nojs

20 ms·

331.
by nojs 2y ago
A case in point is the ubiquity of Pleco in the Chinese/English space. It’s a dictionary, not a translator, and pretty much every non-native speaker who learns or needs to speak Chinese uses it. It has no ML features and hasn’t changed
332.
by nojs 2y ago
A simple explanation may be ease of access to someone who can make the diagnosis
333.
by nojs 3y ago
> Maybe the experts are the problem? Or more likely, the current approach just doesn’t work very well. If everyone with a disease starts taking a medicine, and nobody gets better, it doesn’t mean the medicine causes the disease, despite
334.
by nojs 3y ago
Alternatively, he’s completely relaxed here because he knows what he’s doing is genuinely good and people will support it. That’s gotta be a lot less stressful than, say, a senate hearing.
335.
by nojs 3y ago
Yeah, the old correlation causation strikes again - unfortunately the study itself is paywalled [1]. Realistically how could you possibly prove this isn’t just a case of “people with better brains are less likely to get dementia and also in
336.
by nojs 3y ago
This matches my experience doing it with Elixir/OpenAI/ElevenLabs as well. Depending on the application it’s also possible to fire the whole thing off pre-emptively, and then use the early response unless later context explicitly
337.
Brachistochrone Curve(en.wikipedia.org)
1 points by nojs 3y ago|0 comments
338.
by nojs 3y ago
Why? If the risk of collapse of these countries is less than perfectly correlated with the US, having extra passports still helps.
339.
by nojs 3y ago
> San Junipero, the one positive Black Mirror episode with an actual Happily Ever After romantic ending You forgot Hang the DJ :)
340.
by nojs 3y ago
As devils advocate, maybe this is evidence such QA doesn’t matter as much as everyone thinks? It looks like no actual damage was done and the problem was fixed quickly once users noticed it.
341.
by nojs 3y ago
I mean I recently saw an Airbnb ad with something like pREMOVED instead of pagoda, so I guess it happens to everyone
342.
Smugglers disguise 146kg gold as machine parts(bbc.com)
1 points by nojs 3y ago|0 comments
343.
by nojs 3y ago
Here is a somewhat convincing argument why it can be considered a “fix” for MacOS [1] I have found this to be true with some fonts - they’re excessively thick on MacOS with the default subpixel antialiasing. 1. https://icomoon.io
344.
by nojs 3y ago
Ha, you're right about `font-smoothing` - it appears in the OP CSS like this but I guess they meant `font-smooth` [1] (which has a giant red warning akin to what you said). > As a generalisation, you should not do things like this,
345.
by nojs 3y ago
If anyone is wondering why the body text on this page looks so beautiful, it's this sneaky CSS property: font-smoothing: antialiased; -webkit-font-smoothing: antialiased; -moz-osx-font-smoothing: grayscale Seems like sites w
346.
by nojs 3y ago
If you want to go down the nowsecure.nl rabbithole (often used as a benchmark for passing bot detection) [1] is a good resource. It includes a few fixes that undetected-chromedriver doesn’t. 1. https://seleniumbase.io/help_d
347.
by nojs 3y ago
Thank you for sharing and I wish you good health. Rippetoe and Starting Strength can be brutal about injuries.
348.
by nojs 3y ago
Did you get an MRI or a diagnosis at the time?
349.
by nojs 3y ago
Random Matrix trivia: I recently discovered the “Adam Street bridge” where Trinity says “Because you have been down there Neo, you know that road, you know exactly where it ends” is the underpass right near central station in Sydney. So if
350.
by nojs 3y ago
Could you share more about what happened? How much were you lifting (was it close to your max) and do you feel like your form was correct?
351.
by nojs 3y ago
A fun thought experiment is to consider what would happen if you forgot to check in and the switch fired, and emailed all your loved ones the things you wanted to tell them - would your life get better or worse? Maybe this would function as
352.
by nojs 3y ago
Not necessarily, it might be better to stratify by genre (like buying a separate domain for different product verticals), to make the algorithm think each one is super relevant and specific.
353.
by nojs 3y ago
This is amazing, more detail about how it works: http://b3s23life.blogspot.com/2006_09_01_archive.html?m=1
354.
by nojs 3y ago
It’s not theoretically impossible, it’s just really hard. If we go fast enough, we can even do it within the current lifespan of a human (from the perspective of that human). https://en.m.wikipedia.org/wiki/File:Roundtr
355.
by nojs 3y ago
This thread, for example, is on the HN front page right now: https://old.reddit.com/r/Physics/comments/1bii6h4/if_gravity... Reading the comments, I would say there are plenty of intelligent users still
356.
Quenya(en.wikipedia.org)
1 points by nojs 3y ago|0 comments
357.
by nojs 3y ago
Not sure about Meta but a lot of companies do this to preserve the domain reputation of their main domain, since sending mass or automated emails may degrade it and lead to things being marked as spam.
358.
by nojs 3y ago
One advantage Google has here over everyone else is they can tie it to search crawling and say “sure, you can block Gemini but only by blocking GoogleBot”. That’s a tradeoff very few would make, whereas there’s not much to lose by blocking
359.
by nojs 3y ago
The best startups always look like toys initially.
360.
by nojs 3y ago
There’s increasing evidence you need the UV, eg https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5890609/

More ›