Hacker News new | past | comments | ask | show | jobs | submit | whatrusmoking's comments login

If this makes sense to you, you haven't actually talked to women in this field who trust you.

Scott has a bad habit of setting up bad strawmen and doing literature reviews without talking to domain experts.

(Frankly the weakness of the arguments he's choosing to debunk make me doubt that he even engaged in this topic with good faith. If you think the opposing view point is summarized by ' "And everyone says “Aha! I bet it’s because of negative stereotypes!” ', then you're not serious about this conversation.)


Feel free to get specific.


??? Google's been doing this at scale for a while.


Google switched Google Translate over just last year, November 2016. Considering Facebook is developing all this internally and the scale they operate at and the presumable hardware disadvantages (does FB have any equivalent of Google's TPUs?), switching over in production so quickly is impressive.


Only for very few language pairs, though. But for those translation massively improved, especially for German the old solution was basically unusable.


Google only supports Neural Machine Translation for a subset of languages. Whereas as Facebook is doing it for 2,000 translation directions.


45 languages * 44 other langues ~= 2000 translation directions. There are ~4,500 languages with more than 1,000 speakers ~4,500 * 4,499 ~= 20,254,500 language pairs so that's a very long way from every language.


Frankly though, I imagine the vast majority of those languages are unappealing targets in terms of return on investment.


and much smaller training data..


I guess they really don't need to cover all of them though as most of the people speaking a language with low number of speakers would be able to converse in other mainstream languages as well. I can speak four languages but I rarely converse in two of them as speakers of the two languages are usually comfortable with a mainstream language as well.


The most obvious thing to me is to use an intermediate language - I assume this has already been discussed extensively, anyone have references?

edit: I should have scrolled down https://techcrunch.com/2016/11/22/googles-ai-translation-too...


Even Swedish is barely in the top 100 with 8.7 million speakers.


Google may also have higher bar for translation quality..


Google was doing a statistical method of translation for a long time for Google Translate. Last year they switched over to Neural Machine Translation.


I figure this is a bit of an usual selection for HN, but here's the top of my in-progress list:

Moral Boundaries -- Joan Tronso

Known and Strange Things -- Teju Cole

The Feeling of What Happens -- Antonio Damasio

The Promise of Hope: New and Selected Poems -- Kofi Awoonor

And I just started Neil Gaiman's "Sandman".



Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: