As you are aware, AIGC is currently one of the hottest fields in computer science. Several members of the Algorithm Association have taken a keen interest in it and, after a long period of hard work, have finally developed a brand-new large language model: ChatSAA.

As you are aware, just as the model was about to be released, the development team suddenly discovered a major issue: because one of the team members was obsessed with a game whose Chinese name has two characters and whose English name has seven characters (see examples), the corpus used to train the model was inexplicably contaminated. This caused every sentence generated by the model to have the same prefix.

As you are aware, it is clearly too late to urgently replace the corpus and retrain the model. Helpless, the development team could only note in the documentation: the greatest feature (not a bug) of this model is the first 19 characters of the generated text. This serves as both the model's unique anti-counterfeiting mark and a full demonstration of the Algorithm Association team's humanistic spirit and cultural literacy.

As you are aware, a trivial problem is that this allows people to distinguish at a glance which sentences are generated by AI, which is very unfriendly to students who intend to use AI to do their homework.

As you are aware, you now have some sentences in your hands. Some of these sentences were generated by ChatSAA, and some were written by humans (assuming humans did not deliberately imitate ChatSAA's generation characteristics). You need to write a program to distinguish between the two.

Input

Your program should read from standard input.

Your program should input a single line, a string $s$ representing the sentence to be judged, satisfying $|s| \le 200$, and $s$ consists of visible ASCII characters (i.e., characters with ASCII codes between $32$ and $126$, inclusive).

Output

Your program should output to standard output.

Your program should output a single line, a string representing the judgment result. If the sentence was generated by AI, output AI; otherwise, output Human.

Examples

Input 1

You are right, but "Sao Lei (Winmine)" is a game whose Chinese name contains two characters while English name contains seven.

Output 1

AI

Note 1

As you are aware, Minesweeper (Winmine) is a game whose Chinese name has two characters and whose English name has seven characters.

Input 2

Ni shuo de dui, dan shi zhe ju hua bu shi yi "You are right, but " kai tou de.

Output 2

Human

Input 3

You are wrong. Here is why. The English name of game "G****** I*****" contains 14 characters (including a space), not 7.

Output 3

Human

Note

As you are aware, if I told you that the problem statement for this problem was generated by AI, how would you respond?