbers given (about 50 GWh) https://www.forbes.com/sites/arielcohen/2024/05/23/ai-is-pus....

If you think the numbers I used are wildly off I'd really appreciate any source saying so and I'll update the post with the correct amounts.

Re: Using ChatGPT is not bad for the environment

#275

Post by TulliusCicero » Sat, Jan 18, 2025, 7:28 PM UTC

ben_w wrote: post #201

TulliusCicero wrote: post #151

Earlier quoted context omitted.

> I would agree with you, but Americans intentionally reinforce car dependence whenever it's discussed. They do, because their experience is that transit and biking really do suck and are useless. Which is accurate, for where they've lived. The problem is that you have to convince people that things could be better, when their lived experience is that it's always terrible.

> The problem is that you have to convince people that things could be better, when their lived experience is that it's always terrible. Ironically, international air travel to places where it works great may help with this.

For some people yes.

But a lot of people are mentally stubborn, and seem to have inbuilt excuses of, "it wouldn't work here!" despite not spending even ten seconds thinking about it.

Re: Using ChatGPT is not bad for the environment

#276

Post by TulliusCicero » Sat, Jan 18, 2025, 7:31 PM UTC

Fricken wrote: post #198

TulliusCicero wrote: post #151

Earlier quoted context omitted.

> I would agree with you, but Americans intentionally reinforce car dependence whenever it's discussed. They do, because their experience is that transit and biking really do suck and are useless. Which is accurate, for where they've lived. The problem is that you have to convince people that things could be better, when their lived experience is that it's always terrible.

I ride a bike, and doing so has saved me about $450k in transportation costs over 3 decades. The effort it takes to earn $450k is something to include amongst the unpleasantries and pathologies associated with driving. Now, of course, I've had my whole life to set up my whole life the way I want it, and with a little foresight it really wasn't that difficult to set it up in a way that facilitates getting around on a…

I've ridden bikes for transportation in a handful of regions in the US. It is essentially always bad, it's just a matter of some people being able to look past or tolerate the awfulness.

Yes, some people are okay sharing a lane with cars, or using a strip of paint for protection, but accepting a poor status quo doesn't stop the status quo from being poor.

Occasionally there's a route that's legitimately good, almost always due to having an off street trail you can use for the bulk of the commute. Those can be great.

But once you hit bike lanes in an urban area, it's virtually always terrible. A lot of Americans are just used to the terribleness and don't notice it anymore, the same way they don't notice how almost every neighborhood is designed to be unwalkable (and how most are designed to be economically segregated besides).

Re: Using ChatGPT is not bad for the environment

#277

Post by andymasley » Sat, Jan 18, 2025, 7:31 PM UTC

Retric wrote: post #272

andymasley wrote: post #266

Earlier quoted context omitted.

10 gallons of fuel's worth of energy could be used to ask ChatGTP 100,000 questions (assuming 3 Wh per question) or power a digital clock (1-2 W) for 35 years. If you assume you ask ChatGPT 8 questions per day, it's using exactly as much energy as a digital clock. Personal use of ChatGPT is much closer to the clock.

The marginal difference of someone sitting or not sitting on a specific flight isn’t 10 gallons of fuel. It’s the capacity of the entire aircraft making that trip / number of passengers that gives 10 gallons on average because if more people make a trip eventually you need an extra aircraft. So on an apples to apples comparison ChatGPT including training, hardware, etc uses a fuck of a lot more than just 3 Wh per que…

The training cost adds a minuscule amount to the marginal cost of each search. The best sources I can find on training GPT-4 say it used 50 GWh of energy to train. GPT-4 is being used for a lot more than ChatGPT, but even if we assume that ChatGPT is all it's used for if we divide the total cost of energy by the total number of ChatGPT searches it increases the cost of each search by 10% (I go into this in the post) raising the marginal cost including training from 3 Wh to 3.1 Wh.

Re: Using ChatGPT is not bad for the environment

#278

Post by curvaturearth » Sat, Jan 18, 2025, 7:34 PM UTC

shafyy wrote: post #261

wmf wrote: post #68

Earlier quoted context omitted.

Voluntary conservation was only working by accident and guilt tripping never works. The grid needs to become clean so that we can have new industries.

Even if grid the was 100% renewable, this does not mean that there's no environmental cost to producing electricity. As a society, we need to decide what is important and try to minize energy consumption for things that are not important. And shoving LLMs into every nook and cranny of every application, so just tech giants who run the data centers can make more money and some middle managers get automatic summaries o…

Having LLMs everywhere haven't helped me much, it just gets in the way.

Re: Using ChatGPT is not bad for the environment

#279

Post by TulliusCicero » Sat, Jan 18, 2025, 7:38 PM UTC

AstralStorm wrote: post #215

lugu wrote: post #175

Earlier quoted context omitted.

I live in place known for its rainy weather, 15 km from work (because of the housing crisis). Being overweight, biking to work never crossed my mind for two years. Once I tried to commute during weekend, as a challenge. I realized a few things: - same duration as the train - it give me the exercise I needed - it relaxes me - it is free since I already have a bike Yes, it still take me 50 min to commute, but now I enj…

And if you need help and can afford it, ebikes exist. Some of them really cheap. Hopefully not skimping on safeties though.

Ebikes are really great and I often ride one to work. They do help with speed, range, motivation, and of course, hills.

They don't do anything about unsafe infrastructure though, which is by far the US' biggest problem for biking. To say that American bike infrastructure is garbage tier is an insult to garbage, that's how bad it typically is. It's extremely unsafe, and people can feel it, they can tell, hence why hardly anyone actually bikes to work.

Re: Using ChatGPT is not bad for the environment

#280

Post by hmottestad » Sat, Jan 18, 2025, 7:39 PM UTC

sshine wrote: post #223

hmottestad wrote: post #76

Earlier quoted context omitted.

Regular voting is usually what affects things such as the transportation infrastructure in your country or city. It’s a slow proceed though. Here in Oslo there has been a lot of investment in bike lanes, but just because one part of the local government builds more bike lanes doesn’t mean that other parts of the government will follow suit. Police still doesn’t care about cars illegally blocking the bike lanes. The p…

Speaking of Oslo and bicycles, I just want to add an amazing statistic that surfaced a couple of years ago here on Hacker News: Oslo has a zero pedestrian and bicycle mortality rate! https://thecityfix.com/blog/how-oslo-achieved-zero-pedestria... > In 2015, the political climate and public will in the City of Oslo changed the tone on accepting continued surface transportation fatalities. The mayor, city council and t…

We’ve had at least one cyclist killed since that article came out.