Ah, I’m bad at reading. Editing. Thanks!

Thank you for fleshing out your world model and theory. I think that this model falls short of a source (and contradict some other AI-pessimistic economics predictions; namely a crash in computing cost and in crypto), but could be developed into something I’d find compelling.

Let me brainstorm aloud about what I think this world model predicts that we might have data on…

Did we see a crash in ISP prices, home and industry internet use, domain hosting, or other computing services in the dotcom bubble? That situation seems extremely analagous; but my vibe was that several of these did not drop (ISP price I suspect was stable), and some of these saw a dip but stayed well above early-internet rates (domain hosting)? I feel like there’d be a good analogy here, but I’m struggling with a way to operationalize.

I mentioned a use for compute that your reply didn’t cover: crypto mining. Do we have evidence that the floor on crypto is well below datacenter operating costs (across exploitative coins as well)? I vaguely remember a headline in this direction. Another use case I don’t see drying up: cheating on essay assignments.

More broadly, this model predicts that all compute avenues are much lower payoff than datacenter operating costs. I think I’d need to see this checked against an exhaustive HPC application list. I know that weather forecasting uses up about as much compute as AI for some supercomputing clusters.

Governments have already issued rather large grants to AI-driven academic projects. I suspect many of these are orders of magnitude larger than the size of academic AI 6 years ago. (I’ll also quickly note that libraries are better than google search has ever been for finding true facts; yet google search has remained above library use throughout its existence.)

I know at least a few HPC folks talking about building their datacenters far north, where cooling can be done by opening a window.

? Massive GPU server racks are relatively easy to repurpose for several things. The most likely (if sad) is crypto mining, but there’s also expensive weather simulations, cloud gaming, video hosting, etc.

Requesting a source that these centers are hard to repurpose. I find myself pretty skeptical. Computers are generally multipurpose and easy to swap tasks on.

I do feel obligated to at least expose folks to numeracy, even if they won’t listen. Hope it’s not giving them a repeated trauma somehow.

Which feels to me like terrible policies still. Make big projects pay their costs please!

Growing inefficient cattle crops in a desert to preserve water rights: not necessary.

Flying Coast to coast for a business meeting that could be an email: not necessary.

Manufacturing those cheap scissors that break after 2 uses: should be a crime (not necessary).

All of these subcases have comparable emissions and externalities to the data centers (at least by my fermi estimates).

(I know it’s not the point, but a reminder that data center climate impact, including heat, is nowhere near flight, agricultural waste, or construction. Hate it for its own reasons, not for fake ones.)

Thank you!

Yes, TB was an example for a kind of error we make in morbid, culturally heavy places, I agree it is not a perfect analog.

The claim that: avoiding grieving a pet by cloning it is bad for your mental health.

I’m also interested in how it is bad, and how it compares to and with other treatment. I have the same gut instinct as you, I think, that pet cloning is not a good grief strategy. But I don’t have data, and wasn’t online much when pet cloning was a big topic. Cultures deal with death in a variety of ways, yet we have strong gut feelings for how grief should be done. I also find the idea of eating the recently dead pretty gross, for example, but this is a key step in the grief process for several cultures (and they seem to deal with grief fine).

Not all that long ago, tuberculosis was an incurable and slow killer. People thought it was the coolest death, that the pale complexion was beautiful, and that lying in bed slowly dying of TB was the best way to write poetry, discover truth, and understand philosophy. Humanity had a lot of cope around TB. Now we can eradicate it, and I think the romanticized view of TB looks pretty bad today. No sane person gets TB intentionally to write better.

Just checking: do we have a source on this? Or is this like accepting death by tuberculosis: we’ve romanticized a bad time.

Not your weights, not your wifu. Self host your dead relatives.

deleted by creator

I note the large investiment from Saudi Arabia UAE going towards Canada mentioned in the article; I’m not sure stopping fossil fuels is quite what he has in mind ='(

idk; when you talk to writers they complain about ideas being cheap and written works hard. I think for many concepts, there were poorly done versions well before something made literary history?

Though we’ll need to focus on different damnation for both, as this OP is at best a misunderstanding. See top replies.

Reminder to look at top comments, ideally at a bit of a delay. See this reply.

Thanks!

I’m told that the claim is false, see this reply by Petersson, which makes substantially more sense to me. I’m not on twitter, but allegedly many folks can translate Hebrew tweets.

If you could not because of rampant TOS violations then OP (or anyone else on twitter) could easily make a much better post, with much better evidence, that Hebrew speakers had gone rougue. We’d also see more than an LLM claiming this was the process, we’d also see people showing tweets they can’t translate (which did translate to something awful). If you or a friend are on twitter, please replicate the claim: find a hebrew tweet and post a screencap showing you can’t translate it.

I don’t understand this from a Grok, twitter, or Israeli perspective, and I think this is terrible evidence for the claim being made in the title/op.

If you wish to argue that the majority of Hebrew writers on twitter are nazi’s, surely you could scrape the hebrew language tweets, apply basic translation, and do some clustering. Even just doing this on a handful of highly circulating recent examples.

Twitter is not exactly opposed to nazi content recently, so I don’t understand the claimed policy.