Alpha3031

6 ms·

61.
by Alpha3031 3mo ago
Previous changes to society were also made by people within society so I don't see how that follows.
62.
by Alpha3031 3mo ago
> escaped the desires of its human creators I wonder if we will ever solve the corporate-superorganism-alignment problem and why the AI alignment people rarely (as far as I can tell anyway, though I admit I'm not in the field) talk
63.
by Alpha3031 3mo ago
As we all know, the US is entirely free of geopolitical interests, which is why there was a US–Japan trade war in the late 80s.
64.
by Alpha3031 3mo ago
I was under the impression Axios did that kind of shit even before LLMs, though I admit I haven't refreshed my memory on older Axios articles before this statement.
65.
by Alpha3031 3mo ago
There's an Merriam-Webster article on it: https://www.merriam-webster.com/wordplay/lite-word-history
66.
by Alpha3031 3mo ago
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.
67.
by Alpha3031 3mo ago
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.
68.
by Alpha3031 3mo ago
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.
69.
by Alpha3031 3mo ago
Fairly sure "may" means something different to "is" but maybe I'm just not the target audience and am reading it wrong.
70.
by Alpha3031 3mo ago
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 o
71.
by Alpha3031 3mo ago
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 distillatio
72.
by Alpha3031 3mo ago
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 Te
73.
by Alpha3031 3mo ago
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 fa
74.
by Alpha3031 3mo ago
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 su
75.
by Alpha3031 3mo ago
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/
76.
by Alpha3031 3mo ago
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?
77.
by Alpha3031 3mo ago
+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 b
78.
by Alpha3031 3mo ago
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)?
79.
by Alpha3031 3mo ago
Which militaries are you thinking and how much do you anticipate them spending and/or being willing to try and justify?
80.
by Alpha3031 3mo ago
> 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&
81.
by Alpha3031 3mo ago
> 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
82.
by Alpha3031 3mo ago
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&
83.
by Alpha3031 3mo ago
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 wit
84.
by Alpha3031 3mo ago
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
85.
by Alpha3031 3mo ago
> 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 circu
86.
by Alpha3031 3mo ago
Why bother when they'll more likely than not run it again in a few months anyway?
87.
by Alpha3031 3mo ago
I really don't see the point of playing these word games. I had assumed you meant the government, yes.
88.
by Alpha3031 3mo ago
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&#
89.
by Alpha3031 3mo ago
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 t
90.
by Alpha3031 3mo ago
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.

More ›