How far has Artificial Intelligence progressed at the moment?

How advanced is artificial intelligence actually?

Have you ever thought about how often you come across artificial intelligence in your everyday life? If you think about it, a lot of things come to mind. Such as: Google search, navigation systems or voice assistants such as Alexa.

Humans are constantly developing and leaving their tasks to the AI. 12% of German industrial companies use AI. How far has AI actually developed? I report that and much more in this article.

The definition of AI

The first question that people often ask themselves about artificial intelligence is how do you define artificial intelligence in the first place? Because if the term “intelligence” alone is not easy to define, how should one define artificial intelligence at all?

Several definitions are presented in the book “Artificial Intelligence” by Uwe Lämmel and Jürgen Cleve. A definition describes that the AI ​​takes over activities and tasks of people, regardless of whether the activity is complex or not. Algorithms should adapt to human behavior.

Multiple categories of AI

AI is categorized according to several factors. On the one hand, there is “Weak Artificial Intelligence”, which specializes in only one task. This includes, for example, the chess AI "Deep Blue" which defeated the world chess champion in 1997. Cell phones and emails are also weak AI. There is also the “Strong Artificial Intelligence” and the “Strong Artificial Superintelligence”. Strong Artificial Intelligence is on the same level as human intelligence, but strong superintelligence should be even stronger than human intelligence. However, strong AI still does not exist these days.

In order to build a strong AI, one would have to imitate the human brain. This is done through artificial neural networks with the help of deep learning.

Deep learning

Deep learning in German “deep learning” is a modern technology that is part of what is known as “machine learning”. Artificial neural networks are formed that are inspired by us humans. Various stimuli are transmitted in our brain. If we see a face or a car as an example. This transmission of stimuli is of course subconscious. Thanks to our experience, we can determine what kind of people or cars it is, because we know the characteristics and so we try to teach the AIs.

Images and symbols are stored in the network and then assigned to the respective people. The more often the AI ​​is trained, the lower the error rate, which is similar with deepfakes.

In my article on deepfakes, I explained that deepfakes are created by training an AI. The AI ​​analyzes multiple frames to mimic the person's facial expressions, thereby manipulating videos of people in public life.

Conclusion

We live in a digitized world that is full of artificial intelligence. Although there is still no artificial intelligence that is on a human level, the fact is that science is unanimous and it is only a matter of time before there will be “strong artificial intelligence”.

If you have any further questions regarding technical knowledge, please do not hesitate to contact us.

Did you like the article?

Subscribe to our newsletter for IT decision-makers!