There are many myths surrounding artificial intelligence. Most of them appeared due to misunderstanding of the phenomenon, but some of them are intentionally cultivated by conservative people out of fear of everything new and incomprehensible.
Myth 1: AI Works Like the Human Brain
In fact, artificial intelligence is a discipline of computer engineering. It consists of software tools, and some forms of it may at first glance seem smarter than humans, but don’t think that AI in its current form is like our intelligence. Even though some forms of machine learning have been inspired by the human brain, AI and our intelligence are not equivalent.
For example, AI image recognition technology is more accurate than most humans, but it will be useless if someone wants to automatize playing Play casino games or searching for a flat with certain filters in any other task because it aims to solve only one, unlike the human brain.
Separately, let’s mention machine learning using neural networks, which many believe can learn just as well as humans. In fact, while neural networks are powerful, they are a far cry from the complex structure of the human brain or imitation of our capabilities. Neural networks can solve many types of tasks, but they are not capable of creativity, for example. In other words, this technology can generate a beautiful picture based on analyzing multiple images of real artists, but it will only be imitation, not creativity.
Myth 2: AIs Learn on Their Own
Many people may think that AI learns and processes information by itself, representing something like Skynet from Terminator. But without human help, modern artificial intelligence is unlikely to be capable of even insignificant steps. Programmers form tasks, prepare data, define data sets, eliminate errors in training data, and constantly update the software to integrate new information and set priorities. Without human intervention, AI cannot perform serious tasks at this stage of development.
Myth 3: AI Will Replace All Professions
At the moment, AI can replace humans in a very limited list of tasks, such as industrial equipment management, Big Data analysis, or tech support for your bank. But even here you cannot do without human intervention (at least it determines all inputs), so smart machines are not yet able to have a significant impact on the labor market.
But the development of AI is still fraught with the disappearance of some human activities and even entire professions. However, this is absolutely okay for general technical progress: we no longer have telephone operators, coachmen and water carriers, but we have robotic hotlines, Uber, and water delivery services, which also require human participation. Similarly with artificial intelligence, the emergence of which automatically created a demand for specialists in its development. By the way, a great reason to start exploring this niche if you can’t get rid of the obsessive fear of losing your job.
Myth 4: AI Can Make Sense of Any Messy Data
The quality of the data is essential to AI. Many people think you can just feed any data to artificial intelligence and it will give you the right answer, using powerful machine learning algorithms to do so. In fact, the AI can’t understand data that is too general and hasn’t been pre-processed in a way that the system can assimilate it. So, instead of pouring any information you can get your hands on into the AI, you need to carefully select high-quality content. Without this, even a powerful AI will not be able to learn the best way to solve a problem, because it does not have the creativity of a human.
Otherwise it will turn out like with Instagram AI, called @neural_tp. It’s a neural network trained on the Instagram accounts of thousands of real users. This artificial intelligence has mastered Elon Musk’s personal photo album of portraits and substitutes his face for random photos of girls found on their personal Instagram pages. It looks a little weird and even a little repulsive. But the captions to the posts on this account are a great example of metamodernism.
Myth 5: Everything Will Be Just Like in Sci-fi Movies
Unlikely. To make the narrative more exciting, writers and screenwriters purposely portray an artificial intelligence that resembles us in nature – possessing not all, but some of the human emotions, such as fear and anger.
However, this may not be the case. Who guarantees that an artificial intelligence will not turn out to be completely unlike humans?
Besides, movies try to construct a story where humans and AIs desperately fight each other in order to rock the viewer on emotional swings. In reality, instead of waging war, the artificial brain will simply launch nuclear bombs or find another way to instantly get rid of the Homo sapiens species.
Myth 6: The Term AI Has a Certain Meaning
Over the 70+ years of its existence, the scientific field called AI has spawned one of the main areas of scientific and technological progress. In today’s world, AI embodies thousands of research objects, products and services using intelligent algorithms in hundreds of applications. As a result, the term AI has become a general characteristic of all these research, products and services. That said, the specific meaning of the term by itself says little in isolation from what it refers to.
Myth 7: Ethics Must be Developed for Safe AI
Everything we call AI is just tools, and applying ethical principles to them is pointless. And even developing such principles for the future – just in case AI ever becomes agency – is a futile endeavor. The AI-equipped power saw of the future doesn’t need machine ethics to keep people from sawing.
Myth 8: Legal Regulation of AI Is Impossible
Arguments as if AI is too difficult to regulate and that it will kill innovation are untenable and extremely dangerous. Defining a legal framework that restricts the development and use of dual-use technologies is the de facto standard in today’s world. Without it, the survival of humanity in the face of ever-increasing military, terrorist and technological risks is unthinkable.
Why Shouldn’t We Be Afraid of Myths?
When you watch sci-fi movies, robots are very reminiscent of humans. In fact, in real life, machines will never possess emotions such as fear, anger, or genuine joy. Artificial intelligence will never resemble humans. It means that humans will still be more cunning than robots and will be able to use their emotional component to their advantage.
For now, the world is still just observing the development of artificial intelligence. In time, we will be able to see how far scientists should go in their developments. We will never be friends with robots, but they can be very useful to those around them. And while scientists on our planet are working on creating super-smart machines, it’s worth thinking about our own intelligence. After all, it’s a great advantage to have developed intelligence.