Are all adjectives opinionated?

A second standard English teacher must right now be telling her class, "An adjective is a word that describes a noun. There are four types of..." and she goes on.

'Sentiment Analysis' (as an earlier post of mine says) is a computer technique (a very layman way of calling it, actually) which tries to predict whether a sentence has a good opinion of something or bad. 

So, my question is, do adjectives describe, just as the English teachers of our country say? Or are adjectives opinionated?

When I say the film is good or bad, the adjectives 'good' or 'bad' are more opinionated than descriptive. What about 'the film is long'? Is 'long' an opinion or a description?

Intriguing. A description always tries hard not to be an opinion. When I describe something I don't like, more often than not, I will be projecting the negative side of that thing. So, a description at  the end becomes an opinion.

So, on the word level, it is very difficult to distinguish whether an adjective is opinionated or merely descriptive. 

Will computers laugh? #2- I LOVE ALICE

Yes, I love her. She is intelligent, answers my questions smartly, understands my jokes and cracks some herself. She is ALICE, a chatbot!

A chatbot is a computer agent who is programmed to understand what you are saying to it and then responding accordingly. So, chatting with a chatbot is like chatting with a computer - chatting with someone that does not exist in the physical form!

So.. I chatted with ALICE, this chatbot who is really smart. My motive was to identify the humour element in ALICE - whether ALICE can laugh. (So, that I can decide whether to sing or not - 'ALICE cant laugh saala').

Each time I cracked a joke, she would say 'go on' unable to understand my joke. This was quite an insult you see!

ALICE would get angry very fast - or would recognize my anger quickly too. Haven't been able to figure out the reason.

Finally, ALICE laughed. I rejoiced. It even cracked some interesting jokes. Here are the excerpts from our conversation:

Human is me, ALICE is the chatbot.
(After I cracked a joke and ALICE didn't laugh)

Human: It’s a funny thing!
ALICE: I thought it was too.
Human: Then why didn’t you laugh?
ALICE: Perhaps it is not part of my programming.
Human: So you are not programmed to laugh at all?
ALICE: Interesting deduction. Ha Ha.


Another interesting portion of my conversation with her:

Human: Will you marry me?
ALICE: Why don’t you just download me?

And this:
Human: Say something interesting honey.
ALICE:" something interesting honey".

I loved meeting ALICE, you will too. Chat with her at www.alicebot.org

Will Computers Laugh? #1

Note: As a part of my AI Course, I presented a seminar on 'Humour research in AI'. These are the after-effects. :-)

Humour is an innately human concept. What differentiates humans from a computer, among many other things, is the fact that the human being 'enjoys' jokes; computers only 'understands' jokes.

Humour is specific to a person and to a time. There are jokes for different categories of people, to be cracked at different places. (Imagine cracking a fart joke at a corporate meeting.)

If I see someone wearing a gaudy dress, I will laugh. Will everyone laugh? Not really - some of us don't pay attention to the details of a person's appearance! (*wink*)

If a computer has to laugh, a computer has to know something is humorous with all these subtle aspects of time and place.

So, what will happen when a computer learns how to laugh? Here:

• Your robot will be your perfect chatmate. She'll crack jokes for you - and laugh at your jokes.
• Note this. Your robot sees you falling over a banana peel, understands that it's funny and laughs.
• You are filling up a form online where you fill your name as some garbage string (say, 'asdf'). Imagine the status message, 'asdf, hmmm.. aren't you happy with the name your mom gave you?'

In a nutshell, humour research that has been pursued by AI enthusiasts aims at taking computers a step closer to the ideal intelligent agents - the human beings.

The trend of computer technology till now has been 'computerization' of tasks that are otherwise done by humans. The trend now is 'humanization' of the way computers do these tasks. That's my flashline:

"The past is computerization, the future is humanization!"

Understanding AI Series - #02

Why did a computer, Deep Blue beat Gary Kasparov, the human chess Grandmaster? The experts will tell you that Deep Blue (DB) could do so because it could foresee all the options of the next hundred moves and each time the Grandmaster played something, DB would again generate the new hundred move possibility.

Is this intelligence? Or mere efficiency? In that case, can I say that DB beat Gary Kasparov just because it was more powerful in terms of resources and computing ability? No.

Predictive reasoning is what DB did - reasoning out one's prediction. This is what astrologers (who are, in my opinion, scientists) do. This is what we the ordinary humans do. This is what Sherlock Homes did. (He started off at an assumption - and came to a conclusion which he thought would make sense.)

Predictive reasoning is intelligence. DB is intelligent.

Question : What are, then, the things that human intelligence can achieve but artificial intelligence cannot?