Hacker News Reader

Comment by ams6110 | Hacker News Reader

It's really not that remarkable, given what parts already exist. In pseudo command line:

   speech_to_text --language=english | google_translate --from=english --to=swedish | text_to_speech --language=swedish

All this product does is hook up a few pieces that Google had already developed.

View on HN

If you seriously believe that it took minimal engineering effort to integrate those parts into a low latency real time system I don't know what to tell you. These kinds of dismissive comments on HN never cease to amaze me.

Well, yeah, but without that google_translate part, they wouldn't get nowhere.

Bing Translate, even though it's been used by default on both Facebook and Twitter, will probably never give you good enough results to have a conversation with a stranger. Google Translate probably will.

I mean, if there's anyone that combine those couple of parts, it's Google and no one else.

It is remarkable because it happens in realtime.

Spoken language is actually very slow compared to what computers can handle. Google Translate would have no trouble keeping up.

Reply on news.ycombinator.com