今日は、今週一番の不安だった、システムシンキングのプレゼン。僕のグループのお題はサプライジング モデリング、要は人間同様にパソコンが物を考えたり、予測するシステムについてです。僕のパートは技術的な側面。
結構長い間準備をしました。まず発言原稿を作り、ネイティブにチェックしてもらいました（後ほど掲載）。そしてそれを自分なりに噛み砕き、発言用に変更。そしてプレゼン資料作り。もう吐き気がするような作業です。今週は毎晩一時間ほど、ストップウォッチを片手に4分以内に収めるように練習。
そして本番。ほとんど噛むことなく3分50秒程度で終了。全体では25分程度でしたが、我グループの評価はフィードバックによると、なかなか良かったです。何枚かに僕が良く練習したと思う、いい活躍だったと書いてあり、ほっとしました。
テストと違い失敗するとグループの評価が下がってしまうので、なんとか貢献できてよかったです。グループの中ではクリシュナ（インド人）が一番よかったと個人的には思いました。僕はまだ自分のことを伝えるのが精一杯で聞き手とコミュニケーションをとることはできません。このあたりが今後の課題です。
今日は美味しいビールが飲めそうです。
（以下がドラフトです）
Ｉ would like to talk about the technological aspect of surprise
modeling. Since before 1940, there has been much discussion in the
fields of philosophy, mathematics, logic, and psychology about how to
make a “machine that simulates human intellectual activity.” In the
first half of the 1940s, the computer was developed and a machine that
could one day accomplish that goal became a possibility. In 1956 this
field was named Artificial Intelligence (AI) by John McCarthy.
In order for this field to develop, there needed to be more development
in the processing power of computers and also the development of better
programming languages that were capable of coding the complexities
inherent in human intellectual activity. As this development has
occurred, the AI field has been able to develop as well.
Until 1990, the focus of the field was on imitating human activity.
The computers were programmed lacking applied skills and the ability to
learn. This level of AI is called Weak AI. One example of this type
of AI are the automatic translation services on the internet like
Babbelfish.
Since 1990, the research has focused on machines that can think or
anticipate similar to the way that humans can. This level of AI is
called Strong AI. However, this is still a new area of research
because there are many areas of the human brain and its cognitive
systems that science is still trying to understand. Also, there are
still limits presented by the processing power and programming
languages as well.
Currently, the AI field can create programs and machines that can
forecast things in the short term but anything longer than a few days
is beyond the current technology. One example of this technology is
SmartPhlow, which we will discuss later. Current attempts to develop
Smart AI that can predict better and further in the future continue,
but there needs to be more advancement in both the technical and
cognitive fields in order for further advancement to be possible.
、食えるじゃん・・・