MauranKilom

10 ms·

61.
by MauranKilom 3y ago
Because it's most likely using a conditional move.
62.
by MauranKilom 3y ago
Given everything you said is true, under those assumptions 3% of those asteroids that we identify as being in said 2km box will hit earth, unless the forward simulation is wrong (implausible) or the measurement error distribution is subst
63.
by MauranKilom 3y ago
Then what does 3% mean? Surely it means "given the data we have, one in every 33 will hit". Since that empirically doesn't happen, it must be that "the data we have" has a very low prior probability of being real. I
64.
by MauranKilom 3y ago
I'm not sure I understand your point... The object mass does not impact its trajectory (unless it either touches our atmosphere or is so massive as to measurably change earth's orbit). The gravitational force earth exerts on the
65.
by MauranKilom 3y ago
This only works if there is nothing between "no impact" and "you die the same day as the impact". But we know that's not the case.
66.
by MauranKilom 3y ago
Unless one in every 33 asteroids that have 3% impact probability at some point in time actually impacts earth, there is clearly some unwarranted assumption in the error bar/distribution calculation. "The measurement data has noise
67.
by MauranKilom 3y ago
If you look closely, you can see that the model version is attached to each image, and the original blog post clearly states that asking the same model over multiple days is how they intend to track the variance. No bad logic there imo.
68.
by MauranKilom 3y ago
Yes, nothing about GPT4 changed today . But that's not the goal of the project (although I can't speak for the intentions of the submitter here). Currently there are two different GPT4 models represented in the samples, with quit
69.
by MauranKilom 3y ago
For reference, normal English writing has about 1 bit of entropy per character.
70.
by MauranKilom 3y ago
> The only reason you would want bit-reproducibility is because you haven't done the numerical analysis and have no clue how many digits of your "answer" to trust. I can confidently say that this is not the only good reaso
71.
by MauranKilom 3y ago
...or use fixed-point arithmetic. Which, if I understand correctly, is basically the go-to of modern multiplayer-enabled game engines.
72.
by MauranKilom 3y ago
(Note that it's the StackOverflow users voting on questions - the same ones that you are soliciting answers from. Mods have actual moderation tasks to deal with.)
73.
by MauranKilom 3y ago
CMIIW but the main thing to make acquisition times more reasonable is higher magnetic field strength. Which, leaving all the technical questions of achieving it aside, comes with other fun constraints like requiring heavy shielding for the
74.
by MauranKilom 3y ago
Tangent: To deny those bare-metal instructions with static analysis, you might also have to flat out deny certain sequences of instructions that, when jumped to "unaligned" would also form the forbidden instruction. That might bre
75.
by MauranKilom 3y ago
Not a frontend person, and unlikely to become one anytime soon, but maybe someone can shed some light into the downsides of Svelte? The article fails to mention any (it's apparently "win-win"). Presumably if Svelte were the b
76.
by MauranKilom 3y ago
You might be interested in this: https://en.wikipedia.org/wiki/Fechner_color
77.
by MauranKilom 3y ago
> So, how do you explain that subjects do not change in colour when you move closer to them or further away, or when you divert your gaze (cone cell spacing is far from uniform across the retina)? To be clear, this is not a rejection of
78.
by MauranKilom 3y ago
"Only 5% of users use this feature, so we will remove it to save development efforts." As seen in Firefox..
79.
by MauranKilom 3y ago
I vaguely remember that continuous processes are strongly preferred over batch processes. Especially for such things as "provide energy".
80.
by MauranKilom 3y ago
That question makes pretty little sense without fixing some constraints like "what other files exist" and "what language are you allowed to reproduce them". In a world where every single file is exactly "abcd",
81.
by MauranKilom 3y ago
"impregnate" is not wrong per-se, but it is weird in placing the emphasis on only one of the two people required. Usually you'd say something like "trying for a child", I believe.
82.
by MauranKilom 3y ago
But they don't losslessly recreate the training data.
83.
by MauranKilom 3y ago
Another interesting facet is that, according to some schools of thought, compression and AI are equivalent problems: The better you understand something, the less you need to memorize to reproduce it. https://en.wikipedia.org&#x2
84.
by MauranKilom 3y ago
> The error of moderators and SO here is to discard AI generated answers because some wrong (but sounding right) answers it can generate... when often AI answers are also correct and even at times out do what a single human would have fo
85.
by MauranKilom 3y ago
> I don't see anything wrong letting the AI generate an answer and edit the rough/wrong parts if necessary. And this is not what the previous moderation policy was trying to prevent. What it was trying to prevent is answers fro
86.
by MauranKilom 3y ago
The difference in how much effort it is to create a correct-looking/sounding answer without an LLM, and how little effort it is with e.g. ChatGPT. It's a force-multiplier on the side of people creating bad answers. Worse actually,
87.
by MauranKilom 3y ago
If the place is filled to the brim with GPT-generated nonsense, how do you plan on finding those answers?
88.
by MauranKilom 3y ago
...the mods are part of the volunteer community, and their role largely already is more about keeping the bad apples out than establishing whether a particular response is good or not. Where do you get the idea from that the mods are incapa
89.
by MauranKilom 3y ago
But this is a problem of scale. If I wrote a bot that posted 10k AI generated questions and answers per hour, you would surely agree that it would be unreasonable to require every answer to be human-reviewed before it can be removed - it wo
90.
by MauranKilom 3y ago
> they can't just be easily supplanted by free alternatives as can moderators. That's an incredibly naive take. Do you think there is an abundance of moderators just waiting to fill the ranks? Several SE sites are struggling

More ›