Misconceptions About Artificial Intelligence

Given the promise that artificial intelligence holds for economic growth and societal development, it is important that policymakers not only pave the way for the development of artificial intelligence, but also actively support its further development and use.Over the last decade, significant advances have been made in computer science that enable software systems to process and compile new information to continually improve the way they work. This improved artificial intelligence is making computers an increasingly powerful and valuable complement to human capabilities: in medical diagnoses, weather forecasts, supply chain management, transportation and even personal holiday making (deciding which holiday choice suits your style).

Although artificial intelligence is becoming increasingly commonplace (most smartphones have some version of artificial intelligence, such as voice detection technology), the public still has difficulty understanding this technology. As a result, various critics, motivated by fear of technology, opportunism, or ignorance, have jumped into the “intellectual gap” to warn politicians. Artificial intelligence will cause a parade of disasters sooner than we think. For example; mass unemployment, the misuse of “algorithmic bias”, the end of privacy, the diminishing need for humans, even situations where machines decide that the Earth is a better place without us and want to destroy humanity (like Skynet). Indeed, these voices have become very loud, and the hungry media is eagerly supporting it, indicating that we are moving to a place where these narratives can be accepted as true. If AI becomes thoroughly demonized (in fact, engineering luminary Elon Musk has openly warned that AI could be a major threat to humanity’s existence if it does not stay as designed and falls out of human control), there will be a risk that politicians will try to block the development of AI.

This would be an unfortunate ending because the truth is that AI systems are no different than shovels or tractors. They are both tools of service for humans, and we can use them to further improve our lives. Given the promise of AI for economic growth and societal development, it is important for policymakers to provide support for the development and use of AI. The costs of not developing AI or developing it slowly will be huge: lower growth in per capita income, slower development in areas such as the environment and health, reduced improvement in various public and private goods services. This report highlights the priority of artificial intelligence and explains that five myths are unfounded. He says that rather than spreading fear about artificial intelligence, politicians should attach great importance to the development of artificial intelligence.

MYTH 1: ARTIFICIAL INTELLIGENCE WILL DESTROY MOST PROFESSIONS

Fact: Artificial intelligence, like previous technologies, increases productivity and has no effect on the average number of jobs or the unemployment rate.

As many people say, since artificial intelligence is a great power in terms of productivity, it will destroy jobs faster than the economy can recover and create a lower class of unemployed, which will make the “machine owners” the elite class. These are not new predictions and are as wrong today as they were in years past.

Apocalyptic views that AI will kill jobs suffer from two flaws. The first is that they greatly exaggerate the capabilities of AI to replace humans. In fact, this is quite difficult for technology, whether it is artificial intelligence or something else, the fact is that the rate of development of US production efficiency has been decreasing historically. It is very difficult to automate a lot of tasks with AI because in reality AI is currently “Limited AI” programmed to do one thing very well. In other words, the introduction of artificial intelligence in many professions will not lead to job loss, but instead can increase quality, product and innovation.

Second, even if AI were capable of doing these things, there would still be a lot of job opportunities because if a company reduces producing products at high efficiency, the cost will also decrease. These savings translate into lower costs and higher salaries. This allows more money to flow into the economy. Therefore, historically, there has been a negative relationship between productivity and unemployment rates.

MYTH 2: ARTIFICIAL INTELLIGENCE WILL MAKE US STUPID

Fact: Artificial intelligence will help us make more rational decisions.

Beyond misplaced fears that these smart machines will take our jobs, some dystopians claim that AI will make us completely dependent on machines, turning us into helpless automatons who lose our own abilities, so that when the machines fail, we will not be adequately equipped to take back control. When artificial intelligence begins to perform routine tasks that humans have done before, some skills may become more redundant. For example, the importance of horse riding skills has decreased since the introduction of automobiles, but this will lead to the opening of new skill areas. Moreover, the question is not whether these systems will make mistakes, but whether they will make fewer mistakes than human-controlled activities. The answer to this is yes, it will make fewer errors, otherwise these systems would not be used, which would be a great blessing to humanity.

MYTH 3: ARTIFICIAL INTELLIGENCE WILL DESTROY OUR PRIVACY

Fact: AI will not be able to impact privacy because most information applications are tied to laws and regulations.

If these smart machines crack large amounts of data, they can definitely destroy our privacy. But there are many reasons why these opponents are wrong. First, although AI systems have the ability and even the need to collect and analyze more information, the privacy threat is somewhat greater than for non-AI systems that collect and analyze large amounts of data. Moreover, the privacy-protecting rules that already apply to data will also apply to data analyzed by AI today.

In short, this is a simple policy question, not a technology question. If we do not want government agencies to collect certain data, parliament can mandate it and courts can enforce it. In other words, whether institutions have a machine learning system or not is irrelevant in this regard. Additionally, combined with AI data analysis, it can be provided without the risk of personal information being disclosed.

MYTH 4: ARTIFICIAL INTELLIGENCE WILL CAUSE MISUSE AND PREJUDICE

Fact: In most cases, AI will be less biased than humans.

Machine learning systems are more complex than traditional software systems. It was pretty clear how advanced systems based on old rules made decisions. Some critics argue that this level of complexity will lead to “algorithmic bias,” which will lead to intentional or unintentional exploitation of institutions. Because organizations will hide behind algorithms and use their complexity as a cover to justify abuse, discrimination, or other unethical or harmful behavior.

Artificial intelligence systems are the same in any technology.
may be used unethically or irresponsibly. But
Based on this concern, those who are against artificial intelligence are missing a point: Machine
Learning systems are independent of the company or software developer that develops them.
they are not. If a company systematically abuses certain groups
If it wants to discriminate, it doesn’t need artificial intelligence to do so. Moreover,
an algorithmic system produces unintended and potentially discriminatory consequences
If it does, it is not because the technology itself is malicious, but simply because
to comply with instructions determined by humane decision making or to avoid more bias
It is based on real-world datasets that reflect As a result, in many cases
these systems are less biased than human decision-based systems.

MYTH 5: SMART MACHINES WILL DESTROY THE HUMAN RACE AND TAKE OVER

Fact: We’ll be lucky if smart machines get smart enough to make us a sandwich.

There are some claims that these machines will become extremely intelligent and make better decisions without humans. It’s sad to see the public become technophobic because we take these science fiction claims seriously. The view that smart machines will kill us exaggerates the pace of technological development, as the processing power of silicon computer chips becomes slower and slower. The pace of progress in artificial intelligence outside of deep learning is relatively limited. Moreover, the human brain and machines are completely different systems, with even major advances in computing unlikely to produce a machine with the intellectual capacity, imagination or adaptability of humanity.

As MIT computer scientist Rodney Brooks puts it, “We greatly overestimate the capabilities of today and the decades to come because we generalize the capabilities of machines by looking at their performance.” Even if human-level intelligent machines were built, which is not possible, they would remain under the control of humans. Because we don’t produce them unless they are substantially safe and the benefits outweigh the costs. (Just like all technologies released)

Article Author: Robert D. Atkinson, 2016, NCSSS Journal