BB84

WIMP is only one model of dark matter. A favorite of particle physicists. But from a purely astrophysics point of view there is little reason to believe dark matter to have any interaction beyond gravity.

But it is a model we invented no? To explain the astrophysical and cosmological observations.

Among all those observations, a commonality is that it looks like there is something that behaves like matter (as opposed to vacuum or radiation) and interact mostly via gravity (as opposed to electromagnetically, etc.). That's why we invented dark matter.

The "it is unsuited" opinion in this meme is to poke at internet commentators who say that there must be an alternate explanation that does not involve new matter, because according to them all things must reflect light otherwise it would feel off.

Once you believe dark matter exists, you still need to come up with an explanation of what that matter actually is. That's a separate question.

(I'm not trying to make fun of people who study MOND or the like of that. just the people who non-constructively deny dark matter based on vibes.)

Particle physicists love the Weakly-Interacting Massive Particle dark matter model. But from a purely astrophysics point of view there is little reason to believe dark matter to have any interaction beyond gravity.

I'm still far from convinced about MOND. But I guess now I'm less confident in lambda CDM too -_-

I'm inclined to believe it's one or many of the potential explanations in your second link. But even then, those explanations are mostly postdictions so they hold less weight.

MOND is a wonderful way to explain rotation curves but since then with new observations (bullet cluster, gravitational lensing, ...) MOND doesn't really hold up.

I've heard of something similar that is able to predict an effect of dark matter (the rotation curves), but AFAIK it couldn't match other observations (bullet clusters, etc.) correctly.

Do you have a link for the model you're talking about. I'm curious.

Can someone explain why I am being downvoted and attacked in this thread? I swear I am not sealioning. Genuinely confused.

@sc_[email protected] asked how request frequency might impact cost per request. Batch inference is a reason (ask anyone in the self-hosted LLM community). I noted that this reason only applies at very small scale, probably much smaller than what ~~Open~~AI is operating at.

@[email protected] why did you say I am demanding someone disprove the assertion? Are you misunderstanding "I would be very very surprised if they couldn't fill [the optimal batch size] for any few-seconds window" to mean "I would be very very surprised if they are not profitable"?

The tweet I linked shows that good LLMs can be much cheaper. I am saying that ~~Open~~AI is very inefficient and thus economically "cooked", as the post title will have it. How does this make me FYGM? @[email protected]

What? I'm not doubting what he said. Just surprised. Look at this. I really hope Sam IPO his company so I can short it.

LLM inference can be batched, reducing the cost per request. If you have too few customers, you can't fill the optimal batch size.

That said, the optimal batch size on today's hardware is not big (<100). I would be very very surprised if they couldn't fill it for any few-seconds window.

Okay that sounds like the best one could get without self-hosting. Shame they don’t have the latest open-weight models, but I’ll try it out nonetheless.

Interesting. So they mix the requests between all DDG users before sending them to “underlying model providers”. The providers like OAI and Anthropic will likely log the requests, but mixing is still a big step forward. My question is what do they do with the open-weight models? Do they also use some external inference provider that may log the requests? Or does DDG control the inference process?

Stop depending on these proprietary LLMs. Go to [email protected].

There are open-source LLMs you can run on your own computer if you have a powerful GPU. Models like OLMo and Falcon are made by true non-profits and universities, and they reach GPT-3.5 level of capability.

There are also open-weight models that you can run locally and fine-tune to your liking (although these don’t have open-source training data or code). The best of these (Alibaba’s Qwen, Meta’s llama, Mistral, Deepseek, etc.) match and sometimes exceed GPT 4o capabilities.