I live in Saint-Petersburg, Russia, and I want to do something useful to AI-Alignment movement.
Several weeks ago I started to translate “Discussion with Eliezer Yudkowsky on AGI interventions”. Now it’s complete and linked to in couple thematic social network groups, and I currently translate old metaphorical post almost as a rest, after which I plan to follow the list from Rob Bensinger’s comment.
My English is not very good, I know a lot of people with better language proficiency, but it seems to me that I am on Pareto Frontier on “translation skills + alignment-related knowledge + motivation to translate”. Otherwise, I think, there would be more than one alignment-related translation on lesswrong.ru over the past two years!
I would like to know if this is helpful or if I should focus on something completely different.
I think this is pretty useful. I feel a bit awkward that I kinda have no idea how good the translations are, but I expect it’s worthwhile and potentially quite important.
Yes, but I think much more useful might be for someone to do this for Chinese.
(I found this post after your comment under my question) Some time ago I was reading posts translated by people on the Russian site lesswrong and regretted that I don’t know English, since not everything is translated, and there are discussions in the comments on the original site. But then, in an unrelated example, I noticed that Google Translate has been working TOO well lately. And now I read the original lesswrong through it and most of the time I don’t notice that initially everything was not written in Russian. And I regret that I didn’t start doing it sooner. In general, Google Translate copes with the task on average no worse than a person, but only does it instantly. So lately I’ve been worrying that it would be far more efficient to post Google translations than to do it yourself. Also, it seems like it would be a good illustration: in the last five years, Google Translate has reached such a level, are you really sure that AGI is not a threat in the next 5 years? The bottom line is that I suspect that if people are simply told that it is better to use a translator, then they will not actually believe and will not read the original LessWrong. From this point of view, it would be useful in this way to show people how well it works and ask them to just try with detailed instructions on what to do. It may also be worth creating a script that will publish automatically translated articles on the Russian version of the site, again with a note at the end that everything was done by a machine, here is a link to the Google translated page of the site, you can surf.