Comprehensive coverage

The computer argues with the debating champions and ends up in a draw

At a press conference held last night (Tuesday), IBM presented two confrontations between the Project Debater robot that was managed and developed from the IBM laboratories in Haifa (see review) and human professional judges - one dealt with privacy and the question of the continued surveillance of countries on their citizens, and the other with genetic engineering - is it good or Bad?

Extended coverage on the science website:

Chaya Goldlist Eichler and Yaer Bach - two professional speakers during the debates with Project Debater, IBM's artificial intelligence system designed to compete in debate competitions. Photography: Or Kaplan
Chaya Goldlist Eichler and Yaer Bach - two professional speakers during the debates with Project Debater, IBM's artificial intelligence system designed to compete in debate competitions. Photography: Or Kaplan

At a press conference held last night (Tuesday), IBM presented two confrontations between the Project Debater robot that was managed and developed from the IBM laboratories in Haifa (see review) and human professional judges - one dealt with privacy and the question of the continued surveillance of countries on their citizens, and the other with genetic engineering - is it good or Bad? According to the rules of the field, each of the parties presented a position on the subject, preferably one based on the facts of one for and one against. This is according to Dr. Noam Salonim from the IBM research laboratory in Haifa who initiated the project and managed IBM Global to invest in it, this is the next big challenge after Deep Blue's victory over the world chess champion Garry Kasparov, and Watson's victory in the Jeppardy TV game in 2011.

debate "Supports people's right to privacy"

The Debate was conducted entirely in English, with the first issue that comes up for discussion at the demonstration event being: "The use of means of monitoring the public on a large scale must be ended." The topic was chosen at random, from a predefined list of topics - and was presented to the contestants, human and machine, about ten minutes before the start of the Debater.

Before the debate itself began, the journalists present at the event were asked to vote using a dedicated mobile application, which allowed them to express their position on the issue.

The Debater, who opened the debate with a four-minute speech, argued eloquently for an end to large-scale surveillance. She "supports the human right to privacy," and quotes the Declaration of Human Rights that supports this right. According to her, "Widespread surveillance endangers freedom of expression and contradicts the American Constitution. It prevents the free flow of information around the world. It is not justified, not required and has no place in modern society. It violates the Fourth Amendment to the US Constitution. The NSA's wiretapping program is unconstitutional. Listening on a large scale becomes a dangerous habit, and does not actually give greater security."

The human contestant, Yaar Bach, believes that "people give up their privacy on the Internet anyway - whether in computer games or in order to receive services they are interested in. Enough people are willing to give up their privacy in exchange for a small improvement in their comfort." For him, privacy is only important if it significantly affects the direct and immediate environment of humans. "Governments may use your private information against you - in totalitarian countries. But this is not true in democratic countries."

Awake you and the Project Debater during the heated debate. Photography: Or Kaplan
Awake you and the Project Debater during the heated debate. Photography: Or Kaplan

When the machine talks about the stench of a wilted bouquet of flowers

In the response round, again for another four minutes, the Debater opened by stating that "My opponent says that surveillance is a valuable matter. I present a different position... wide surveillance creates a lack of balance between governments and citizens. He is dangerous to democracy. It makes us more vulnerable to terror." In direct reference to the claim made by Bach, according to which the use of surveillance on a large scale is already a fact, the machine noted with humor that "adhering to the sestos quo is like guarding a wilted flower. We need to get rid of him before the whole room stinks."

Bach, in his second round speech, wondered "Why should a machine care when the room stinks?" He pointed out that surveillance measures deter some potential criminals - and help police activity. "In the absence of such surveillance, the police are forced to act on the basis of general statistics. This creates a cycle of crime where certain populations do not trust the police. The more we allow the police to focus on criminals and not on entire communities - the situation will be better."

The Debater's closing speech focused on summarizing the arguments he put forward, while the system reiterated and emphasized the claim of the illegality of the widespread use of surveillance means.

After the debate, the guests of the event were asked to vote again, and to report who they thought presented a better speech, who managed to expand and enrich their knowledge - and whether they changed their positions.

Almost half of the voters reported that the quality of the system's speech construction was equal to that of the human debater. Only a minority stated that the system enriched their knowledge on the subject. Each of the parties managed to convince one listener to change their positions.

The machine explains why it is right to use genetic engineering

The argument according to which "we must expand the use of genetic engineering" was the topic defined for the second Debate as part of the event held at the IBM research laboratories in Givatayim. This time, Chaya Goldlist Eichler, who has been competing in Debate for 15 years, appeared in front of the machine. "We don't make life easy for the debater," said Dr. Aya Sofer when she introduced the contestant.

Debater started the discussion on the domain definition. And he claimed that this is a reliable technology - more reliable than old alternatives. In a kind of humor, the system announced that it was an efficient and safe process - "just as I am safe as long as the electricity works".

In an eloquent speech and based on facts and citing numbers, the Debater states that "there is no scientific proof that genetic engineering is not safe. The FDA has confirmed the safety of consumption of genetically engineered products. Genetic engineering is used as a tool to develop better medicines." The system cites data according to which the majority of the public in the US supports the use of genetic engineering in a variety of fields, brings examples of the fight against grain pests through genetic engineering - and examples of drugs developed with genetic engineering and approved for use in the US. "Genetic engineering has the potential to help poor countries, and feed the world." The system claims that this is a safe and useful technology - which also allows for the development of new drugs. Finally, the Debater predicts that the human competitor will show reservations and fear of future risks. "Life is too short", she concludes.

Chaya Goldlist Eichler does focus on the risks, and mentions that those who developed the antibiotics did not foresee the emergence of drug-resistant superbugs. "I am not ready to accept the word of technological progress when it comes to the safety of technological development" - she stings the machine, and wonders what genetic engineering will do to human society. "The results of the changes may not be felt for decades - but they may lead to cancer, for example. It is forbidden to fill the world with dangerous food and dangerous products.

In his response speech, the Debater notes that "people are afraid of any new technology". It refers to the arguments put forward by the human competitor regarding the impact on the environment - and requires research proofs. According to him, "genetic engineering of crops makes it possible to produce larger crops, and is essential for a higher yield". The system cites examples from Canada and India, where it has been proven that genetic engineering enables better adaptation to climate change, and cites studies that support genetic engineering. "Nuclear power and genetic engineering are not as dangerous as the continued use of traditional fuels."

Goldlist - in her response round - wonders what will happen when the ecological environment changes due to genetic control of mosquitoes? "We are satisfied today - but it is difficult to know what the long-term results will be. Genetic engineering is much more than modern medicine. Genetic engineering of children before they are born carries a dangerous potential. Do I have the right to choose the color of my child's eyes or hair just to resemble my family? This is dangerous, because the company may say that there is an advantage to a certain person. We must take care of the next generation of humans - and especially for those who will not have the means for genetic engineering, to make their children more successful and more successful. It is a danger that changes the way human society looks and its future.

The Debater opens his closing speech with the statement that "My competitor is romantic, a little naive and conservative. People should use their strengths." The system repeats and summarizes the arguments it put forward, according to which genetic engineering helps in the fields of food and agriculture, medicines and medical treatment.

Goldlist signs off with the statement that "the dangers cannot be ignored. When we manipulate the food, the medicine and our body - we put our future in danger. Democratic society worked to protect itself - but genetic engineering is ushering us into an era where money will buy you unfair advantages."

The voting results at the end of the second Debate revealed that the listeners appreciated the Debater as the one who provided them with a better enrichment of knowledge on the subject. Even in the quality of the speech - there was almost equality. In the persuasion index - it turned out that more listeners moved to support the arguments presented by the human competitor.

In response to the question of how the discussion topics are determined, and how it is decided which of the parties will be required to present this or that position, Dr. Salonim explained the processes of selecting the discussion topics: in the first step, the system is allowed to run on the data base and locate those topics where there is material available for its use. Among the topics that she reports on her ability to deal with - topics that are of interest are selected. The Debate itself is conducted in the format of a parliamentary debate, where the arguments in favor are presented first.

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismat to prevent spam messages. Click here to learn how your response data is processed.