July 31, 2020, ainerd
Artificial Intelligence Humor
Artificial Intelligence Humor
Humor is a common human trait that is difficult for engineers in machines to replicate. For example, we reported on an experiment in which robots competed against humans in a reading test and a live debate.
Computer scientists are also willing to teach robots technical skills that help them recognize, process, and respond to humor. AI experts say that attempts to make robots understand Humor often produce funny results rather than the opposite of what they should. Robots designed for Humor may struggle to understand when they are ‘making things funny’, an expert has warned, and could even lead to killing themselves.
Artificially intelligent machines that can’t capture context, timing and timing could have disastrous consequences for badly-timed jokes, according to a new study.
This could lead to automated software that kills someone or does funny things, according to a new study published in the journal Proceedings of the National Academy of Sciences.
There are good reasons why artificial intelligence should try to learn humor, says Miller of Darmstadt University. There is still a great deal of uncertainty about what machines can do and what people can do, but also about the limits of what they can do, “says Slonim. Humor could therefore be the key to the ultimate test of machine intelligence, which is to see if an independent assessor can spot the difference between a person and a computer they interact with, he says.
If we can get machines to understand sarcasm, he says, they will be related to it, but computers don’t get jokes (or metaphors, for that matter) because they have no expectation of undermining it. To see if computers can understand jokes, computers trained to solve crosswords or play chess can be tested to see if they can do the kind of things people do.
As a computing machine, the human brain remains uniquely efficient: according to some estimates, it can perform 38 trillion operations per second at 20 watts.
Mazlack and Taylor are working together, and then separately, to develop a program that can identify and even invent jokes. Rayz’s research involves trying to give artificial intelligence an understanding of what makes a joke a “joke,” looking at facets like delivery, context, and emotion.
Finally, it must have a single meaning used in wit, and it must be able to combine them and interpret the meaning of what is said. To understand, the computer must also understand the situation in which a joke is described. Artificial intelligence (AI) can do all this when it comes to telling jokes, but there’s still a long way to go.
Graeme Ritchie of Joking Computer tweeted a joke of the day from Edinburgh University: “A robot entering a bar joke is one of them.
Rayz spent 15 years getting a computer to understand humor, and the results were sometimes ridiculous. He has forgotten what he understands, but if you cross “fall” with “fall” in a dictionary, you get the same result: a pun with a different meaning. A computer can create and understand puns, the most basic of humor, based on various meanings of similar-sounding words.
When Rayz asked the computer why he thought it was a joke, his answer made sense: “Because it’s a pun on the word” fall. “
The reason so many scientists and scientists explore the fascinating world of computer humor is not just bizarre, no matter how enjoyable it must be to watch a machine strive to make a decent joke. It turns out that computers, while infinitely better, are not great at cracking jokes. Many argue that we are missing a fundamental part of ourselves in comedy – we are just funny. We look at a joke, we hear it and we look for it – but we’re just not funny about it.
Developers, scientists, product managers and scientists are working to make computers better in natural language processing and artificial intelligence in general. They must deal with natural language processing and teach computers to process, analyze and replicate the structure of language. Computational humor is the study of computer linguistics or artificial intelligence, which computers use for humor research.
Modeling humor is an important step in trying to model human thinking, but we don’t seem to be able to get computers to detect malignant tumors or prevent planes from crashing. Studies on the general scheme of information processing have shown that specific disturbances can be made dependent on a variety of factors, such as the type of information, the type of data and the context. They correspond exactly to the theory of dissolution of discrepancies and can detect humorous effects for psychological reasons.
There have been some researchers who have tried to make computers understand humor, but in vain, at least according to the current state of technology in this field.
Then, Mazlack said, you have to appreciate puns, understand rhymes and grays, and assess the audience’s ability to understand the joke. One thing to expect: don’t be surprised if you get a joke with lots of words but not a punchline or a pun.