geertj

9 ms·

61.
by geertj 2y ago
You must not have tried FSD 13.x with AI4 hardware. I commute to work every day from the suburbs to the city with a ~25 min one-way commute with zero disengagements. Edit: Elon mentioned in the last earnings call that if you are on AI3 hard
62.
by geertj 2y ago
Thank you for clarifying!
63.
by geertj 2y ago
I will go out on a (small) limb here and note that it really looks like the intention of this submission is is to associate a technical issue with DOGE and find a friendly audience to further a point. Normally a DNS issue of a government si
64.
by geertj 2y ago
> I’d be curious if this would just be a linear slowdown proportional to the RAM MHz or whether CAS latency plays into it as well. Per o3-mini, the blocked gemm (matrix multiply) operations have very good locality and therefore MT/s
65.
by geertj 2y ago
This runs the 671B model in Q4 quantization at 3.5-4.25 TPS for $2K on a single socket Epyc server motherboard using 512GB of RAM. This [1] X thread runs the 671B model in the original Q8 at 6-8 TPS for $6K using a dual socket Epyc server m
66.
by geertj 2y ago
It's okay to ask of course. My statement was that the ask does not have to be reciprocated. Personally, unless there's specific claims about data proving something, I usually assume claims on discussion sites to be somewhat inform
67.
by geertj 2y ago
Can I mathematically disprove that if I take the Lagrangian of the standard model and combine it with with Einstein's field equations, and chase down emergent complexity over many levels, that the properties that DEI selects for are ac
68.
by geertj 2y ago
Hiring will always have a subjective aspect to it, which is just the nature it. And I do agree that techniques like being conscious of bias and trying to interrupt it are useful. But yes, I do have an answer, which is that excluding factor
69.
by geertj 2y ago
> DEI can be implemented badly, but that doesn’t mean that DEI is bad I really doubt that DEI per your definition above can be implemented for a general benefit. Fundamentally you need to show that not hiring the best candidate is best.
70.
by geertj 2y ago
GP has no obligation to do a literature search for you. This is a discussion site, not a scientific journal, so it is okay to have opinions based on other sources, such as anecdotes from personal life, first principles, or common sense&#x2F
71.
by geertj 2y ago
I do! Great advice. Going off on a tangent, when I recently visited my Microcenter after a few years of not going there, it totally gave me 80s vibes and I loved it. Staff fit the "computer nerd" stereotype accurately, including j
72.
by geertj 2y ago
Any advice on how to buy the founders edition when it launches, possibly from folks who bought the 4090 FE last time around? I have a feeling there will be a lot of demand.
73.
by geertj 2y ago
Piling on, but I wanted to find a song I had once heard when I was a kid. I did not know the exact lyrics, just that it was woman talking about her husband leaving for space. I had tried to find the song using regular search engines multipl
74.
by geertj 2y ago
Why linear algebra? Honest question, i have not seen this as a specific requirement before.
75.
by geertj 2y ago
> I grab my popcorn bucket and eagerly await the Jepsen report about it I am the same, as distributed consensus is notoriously hard especially when it fronts distributed storage. However, it is not imposssible.. Hunter and I were both in
76.
by geertj 2y ago
I know this theory is being put forward but it never made sense to me as 1/ severance is at most a few months of salary, and forcing people out this way will also take months, during which time employees will be much less productive, a
77.
by geertj 2y ago
I went through a similar change from PM(-T in my case) to SDE, and I can +1 the parent. The best way to make a mid career change is to find a job where the old and new role can overlap, which I how I did it as well. If you're currently
78.
by geertj 2y ago
I imagine a future AI trained on this going into therapy to uncover childhood trauma.
79.
by geertj 2y ago
> Space tourism on the other hand is a different thing in my opinion. Space tourism is funding part of the R&D required through for new space to be able to do exploration.
80.
by geertj 2y ago
> Except you might be calculating the material, personnel, property costs also. The launch itself doesn't release that much, but everything leading up to the launch might. Yup, and the supply chain for all of those. At scale, that s
81.
by geertj 2y ago
> The whole rocket weighs 550 metric tons fully fueled on the pad, I am having a hard time coming up with a way for that to create 50x as much greenhouse gas. It is in fact impossible. The rocket carries all of its fuel and oxidizer, so
82.
by geertj 2y ago
> sorry if it was obvious to you already It was definitely not obvious to me! What do you make of the third chord under ‘composing’ that would not be playable in a diatonic (?) scale that was used at the time of King David, as the video
83.
by geertj 2y ago
Last month a guitarist named James Hargreaves posted a YouTube video [1] proposing a fascinating theory about the hidden chord that Cohen refers to in the first verse of Hallelujah (“Now I've heard there was a secret chord That David p
84.
by geertj 2y ago
Okay, thanks for that example. The ability of the pipe syntax to re-order the standard SQL pipeline order does indeed provide for more compact queries in this case.
85.
by geertj 2y ago
> It's not imperative. The pipe symbol is a relational operator that takes one table as input and produces one as output. Maybe I used the wrong term. In my mental model, the query planner decides the order in which the query is eva
86.
by geertj 2y ago
Understood. I should have asked my question a bit more specifically: what's wrong with CTEs that wouldn't be an issue with this new pipe syntax . I briefly scanned the paper and it appears there aren't any specific benefits
87.
by geertj 2y ago
Interesting, thanks for sharing. I did some digging on the linux-kernel mailing list and it seems that after the first patch series in Dec ‘21 and the LPC talk there was no further work on this. Too bad if true. Also I wasn’t able to easily
88.
by geertj 2y ago
What’s wrong with CTEs though? I have never thought of them as busywork and start with them if I know my query is going to be a multi-step process. I already thought them as “pipes” (in the abstract, not related to this paper). If the query
89.
by geertj 2y ago
This would work in this specific case where we know that there is a maximum rate at which work is produced. Arguably I was hijacking the thread to discuss a more general problem that I've been thinking about for a while. I have the sen
90.
by geertj 2y ago
> There isn't a significant advantage in having the kernel doing the polling, it would still be busy polling. I was thinking in terms of a generic syscall-less wake functionality where the kernel could do this for all processes in t

More ›