This is what Dr. Noam Lamelstrich-Later, the founding dean of the Sami Ofer School of Communication at the Herzliya Interdisciplinary Center, said, "The human journalist is not a dead man at all. The robot cannot replace him with interesting coverage. We, as humans, have an advantage Huge in terms of creativity, provided we understand where the world is going, it's not easy."
By Yossi Al-Toni, People and Computers
"Artificial intelligence has limitations and therefore human journalism will survive the age of artificial intelligence. At the same time, journalists will also be required to be aware of the changes they must go through in order to operate in the new world," said Dr. Noam Lamelstrich-Letter, the founding dean of the Sami Ofer School of Communication at the Interdisciplinary Center Herzliya
Dr. Lamelstrich-Latter was the key speaker in the tour held by the IBI Forum - Data Science from People and Computers at the Interdisciplinary Center. The meeting was opened by the leader of the forum, Moti Sadovsky, CEO of DCS Strategic Consulting for Information Systems.
Artificial intelligence, said Dr. Lemelstrich-Latter, "has limitations. Today there is a journalistic bot, a computer program that writes the journalistic narrative without human intervention. The bot can analyze a lot of information at lightning speed, with the ability to write a journalistic story tailored to the client's consumer profile and agenda."
"There are currently 12 commercial companies building algorithms that write such narratives, including Narrative Science in the United States, Tencent in China, and Yandex in Russia. The algorithms read, analyze and provide insights. They are learning machines, which correct themselves on the fly," said Lamelstrich-Later.
"When there are people, there are terrible noises"
He quoted Ray Kurzweil, the American Jewish futurist, who said that "in 2029 computers will be able to do all human work and better" and the historian Yuval Noah Harari, who said "Human beings will be worthless. Robots will do everything. People will consume drugs and entertainment."
Lemelstrich-Letter also quoted Prof. Daniel Kahneman, Nobel laureate in economics, who said "I see no reason to put limits on artificial intelligence. It is hard to imagine a situation where when there is enough data - there will be things left that only humans can do. When there are people, there are terrible noises." .
According to him, "Kahneman's big mistake is seeing noises as a disadvantage. I see them as an advantage, because they allow the brain to be creative at higher levels."
"IBM is a pioneer in the field of the artificial neuron, which simulates the natural one. They focused on language and analytical thinking, today they are making a revolution: they are building a neuro-synaptic brain that focuses on the senses. Their goal is to develop a brain, not only on the analytical side, but with senses and feelings," added Lamelstrich- Letter.
"But an artificial brain has algorithmic thinking, a combination of logical circuits of '0' and '1.' New. Albert Einstein said that 'I have never arrived at my discoveries through rational understanding,'" he said.
According to him, "There are three levels of artificial intelligence. The narrow one, the weak one, can only operate in context, from one defined field such as inventing drugs or identifying software. Its weakness is that it is impossible to transfer knowledge from one field to another. The second, general artificial intelligence combines insights from different fields With the ability to think abstractly, the third form is an artificial brain that is more creative than a human brain in every possible dimension, Super AI," explained Lamelstrich-Later.
He also said that "artificial intelligence has limitations. Despite all the progress in its development in the last 70 years - all the algorithms that have been developed do not go beyond narrow artificial intelligence. They fail to imitate the human brain, which can connect areas that narrow artificial intelligence cannot."
"Studies show that there is a direct connection between imagination and subconsciousness and creativity and innovation at a high level," noted Dr. Lemelstrich-Letter, citing cognitive science researcher Margaret Boden, according to which "one should not expect breakthrough innovation from artificial intelligence - it operates within the space of concepts defined for it." .
The activity of the journalist
Dr. Lemelstrich-Letter described the work of the journalist. His first activity, he said, "is collecting data. Not only he does this, government agencies and technology companies also do this - and in a much better way than the journalist. In the second step, we will analyze the data. Artificial intelligence helps journalists, analysis tools generate insights and enable predictive capabilities, for example by scanning social networks. Here, the role of the human journalist is very important, in that he knows how to pose the right questions."
"The third action is the construction of the narrative. The journalistic story is undergoing a revolution: the new journalistic story consists of a combination of platforms. The robot will analyze the information and write a draft. Journalism becomes immersive journalism (embedded, embedded), in which the consumer wants to be part of the experience. The new journalistic story consists of a diverse puzzle of Technologies, including the integration of virtual reality. Artificial intelligence is not able to cope with the integration of technologies, and it will not be there for years "Many. The artificial intelligence that has been developed to date is the narrow type of intelligence, which is unable to deal with the new journalistic story structure," he said.
Lamelstrich-Letter concluded by saying that "the journalist and every content creator in the new era must go beyond his traditional journalistic activity and employ more professional work tools and patterns of behavior. He must take advantage of new technologies; he must use other people who know how to take advantage of these technologies; he must ask questions that are important to democracy and human freedom In all these places - there will be no artificial intelligence and robots."
"The role of the human journalist in the new era is critical. He must tell the new story in a creative and interesting way. The human journalist is not a dead person at all. The robot cannot replace a human journalist with interesting coverage. We, as humans, have a huge advantage in terms of creativity - provided we understand where the world is heading. It's not easy," he said.
2 תגובות
The recent disappointments with Watson's performance in the field of medicine and in general with Watson's applications, show that it will probably be some time before Bina replaces the journalists, so I am obliged to qualify and hedge my words. For the next few years, the professor may be right.
In the first stage, Bina's technology may not be ready yet. In my opinion, these problems are challenges.
https://www.ynet.co.il/articles/0,7340,L-5331307,00.html#autoplay
What is said is about the opinion of the said in the above article only, and not as defamation, and K.A. is entitled to express his opinion differently from it and to find evidence contrary to the issue of intelligence.
The lecturer speaks from the thoughts of his heart. Today there is no software that can replace the person, at least not overtly, maybe in deepmind there is. In my opinion, either there isn't or there is at the level of a small child.
To say that it is impossible means that there is a limit to the human creativity of artificial intelligence. And there is no limit at all in any field.
As scientists understand the human brain, also through experiments in artificial intelligence, so the intelligence will advance.
In other professions such as deciphering an MRI expert, and playing 60 strategy games, the brain already works better than the best person in the field. The professor wants to say it will be a sad day if it happens and that is true, but as someone who works in the field it will happen.
They will be different from us, colder than us. Depends on where we direct the development. Examples of how intelligence research has deepened our understanding of what intelligence should be like. A few years ago, Dr. Iain Goodfellow wrote an article about Binah in which instead of one player there are 2 players who compete with each other. Playing against each other accelerates learning. The Binah created shortens the training time from a million learning examples to 10,000. Besides that, this Binah creates films in a field called styling where you choose the style for example Van Gogh and a regular film and the intelligence turns the film into a Van Gogh style film.
Since then, the field of multi-agent deep learning research has opened up. Another researcher, for example Dr. David Silver, came and said - the brain will compete in a game against itself in which it receives reward and punishment. This created software that achieves much higher performance achievements than deep learning by itself and is called reinforcement learning. The sad and amazing thing is that the size of human creativity is the size of the brain's progress And if you multiply by hundreds of thousands of research teams, the evolution there is fast.
Another example: it was clear that as more layered neural layers are added, the abstraction in internalization increases and with it the performance increases. But above 5 layers the learning entered saturation. From 1998 to 2012 the issue was stalled. In 2012, Alex Kritsevski together with Joshua Bengio from Toronto presented a network in which there are bypasses of a neural layer to its non-consecutive layers. As a result, today it is possible to connect 128 and more layers without saturation. It took 20 years, and then a revolution happened.