Attended the Singularity Salon #26

It was about Game AI this time.

The instructor was Yoichiro Miyake (Lead AI researcher of Square Enix technology promotion department)

By the way, I don't play games at all, so I was not interested in the game part. However, I participated because I was interested in AI. The talk was extremely interesting even for people who don't know about games.

Apparently in the game industry, there is a path where a student researches about game logic > get a job at game company. I hope there will be similar path made in the translation industry as well.

The most impressionable word was "Games can no longer be created by humans alone (needs AI)". I felt that the day where the same can be said for translation would come soon.


Above is translation of an article "＃２６シンギュラリティサロンに参加しました。" dated December 18, 2017
Translation by Hiroko Matsuda

TM + MT + AI

I always thought it would be cool to combine translation memory, machine translation, and artificial intelligence.

Maybe I waited about a year. A situation where we can really use this is coming true. Let's keep a close watch on the next move.



Above is translation of an article "TM + MT + AI" dated October 17, 2017
Translation by Hiroko Matsuda

Adaptive neural machine translation

It hereeeeeee! (ﾟ∀ﾟ)

The adaptive neural machine translation (ANMT) is here!!

Is this neural machine translation with artificial intelligence? I'll be investigating closely (if I can, I want to try it out as soon as possible).

Above is translation of an article "Adaptive neural machine translation" dated November 2, 2017

Translation by Hiroko Matsuda

Something that AI can't do: Picture story telling (Kamishibai)

On November 9th, I went to a Thursday salon event "Time in need of elevator-pitch--from a picture story teller: Finding hints of making your presentation have an impact with one phrase"  (presenter: Hajime Fujii) held at the Knowledge salon.

According to the presenter, presentation of paper story telling is still not possible for artificial intelligence. In the future, it might be possible to detect body temperature and facial expressions with a sensor to change the content of the picture story according to the detected values.

Something that was intriguing about the presentation was that picture story tellers change according to the children's facial expressions. When children lose their interest, they will walk away, so it's a harsh situation.

I found out that there are two types of jobs: jobs that can be replaced by AI and jobs that cannot be replaced by AI. The enemy of human kind is not only machine but also human itself (like emotions, interests, and concerns).

Maybe my next presentation will be done in a style of picture story telling.

Above is translation of an article "人工知能にできないこと～紙芝～" dated November 10, 2017
Translation by Hiroko Matsuda

Implementation of "challenge system" in the professional baseball

The following article was released in a Japanese sports magazine on November 13, 2017.

"It has been released, on November 12, that the 'challenge system' will be implemented from the next season in pro-baseball. From this, when there is an objection to the judge's ruling, a replay verification can be requested by the coach, which was only done when a referee determines the need to do so in a system similar to the major league baseball. The 'Japanese version challenge system' which specifies the number of requests that can be made will be officially decided and released at the executive committee on the 13th".

Apparently, video verification will be actively utilized in the professional baseball as well. I was actually watching the Japanese series game on TV that probably lead to this decision. "Out" was initially called by the judge, but the runner immediately claimed: "I was faster, I was not touched".

The judge might not be needed in the near future. If we put sensors on the batter, fielders, pitchers, catcher's uniform, shoes, gloves, bases, mounds and such, not only the runner's out/safe but also strikes and batter's out can be automatically determined. If the sensors are placed on the ground, out-field poles and also on the ball, it is possible to automatically determine home run or foul.

If this happens, judges will not be needed in the future. The member exchange before the game is done by both coaches hand by hand in presence of the judge. If the members are input on a terminal device, the official ceremony will not be necessary. The witnessing judge does not need to be a human. If a third party is necessary to be the witness, AIBO or a robot by Yahoo! will be sufficient.

Personally, I hate the scene where the coach intensely argues with the judge, so I love the idea of automatic determination. Plus, I think the fun of professional baseball will not change even if it is automated.

Above is translation of an article "プロ野球で「チャレンジ」導入始まる。" dated November 15, 2017

Translation by Hiroko Matsuda

翻訳はあと１０年で消える職業だと言われた根拠となる論文はこれだ。

ＡＩが発達すると翻訳者が失業するという翻訳業界を激震させた記事が出た。

オックスフォード大学が認定　あと10年で「消える職業」「なくなる仕事」

この記事である。１０年で消える根拠になったのが以下の論文である。

THE FUTURE OF EMPLOYMENT: HOW SUSCEPTIBLE ARE JOBS TO COMPUTERISATION?

Prepared by: Carl Benedikt Frey and Michael A. Osborne
Published on: September 17, 2013

「あと１０年で翻訳の仕事がなくなる」かどうかを理解するためには、しっかり読んでみる必要があると思います。

Translated> AI in schools 20170803

It seems that the teachers at school are attending seminars and studying sessions on the use of artificial intelligence (AI) to study how children should live in the future.

The research is proceeding to the extent where manufacturers are researching how they can utilize the AI.

In contrast, in the translation industry, there is no active opinion about the use of artificial intelligence (perhaps it’s just not out on the surface). In some cases, machine translation is even rejected…

I think it’s only a matter of time before AI and machine translation become industry standards. We need to prepare for the day when this happens.

Networking with AI related people 20170731

The long road to January 2019 has begun.

Today, I went to a special table in the knowledge salon to see if they can “introduce me to people who are in the field of AI and machine translation”.

I was told: “Please give us some time and we will get back to you”.

Thank you!

Can AI be useful for translation, or useful in project management?

Regarding machine translation, we’ll need to decide which machine translation to use and select internal server (not necessarily in-house server), write up a manual for post editing, and decide on the workflow…

Some people say that patent translation is suitable for machine translation because there are a lot of patterns and set phrases, while others say there are not many patterns or set phrases. I’m in the latter position. I think only about 3% in one specification is set phrases. Many clients will say that set phrases do not need to be translated anyway…

Machine translation will most likely be a standard technology in the industry in a few years. This may happen next year, or the year after next.