(THIS ARTICLE IS MACHINE TRANSLATED by Google from Norwegian)
"Before, they talked about the 'age of reason' and the like. Unfortunately, we live in the age of bullshit, "says philosophy professor John Searle. We are at the conference "Technology and the Human Future" under the auspices of Fritt Ord and The New York Review of Books. John Searle addresses the hype surrounding artificial intelligence – the idea that only machines can rain fast enough, then they will eventually become conscious. It is a well-known theme in the science fiction genre, but can it happen in reality? An entire emerging industry led by Google seems to mean it. "I give lectures to Google employees and others in the industry from time to time, and tell them that they are talking nonsense," Searle told Ny Tid.
Semantics is more than syntax. Searle is known for the Chinese space argument, which rejects the idea of what he calls "strong" artificial intelligence, that is, to build a human-like mind. The argument goes as follows: Suppose you sat in a room and entered Chinese symbols. Inside is a library that tells you what to do when you get specific symbols. You look up the right answers and feed them out again, but you still don't understand a split word Chinese. Searle explains: “A computer program only sticks symbols. It manipulates characters by pattern of a syntax. Human cognition has something more than a syntax, it has semantics. Syntax is not sufficient for semantics, for understanding, and therefore the computer program is not one sense. "
The so-called Turing test, launched by Alan Turing in 1950, is a widely used measure of machine intelligence. The test involves a judge communicating with a human and a machine using a keyboard. If the judge fails to distinguish who is who, the machine has passed the test. Turing predicted that the test would have passed by the year 2000, but so far no machines have come close. Professor of philosophy Luciano Floridi, who has often been a judge in such tests, gives us an idea of why: "If you ask questions like 'What can you do with a pair of shoes?' instead of questions like 'What is the capital of France?', the machine gets into trouble. A person would say, 'I don't know, you can use your shoe to keep a door open or you can nail something on the wall.' The machine will say 'I'm not a lexicon.' "
The ontology of machines. Searle also finds a deeper difference between man and machine by means of ontology – that is, what is about the being of the being. “If you look at the most important concepts in cognitive science, terms like perception, memory, ending, rationality, and even information and calculation, then all are systematically ambiguous. They can have an observer-dependent meaning and an observer-relative meaning. The information on a computer is observer-relative, it must be interpreted by humans, ”says Searle. Therefore, one cannot speak of a computer having information in an independent sense. Until it requires something a computer does not have, namely consciousness.
"I can say that my computer has better memory than previous computers, but that doesn't mean it's going through a Proustian experience of remembering its young days as a box in a Chinese factory," Searle elaborates. He continues: “What I'm trying to convey is that the intelligence of intelligent machines is absolutely zero. To think that a computer's enormous computing power gives it psychological relevance is a fallacy. ” Both Searle and Floridi agree that theoretically there is no obstacle for us to one day be able to build a thinking and conscious machine, but to achieve that we must first understand how our own brain creates consciousness. We are still far from achieving that.
Architecture of the Infosphere. "It's precisely because computers are not intelligent, because they are stupid, that I'm worried," Floridi told Ny Tid. Many of us tend to think of computers and networks as something else, something intangible. For Floridi, the sum of the technology we surround ourselves with is a new architecture that will shape us in very concrete ways. "This new information sphere is not just about communication or money, it is about the new living environment we are going to spend most of our lives in. Even today, most of our conscious time is spent on what is happening online. We have to start thinking strategically about how we design all this as an environment where we spend a lot of time, "says Floridi.
Part of the problem today, as Floridi sees it, is that the development is characterized by carelessness and ad hoc decisions. Computers are good at what they are designed to do. So they may be smart, but they are not intelligent. In order for smart robots to work, we need to organize the environment around them to make sure they can perform the tasks they are intended for. We are the interface that makes the robot successful, for example when we scan the items in the cash register at the store. When we adapt the environment around us to become robot-friendly, it is quickly made that it does not become human-friendly. For Floridi, it is essential that stupid computers be used to facilitate human intelligence – not the other way around.
"I can say that my computer has better memory than previous computers, but that doesn't mean it goes through a Proustian experience of remembering its young days as a box in a Chinese factory."
Technology and class differences. There are several glaring examples that the eagerness to implement smart technology has made life more difficult for people, and in some cases even reinforced existing class differences. Floridi tells of a city where it was decided that the parking lot at the train station should only accept a payment solution based on credit card via smartphone. This should make everything easier, but ended up creating major problems for students and others with low incomes who did not have credit cards. Older people without a smartphone could no longer park there either. Another city used a mobile app to detect holes in the roads, thus getting an overview of where it had to be paved again. Predictably, most holes were recorded in affluent areas with many smartphones, and the city ended up directing resources toward improving the good roads while ignoring the worst roads.
"These are examples that show a lack of clear planning," says Floridi. "Suddenly, the digital divide between those who are lucky enough to be on one side and those who are not becomes clear. This is the very concrete world we live in, and this is where politics can make a difference. Ny Tid asks Floridi if he can give an example of successful use of smart technology. "I'm thinking, for example, of third-generation bicycle sharing programs," he says. "At first they only put bicycles there in the hope that someone would take them, pay, and return them. Now you can actually check where the bikes are; you know which station has them while you are on the train. With smart use of resources, we see that people suddenly start cycling. It is good training, there will be fewer cars, it is economical and it provides a basis for a new business. Say that in the next generation you can order a bike in advance – then we are talking about technology that improves our lives. "
The last point is perhaps the point where Floridi thinks it is most sinful when new technology is to be implemented. For Floridi it does not hold that a smart technology is feasible, it must also be ecologically sustainable and socially acceptable, and not least it must create a desirable situation for man. "If we do this right, it will be a big step forward. We could really handle environmental problems with the help of smart machines. We could find solutions to inequality, health problems and housing shortages. Technology can be an incredible force for good, or it can be a total disaster, ”concludes Floridi.
