Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
maksimum
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
31.
▲
by
maksimum
8y ago
> Facebook has no clue what kind of stuff gets posted on their platform, and their AI isn't powerful enough to detect it. Almost certainly Facebook's ML can identify the themes of almost all content on their platform with high
32.
▲
by
maksimum
8y ago
I also use this approach. To standardize appearance I put appearance modifiers in `notebook_context/__init__.py`, and then in my second jupyter cell from notebook_context import * configure_plotting_for_publication() Example n
33.
▲
by
maksimum
8y ago
I'd like to try out #3. "zah" has a nice ring to it. Do you guys minding not using it?
34.
▲
by
maksimum
8y ago
Not sure about physics terms, but in combinatorics I believe that an example of macrostate (also called an 'event') would be "2 coins out of 3 landed heads" and the possible microstates (also called 'outcomes')
35.
▲
by
maksimum
8y ago
> beta testing, and figure out discoverability I feel like these two are much harder than the others. How does the app store help you with beta testing? Isn't discoverability basically a lottery?
36.
▲
by
maksimum
8y ago
iphone/ipad area: 44% volume: 55%
37.
▲
by
maksimum
8y ago
Great clarification. You're right, reduce was a typo. I definitely like how method call notations can be read left-to-right rather than inside out. Now I'm curious why python developers decided not to add these 'transformatio
38.
▲
by
maksimum
8y ago
Good points, I'd appreciate if you could elaborate since it seems you've thought a lot about this. > Docker can help only if there is a single step in your project. In ML projects you usually have many steps - Preprocess, Train
39.
▲
by
maksimum
8y ago
Thanks! That looks equivalent to python without comprehensions. As far as I can tell the difference in the expressiveness of `map/filter/reduce` is mainly about whether you like or don't like comprehensions. Do you have any e
40.
▲
by
maksimum
8y ago
> if you generated output.data using input.data, then whenever input.data changes, DVC can detect that output.data needs to be regenerated as well To my understanding you could do the same with Docker. E.g. if you COPY your input files i
41.
▲
by
maksimum
8y ago
This is definitely needed and DVC has a few cool ideas. I think the most useful feature missing from existing tools is integrating data versioning with git, and simple commands to tag, push, and pull data files. I'm not sure I buy the
42.
▲
by
maksimum
8y ago
I think one reason people like comprehensions in python is because they provide an intuitive syntax for both map and select/filter. As I understand it, reduce still provides functionality that you can't get from comprehensions alo
43.
▲
by
maksimum
8y ago
> I am constantly running into things which I like just a bit less as I am trying it for side projects. Things just don't quite work the way I expect them to. How much of your like/dislike do you think comes down to familiarity
44.
▲
by
maksimum
8y ago
> I see a flaw in a lot of subscription models in that they lead to companies optimizing for "engagement" rather than maximizing the value of my time as a consumer This is really interesting. I suppose the question is how do yo
45.
▲
by
maksimum
8y ago
i7-3770: real 0m6.800s user 0m6.695s sys 0m0.034s 2x e5-2667 v2: real 0m2.217s user 0m56.088s sys 0m0.016s Seems like it's pretty inefficient with the dual CPU setup.
46.
▲
by
maksimum
8y ago
I think the extra work for a single developer to perform atomic commits is justifiable. How does this work with multiple developers working on the same repo? I'm assuming everyone should work on their own feature branch and send PRs on
47.
▲
by
maksimum
8y ago
I don't think CoL indeces are relevant because salaries are based on how much value a developer brings to the business. Suppose you have two jobs: $140k in NYC, and $100k in Dallas. If only one company can get an H1B, IMO it should cle
48.
▲
by
maksimum
8y ago
Based on the US's history, relatively free immigration has made the country as a whole very successful. It's not obvious that citizens should want protectionist immigration policy.
49.
▲
by
maksimum
8y ago
> Sure, but do the employees see any of that? Do Alibaba employees get generous equity grants? Vacation time? Cash bonuses? SV type salaries? Logically speaking, the employees are probably paid near their global market value. The convers
50.
▲
by
maksimum
8y ago
They mentioned that they retrained after the camera change and it was equivalent to the AIs that beat Mana 5-0 by their metrics.
51.
▲
by
maksimum
8y ago
> Or, what if we slow down the game, so that the human can actually pause the game each second and consider what to do next. That's basically what the computer is allowed to do While this would make it more fair, it would just make
52.
▲
by
maksimum
8y ago
$99 for a font... Seems like the artisinal font industry gets its business model from free-to-play, i.e. whale-hunting.
53.
▲
by
maksimum
8y ago
Are you including the time to build/link the tests? This is especially true if you have a bunch of dependencies. Last time I worked on C++ tests most of my time was spent on getting the tests to link quickly. Managed to get it from 25
54.
▲
by
maksimum
8y ago
At which point it's $26/mo if you go to see two movies. Hardly a deal even in LA or NYC.
55.
▲
by
maksimum
9y ago
> To get to a level of knowledge on anything to get that conviction though, requires a lot of deliberate reading and experience. Nah it's pretty easy, Dunning-Kruger's got us covered. :)
56.
▲
by
maksimum
9y ago
You could upload one yourself and you would get a notice if it gets monetized by the creator.
57.
▲
by
maksimum
9y ago
Is anyone aware of configurations that mimic i3wm using Phoenix/Hammerspoon or modifications to Amethyst?
58.
▲
by
maksimum
9y ago
// When we pretend that we're dead (pretend we're dead)
59.
▲
by
maksimum
9y ago
Is hyper-parameter optimization a cherry on top or one of the core value propositions? IME hyper-parameter optimization doesn't require much in terms of implementation effort (e.g. [1]), but requires compute. I would be surprised if a
60.
▲
by
maksimum
9y ago
Inb4 "your TV failed to turn on because of a timeout while trying to access manufacturer.com for checksum verification."
More ›