Hacker News Reader

Comment by gilrain | Hacker News Reader

[flagged]

I think it's incredibly obvious how it connects to his "argument" - nothing he complains about is specific to GenAI. So dressing up his hatred of the technology in vague environmental concerns is laughably transparent.

He and everyone who agrees with his post simply don't like generative AI and don't actually care about "recyclable data centers" or the rape of the natural world. Those concerns are just cudgels to be wielded against a vague threatening enemy when convenient, and completely ignored when discussing the technologies they work on and like

You simply don't like any criticism of AI, as shown by your false assertions that Pike works at Google (he left), or the fact Google and others were trying to make their data centers emit less CO2 - and that effort is completely abandoned directly because of AI.

And you can't assert that AI is "revolutionary" and "a vague threat" at the same time. If it is the former, it can't be the latter. If it is the latter, it can't be the former.

> that effort is completely abandoned directly because of AI

That effort is completely abandoned because of the current US administration and POTUS a situation that big tech largely contributed to. It’s not AI that is responsible for the 180 zeitgeist change on environmental issues.

> It’s not AI that is responsible for the 180 zeitgeist change on environmental issues.

Yes, much like it's not the gun's fault when someone is killed by a gun. And, yet, it's pretty reasonable to want regulation around these tools that can be destructive in the wrong hands.

This is off topic, I’m talking about the environmental footprint of data centers. In the 2010s I remember when responding to RFPs I had to specify the carbon footprint of our servers. ESG was all the rage and every big tech company was trying to appear green. Fast forward to today where companies, investors, and obviously the administration are more than fine with data centers burning all the oil/gas/coal power that can be found.

Is it off topic?

What're the long term consequences of climate change? Do we even care anymore to your original point?

Don't get me wrong, this field is doing damage on a couple of fronts - but climate change is certainly one of them.

I don't consider it reasonable to want regulation for tools that are as of now as potentially destructive as free access to Google search.

I don't consider you reasonable if this is your best attempt at a strawman argument.

"You can't assert that AI is "revolutionary" and "a vague threat" at the same time""

Revolutions always came with vague (or concrete) threats as far as I know.

> And you can't assert that AI is "revolutionary" and "a vague threat" at the same time.

I never asserted that AI is either of those things

> nothing he complains about is specific to GenAI.

You mean except the bit about how GenAI included his work in its training data without credit or compensation?

Or did you disagree with the environmental point that you failed to keep reading?

I often find that when people start applying purity tests it’s mainly just to discredit any arguments they don’t like without having to make a case against the substance of the argument.

Assess the argument based on its merits. If you have to pick him apart with “he has no right to say it” that is not sufficient.

They did also "assess the argument on its merits" though?

“He just hates GenAI so everything is virtue signaling/a cudgel” is not an assessment. It’s simply dismissing him outright. If they were talking about the merits, they would actually debate whether or not the environmental concerns and such are valid. You can’t just say “you don’t like X so all critiques of X are not just wrong but also inauthentic by default.”

The part where they specifically address Pike's "argument" [0] is where they express that in their view, the energy use issue is a data center problem, not a generative AI one:

> nothing he complains about is specific to GenAI

(see also all their other scattered gesturings towards Google and their already existing data centers)

A lot can be said about this take, but claiming that it doesn't directly and specifically address Pike's "argument", I simply don't think is true.

I generally find that when (hyper?)focusing on fallacies and tropes, it's easy to lose sight of what the other person is actually trying to say. Just because people aren't debating in a quality manner, doesn't mean they don't have any points in there, even if those points are ultimately unsound or disagreeable.

Let's not mistake form for function. People aren't wrong because they get their debating wrong. They're wrong because they're wrong.

[0] in quotes, because I read a rant up there, not an argument - though I'm sure if we zoom way in, the lines blur

This thread is basically an appeal to authority fallacy so attacking the authority is fair game.

>appeal to authority

How so? He’s talking about what happened to him in the context of his professional expertise/contributions. It’s totally valid for him to talk about this subject. His experience, relevance, etc. are self apparent. No one is saying “because he’s an expert” to explain everything.

They literally (using AI) wrote him an email about his work and contributions. His expertise can’t be removed from the situation even if we want to.

having made Go amd parts pf Unix gives him no authority in the realms that his criticisms are aimed at though - environment science, civil engineering, resource management etc

not having a good spam filter is a kinda funny reason for somebody to have a crash out.

The "attack on the authority" is rather flat though.

> nothing he complains about is specific to GenAI

Except it definitely is, unless you want to ignore the bubble we're living in right now.

Someone else in the thread posted this article earlier.

https://nationalcentreforai.jiscinvolve.org/wp/2025/05/02/ar...

It seems video streaming, like Youtube which is owned by Google, uses much more energy than generative AI.

A topic for more in depth study to be sure. However:

1) video streaming has been around for a while and nobody, as far as I'm aware, has been talking about building multiple nuclear tractors to handle the energy needs

2) video needs a CPU and a hard drive. LLM needs a mountain of gpus.

3) I have concerns that the "national center for AI" might have some bias

I can find websites also talking about the earth being flat. I don't bother examining their contents because it just doesn't pass the smell test.

Although thanks for the challenge to my preexisting beliefs. I'll have to do some of my own calculations to see how things compare.

Those statistics include the viewing device in the energy usage for streaming energy usage, but not for GenAI. Unless you're exclusively using ChatGPT without a screen it's not a fair comparison.

The 0.077 kWh figure assumes 70% of users watching on a 50 inch TV. It goes down to 0.018 kWh if we assume 100% laptop viewing. And for cell phones the chart bar is so small I can't even click it to view the number.

And it’s fair assume much of the time watching streaming would instead have been spent on TV

> Unless you're exclusively using ChatGPT without a screen it's not a fair comparison.

Neither is comparing text output to streaming video

Compare generative AI video to streamed video, and generative text to streamed text etc. The differences are closer to an order of magnitude. The comparison to be made is the processing power required to deliver the content, not to display it.

This is based on assuming 5 questions a day. YouTube would be very power efficient as well if people only watched 5 seconds of video a day.

How many tokens do you use a day?

It would be less power efficient as some of the associated costs/resources happen per request and also benefit from scale.

Thankfully YouTube provides a lot more value to society than gen-AI.

This is a subjective value judgement and many disagree.

Doubtful. If you look at viewed content it’s probably 90% views from brainrot content.

To adults? Certainly. But keep in mind that many children are now growing up with this crap glued to their eyes from age 2:

https://www.youtube.com/results?search_query=funny+3d+animal...

(That's just one genre of brainrot I came across recently. I also had my front page flooded with monkey-themed AI slop because someone in my household watched animal documentaries. Thanks algorithm!)

Not for me.

It's not just about per-unit resource usage, but also about the total resource usage. If GenAI doubles our global resource usage, that matters.

I doubt Youtube is running on as many data centers as all Google GenAI projects are running (with GenAI probably greatly outnumbering Youtube - and the trend is also not in favor of GenAI).

Videos produce benefits (arguably much less now with the AI generated spam) that are difficult to reproduce with other less energy hungry ways. compare this with this message that it would have cost nothing to a human to type instead of going through the inference of AI not only wasting energy for something that could have been accomplished much easier but removing also the essence of the activity. No-One was actually thankful for that thankyou message.

I think that criticizing when it benefits the person criticizing, and absense of criticism when criticism would hurt the person criticizing, makes the argument less persuasive.

This isn't ad hom, it's a heuristic for weighting arguments. It doesn't prove whether an argument has merit or not, but if I have hundreds of arguments to think about, it helps organizing them.

It is the same energy as the "you criticize society, yet you participate in society" meme. Catching someone out on their "hypocrisy" when they hit a limit of what they'll tolerate is really a low-effort "gotcha".

And it probably isn't astroturf, way too many people just think this way.

being inside the machine doesn’t exempt you from tradeoff analysis, kind sir

As it so happens Rob Pike performed absolutely 0 tradeoff analysis

Do you really think that the only reason people would be turned off by this post by Rob Pike is that they are being paid by big AI?

No, which is why I didn’t say that. I do think astroturfing could explain the rapid parroting of extremely similar ad hominems, which is what I actually did imply.

Astroturfing means a company is paying people to comment. No one in this entire thread was paid to comment.

Buddy it's not astroturfing if people hate your favorite thing.

This is the most astro-turfy comment ITT

Reply on news.ycombinator.com