Alpha3031

There's an Merriam-Webster article on it: https://www.merriam-webster.com/wordplay/lite-word-history

Well, if you're happy with around (as in within an order of magnitude or two of) 0.1 tokens per second... I believe that's around what people are getting when loading MoE weights from NVMe.

To be fair, the flagship Xiaomi, Vivo, Oppo etc are comparable in price to other flagship devices. You very much pay a premium for the large camera sensors they put in these devices, amongst other things.

I don't think the timeline is plausible for Moonshot to have done any distillation from Fable, which means any distillation training data would have been from Opus.

Fairly sure "may" means something different to "is" but maybe I'm just not the target audience and am reading it wrong.

Paging in experts is mostly reading so on the first order effects it would be fine. These might be second order effects from e.g. write caches needing to be flushed more often (and maybe swapping other applications, if you use a swap file or partition) but it probably wouldn't be too much of an issue.

Apple does have its own models for on-device use and the newest series has a MoE with expert activation swapped per prompt instead of per token which is interesting.

I'd assume it's probably somewhat based off Gemini distillation or something, and I'm not expecting them to release the weights this time at least, but it would definitely be interesting to see an open local model using a similar MoE whether that's at a similar size (20B A1–4B supposedly) or a larger model (something DeepSeek V4 Flash sized, or slightly smaller, would just barely fit on consumer hardware, and could be fun to see how it performs I think).

Referent of "the models are meaningfully different" reads as <top closed, top open> rather than <top closed, cheaper closed> to me, so I'm not sure why we'd be comparing Fable vs Opus/Sonnet or Sol vs Terra rather than the same against Kimi K3.

I am not interested in discussing "facts" with someone who has yet to acknowledge their motte-and-bailey, which for obvious reasons does not engender a reasonable expectation that such a discussion would be approached in good faith. If you wish to make a claim, you are free to explain your preferred definitions yourself. If you don't, there is nothing to debate.

No, because there are public projects that make sense at 3–4% discount rates that haven't been funded, so it would clearly make more sense to direct funding towards those projects first before trying to fund anything that requires a sub-1% discount rate.

More recent research suggests the cerebellum does actually have a significant role in cognition. See for example https://pmc.ncbi.nlm.nih.gov/articles/PMC4089997/ and https://www.sciencedirect.com/science/article/pii/S138904172....

How is the question of "how satellite backhaul is useful in urban areas" unrelated to your implicit thesis that satellite backhaul is useful in urban areas?

+1.4 °C on average is pretty serious warming, I don't know how you got the impression I was "skeptical about the reality of climate warming". I'm even willing to believe that AR6 WGI is more conservative than it should be on GMST and underestimating current levels, seeing as they have previously underestimated things like sea level rise, but I don't think I am out of line for actually asking to see those alternate estimates which put us potentially above +2 °C from the preindustrial reference before accepting such a conclusion.

Why in the world would any MNO or WISP choose satellite backhaul for urban areas when their next tower is presumably only a few kilometres over (as opposed to tens of kilometres in more regional areas)?

Which militaries are you thinking and how much do you anticipate them spending and/or being willing to try and justify?

I answered your comment because it was purely about gender

I was under the impression that it was about the demographics that are unresponsive to efforts to reduce meat consumption, but I suppose I could be mistaken.

I assume you're American

I am not.

So singling out men, or even conservatives, for a tiny difference feels like antagonisation with no basis

I consider the Boterman paper and other similar works (https://journals.sagepub.com/doi/abs/10.1177/136843021882440..., https://www.sciencedirect.com/science/article/pii/S019566632..., etc) to be sufficient basis. One does not need to eliminate meat entirely to significantly reduce it. Also, I consider right-wingers to be more reactionary than conservative.

but being bigoted is wrong in any way.

I think I'm ethically justified in bigoted towards people that want me to die, but whatever,

Many scientists consider +1.5℃ to have been reliably and consistently breached at some point in the very early 2010s.

Do you have a source for that? The 20 year average (not minimum) is considered reasonable to provide an estimate the GWL 10 years ago and is simple to calculate but obviously we'd only have the data going up to 2015 if using those.

I don't think there are any, even current-year GWL indicators that have us at close to +2℃ above the standard 1850-1900 reference. LOESS is probably the simplest one (the others are usually 10 years of observations + 10 years of projections) and that was ~1.31 [1.16 — 1.42] in 2023 according to https://climate.metoffice.cloud/current_warming.html and 1.44 ± 0.09 °C for 2025 according to https://berkeleyearth.org/global-temperature-report-for-2025...

Because gender conservatism is literally one of the leading factors predicting willingness to reduce meat consumption? (Boterman, 2026) Have you literally never seen what nonsense the alt-right churn out? It's everywhere, I really don't see how you could have missed it. Like, this is literally what real right-wingers say:

Conservative pundit Tucker Carlson, for example, produced a documentary in 2022 featuring the Raw Egg Nationalist, a prominent far-right influencer, who said that Impossible, Beyond and other plant-based companies are part of a “soy globalist” conspiracy to criminalize meat consumption and weaken citizens through poisoned food.

https://theconversation.com/plant-based-meat-alternatives-ar...

I'm sorry I'm bigoted against right-wing nonsense but considering they want to literally erase queer people from existence I would think I am at least a little justified in thinking badly of them.

Boterman, Willem Rogier (14 January 2026). "The political ideological dimensions of meat consumption". Political Science. 78(1): 73–99. doi: 10.1080/00323187.2025.2602446. https://www.tandfonline.com/doi/full/10.1080/00323187.2025.2...

True. Review on it from two years ago here: https://www.sciencedirect.com/science/article/pii/S221479932...

Might be worth shifting subsidies away from meat and towards plant protein to a level consistent with the social cost of carbon, but one could probably expect resistance from a certain demographic, maybe loudly exclaiming on social media that it's a secret government plot to emasculate them or something. (and opposition from industry groups as well probably)

I don't think having rich people pay more taxes relative to their income should be considered punitive. Progressive taxation has a long history going back to the first modern income taxes in the 18th or 19th centuries on the basis that low income people have a more reduced capacity to pay (due to non-discretionary spending on necessities being a higher portion of income when it is lower) and I think it's reasonable to expect it out of taxes.

Though, I do recognise that having one regressive tax wouldn't necessarily make the entire tax-and-transfers system regressive (or even more regressive, if that tax money is spent).

And the protein catches fire in contact with air.

FWIW neither lithium metal nor most Li-ion electrode chemistries autoignite in air at STP. The fire hazard is primarily due to heating up to the ignition temperature through short circuits though there are also exothermic reactions with e.g. water that can heat things up sufficiently.

Being the eager electron donor it is, once it is on fire it is very hard to put out, of course.

Why bother when they'll more likely than not run it again in a few months anyway?

I really don't see the point of playing these word games. I had assumed you meant the government, yes.

No, I just don't see any point in debating anything when you presumably agree that nobody is being bankrupted by free ambulance rides given that you called it a strawman. If you wish to defend what you called a strawman, which I didn't think people normally did in good faith but could be convinced otherwise, then I am willing to let you set the definitions you wish and make a case, as is standard for those instigating a discussion, while reserving the right to decline further participation if my disagreement with the position is below a certain threshold and/or any other reason I might find the discussion unproductive.

It's true that obstructions above the ground level would decrease the distance you can get line-of-sight to the next tower at the same tower height (a 40 m pole-style tower could have only 20 km LoS to an equally tall tower with 30 m tree cover vs 40 km for no tree cover, meaning 4× as many to cover the same area; effect is less if you use taller lattice towers at one end or both, of course) but it also decreases the range you can get out of the connection from tower to UE, not exactly to the same extent, but close enough. The change from GEO to LEO for satellite backhaul means that satellite is more technically competitive, sure, but microwave backhaul is possible, at least for basic mobile service (I would not want to use it for fixed wireless without even more tower density)... if you wanted to blanket even the mostly empty land between the two towers with good mobile coverage, which is way more towers than what most MNOs would want to purchase.

It's possible the general public wouldn't care enough one way or the other such that elected representatives could just do it without any change to their electoral fortunes.

Didn't know Perth was a regional area now, my mistake.

I refuse to play these games. If you wish to make the argument that universal healthcare (including cheap or free ambulance coverage at point of use) necessarily leads to financial ruin you are free to make that argument, along with all the necessary support such a position requires. If you do not wish to do so, then I have no issues and thus decline to make any argument otherwise.

I am not obligated to make your argument for you, and I am not inclined to calibrate your motte-and-bailey merely because you feel like interjecting your support for a position you would call a strawman when called out.

Would be nice for oil and gas companies to pay for all the emissions say, starting when they found out about it and decided to lie to the public. Maybe also bring charges against the PR firms they used since given those same PR firms worked for the tobacco industry clearly they won't stop until there are consequences.

Unrealistic, I know, but one can dream.

It was never necessary to run fibre to any specific individual tower. Microwave backhaul works fine of you have good coveragef the area. Less well if you don't. If cell providers wanted to put up enough towers that's a self solving problem.