ma2rten

48 ms·

511.
by ma2rten 7y ago
Is it possible using this or pygsheets to open a sheet that is editable by link without authorizing your account?
512.
by ma2rten 7y ago
Not if you into account the price of health insurance...
513.
by ma2rten 7y ago
They also have GPUs.
514.
by ma2rten 7y ago
Exercise and meditation can help with sleep, so I guess those two should come first?
515.
by ma2rten 7y ago
This is the Dunning-Kruger effect.
516.
by ma2rten 7y ago
I believe that some scholars actually believe that Pali is an artificial language, a mixture between different dialects that was intelligible to people from different regions. There is no question that there are some later additions. There
517.
by ma2rten 7y ago
We truly live in the future.
518.
by ma2rten 7y ago
Google Brain has several offices in Europe.
519.
by ma2rten 7y ago
Alexa and Google Home have them too.
520.
by ma2rten 7y ago
Exactly, this phenomenon is also known as unconscious bias.
521.
by ma2rten 7y ago
Here is the thing. There are probably innate differences between different men and women. There are probably also innate differences between different races. However, this doesn't say anything about individuals. A group of people might
522.
by ma2rten 7y ago
The HN title is inaccurate. The article itself actually says: The draft does not provide for an explicit obligation to provide identifiers in plain text. When providing information, however, "all company-internal data sources must be
523.
by ma2rten 7y ago
Neural networks also led to the AI winter. It wasn't until we had enough compute power that neural networks become popular again. Some people think that reasoning is the same. If we had a database of enough common sense facts there wou
524.
by ma2rten 7y ago
I don't think it works like that. The supposed copyright violation is ongoing.
525.
by ma2rten 7y ago
Instead of trying to crack the compression algorithm directly, it would make sense to disassemble the machine code and try to understand what it does.
526.
by ma2rten 7y ago
Running the compression algorithms is O(kn) where k is the number of compression algorithms. Taking the machine learning approach is O(n).
527.
by ma2rten 7y ago
Have you considered neural network based techniques for both problems?
528.
by ma2rten 7y ago
According to their comment on yesterday's post, they are doing an "advent calendar" of blog posts with one post each day.
529.
by ma2rten 7y ago
There is a browser extension called personal blocklist for this purpose.
530.
by ma2rten 7y ago
They mean auto complete.
531.
by ma2rten 7y ago
Here you go: http://www.peterbloem.nl/blog/transformers
532.
by ma2rten 7y ago
I have spent a few billion dollars and they were not successful.
533.
by ma2rten 7y ago
Google isn't like that. There is no CTO and even people in high levels can change teams.
534.
by ma2rten 7y ago
> I once had to spend 6 months reverse engineering a video output [...] this is the sort of shit you will do at an actual hard-tech startup I don't know about hard-tech, but it seems to me that most startups can't afford to hav
535.
by ma2rten 7y ago
I don't know. I feel that most FAANG employees are too risk averse to start a startup (which is why they work at a big company in the first place). YC actually gave the advise to work at a startup to find cofounders.
536.
by ma2rten 7y ago
Keep in mind that Uber's revenue was $3.6 billion in the last quarter, but they lost $1.2 billion. I wouldn't be surprised if their margin was negative. Really the problem is not that Uber gets paid too much, but that drivers are
537.
by ma2rten 7y ago
I doubt that their margin is much higher than that after factoring in discounts.
538.
by ma2rten 7y ago
Also Google Docs became more viable as a free (as in beer) alternative.
539.
by ma2rten 7y ago
Are you a man?
540.
by ma2rten 7y ago
You can't choose what you are interested in, but you can choose to pursue your interests.

More ›