kevindamm

5 ms·

61.
by kevindamm 1y ago
I think it depends on the effort put into reading and understanding the code being generated. The article makes the assumption that extended use of LLMs leads to a shift of _not reviewing and validating the code_. It is pointed out as the
62.
by kevindamm 1y ago
Transformers can train models with much larger parameter sizes compared to other model architectures (with the same amount of compute and time), so it has an evident advantage in terms of being able to scale. Whether scaling the models up
63.
by kevindamm 1y ago
It's system dependent but Linux will generally speed up or slow down the time advancement until the delta from adjtime(...) matches up: https://linux.die.net/man/2/clock_gettime This clock is not affecte
64.
by kevindamm 1y ago
Satellites experience time dilation because of their orbital velocity and gravitational field being significantly different at their altitude. Without accounting for this, the clock drift would become unmanageable and Newtonian models are
65.
by kevindamm 1y ago
Ironically, those satellites would not be able to communicate effectively without the understanding of relativity that was obtained by looking at things light-years away.
66.
by kevindamm 1y ago
My tool of choice for what TFA describes is Vitepress. Markdown plus about a dozen useful plugins and understands Vue3 components and treats each *.md as an SFC. It hits a sweet spot for me.
67.
by kevindamm 1y ago
Preemptive multithreading is better than cooperative multithreading (which windows 3 used) but then it's de-fanged by allowing the threads and process to adjust their own priority and set arbitrary lower bounds on how much time gets al
68.
by kevindamm 1y ago
If you get into this position what you can do is `ls <tab-completed-path>` or other command to put the filename in the previous command's argument, then you can access it via !$ or !^ (or use !!:1 or your shell's notation fo
69.
by kevindamm 1y ago
Current search trends, and which results get clicked for which queries, are still intrinsically valuable. Mostly for the search index signals as you mention, and other things like updating recrawl rates, etc. Even for established players t
70.
by kevindamm 1y ago
I kind of agree and don't. You could say HTTP+DOM is the API, we're already there. But it lacks the structure and a more explicit regularity (in part because it's meant for human consumption, not programming). And if you w
71.
by kevindamm 1y ago
"make railways great again" is a little easier to pronounce as an initialism but you've got an internal rhyme in your version, so idk
72.
by kevindamm 1y ago
I think you could do it with many board games. Mouse Trap for monads? Poker for permutations? Dice for decision theory?
73.
by kevindamm 1y ago
I almost believed that was a thing, but I don't see any indication that it is. Not a bad idea IMO
74.
by kevindamm 1y ago
I think it was in "Manufacturing Consent" by Edward S. Herman and Noam Chomsky. https://en.wikipedia.org/wiki/Manufacturing_Consent#:~:text=... https://www.goodreads.com/book/show/12
75.
by kevindamm 1y ago
I think a lot of people underestimate the most valuable resource at universities: the access to specialized experts who love to talk with a curious and interested audience. Going through college by only kind of attending class and doing on
76.
by kevindamm 1y ago
Considering the video game industry contributes $66 billion to US GDP (and $184 billion globally) [1] some people would probably see it as an issue. [1] https://www.trade.gov/media-entertainment-video-games-sector
77.
by kevindamm 1y ago
That page also shows their results for quarter turns, there are 3 (or fewer) with 26 quarter turns being the shortest they can be solved in. 36 (or fewer) needing 25 quarter-turns. https://www.cube20.org/qtm/
78.
by kevindamm 1y ago
There are about 490 million of them, the full list can be downloaded from that same cube20 site https://www.cube20.org/distance20s
79.
by kevindamm 1y ago
Except that the transform is relatively easy to reverse (compared to prime factorization) because of the properties of edge and corner pieces and limitations on piece movement that make a kind of spectral analysis possible. Things get a li
80.
by kevindamm 1y ago
This was done in 2010 thanks to the analysis of the symmetries inherent in the underlying group theory, and about 35 CPU-years: https://www.cube20.org/
81.
by kevindamm 1y ago
You can put "... near <location>" at the end of your query to get nearby places. "... near me" also works
82.
by kevindamm 1y ago
I feel like this swings back and forth because neither monolithic nor microservice approaches completely solve the problem. You want some separation because it allows you to scale just the parts that need scaling, without paying the multip
83.
by kevindamm 1y ago
And then in Prometheus that scene is used to emphasize the behavior of an android becoming mis-aligned with the crew and not minding the consequences of subsequent actions.
84.
by kevindamm 1y ago
> Maybe it's not only responsibility, but the capability for one to understand the situation one is in and what is at their disposal. ..and a hint of 'don't be evil' that leads to good outcomes despite what everyone
85.
by kevindamm 1y ago
The more powerful the tool, the more responsible its wielder should behave. Ideally. In reality, that's impossible to enforce.
86.
by kevindamm 1y ago
The first thing this thread made me think of: https://cyc.com/ It's been around for decades.. I'm sure there are other similar systems.
87.
by kevindamm 1y ago
Wireguard is opaque about the independent streams in its connection. So, while they both can encapsulate multiple concurrent streams in one connection, QUIC can do things like mitigate Head-of-Line Blocking and manage encryption at the tra
88.
by kevindamm 1y ago
They compare cost, speed-to-market, schedule, and scalability, but it looks like they ignore failed launches and consider all missions successful? I couldn't find a comparison of the number of launch failures between the two, my recoll
89.
by kevindamm 1y ago
I was curious about this too, so I went looking. I found this video from about a year ago. He seems cautiously optimistic. The second half is a Q&A and he directly addresses this: in the presence of errors he finds more faults in the
90.
by kevindamm 1y ago
I'm just curious, I don't intend to harass anyone. Are you considering including the prompts in the published article?

More ›