
Johannes Simon / Getty Images
Research on artificial intelligence began a long time ago. It has been almost thirty years since Deep Blue defeated world chess champion Garry Kasparov. IBM’s machine at the time could only play chess. In the emergence of AI, it can be seen as a remarkable milestone when a human is defeated in a game created by humans, by a machine also created by humans.
It’s time to clear up some of the confusion out there. Technological development has accelerated so rapidly that it’s becoming increasingly difficult for everyday users to understand and process it. One of the biggest "sceriest monster" is the artificial intelligence — something everyone thinks they know and believe they’re already using correctly… yet most non‑experts have ended up completely misunderstanding it.
With this post, my aim is to help you navigate this AI jungle and find the right path.
Let’s not rush in headfirst; let’s start with the basics. First of all! The question is :
What is artificial intelligence (AI)?
Artificial intelligence (AI) refers to a computer algorithm or program capable of imitating intelligent human behavior.Contrary to popular belief, AI cannot think the way humans do. The AI analyzes large amounts of data, uses statistical algorithms to identify patterns within that data, and provides results to users based on these patterns.
- One of the most common misconceptions!
1.“AI learns by itself” – Not true. Artificial intelligence is trained by humans using datasets; it does not “self‑learn,” but works from data that has been provided to it in advance.
2.“AI understands human language” – It does not. Machine models break words into tokens and generate responses based on statistical patterns.
3."AI is creative” – Only in appearance. What we perceive as creativity is actually the combination of patterns; behind the novelty there is no conscious intention or understanding.
4."AI corrects its mistakes” – This is only partly true: it happens through limited and guided use of memory, not through classical learning.
AI cannot function without large amounts of data, and the algorithms behind it rely on statistical methods that were developed decades ago. Artificial intelligence has become such a popular topic today because enormous volumes of data are now available across nearly every industry, and computing power has reached a level that can meet the demands of AI‑driven applications.
The AI is therefore not a thinking machine, but software created by humans that analyzes data using statistical methods. It can identify patterns within large datasets and, based on these patterns, generate predictions and provide answers.

How does artificial intelligence work?
Artificial intelligence can now be found across a wide range of industries. It powers Google search, the personalized video recommendations on YouTube, the Facebook news feed, traffic management systems, self‑driving cars, and most recently, the development of large language models reached a milestone that became widely known through OpenAI’s ChatGPT. Whatever AI application we talk about, there are two fundamental conditions required for its creation. These are the following:
-A large amount of data
-A training technique, which typically consists of statistical methods
We are living in the era of “big data,” meaning that businesses and governments collect data about the decisions of individuals and companies across every area of life. While large datasets used to exist mainly in finance and economic fields, today they cover virtually every industry. Data on its own is useless, however it must be analyzed, and this process is what we call training artificial intelligence. There are several ways to train an AI. We are not digging in this any deeper for details at this moment.
Let’s talk about something a bit more exciting! Deepfakes, data quality, and the battle of AI vs. AI
What the hell is that deepfake? A deepfake is a type of artificial intelligence that creates very realistic but completely fake images, videos, or audio. It can make it look like someone said or did something they never actually said or did. Deepfakes work by learning from many real pictures or recordings of a person, then using that information to generate new, convincing fake content.
For example, imagine a video where a famous actor appears to say something shocking — but in reality, they never said it. A deepfake can take real videos of the actor, learn their face and voice, and then create a completely fake video that looks and sounds real.

Source: Shamook (2020)
Or....there is another simple example, the deepfake can be a voice too. Imagine getting a phone call from someone who sounds exactly like your boss, asking you to send an urgent payment. But it’s not your boss at all — it’s a deepfake voice created by AI, using just a few seconds of real audio.
Deepfake technologies and AI‑generated content are becoming increasingly convincing. But this also means that reliable, original data sources are starting to run out, leaving less real material for AI systems to learn from. The big question for the future is this: if AI begins learning from AI‑generated content, could that degrade quality? And what might the solution be? One possibility is AI‑driven cybersecurity — in other words, AI defending against AI.
We can’t ignore the risks of AI either, so let’s touch on them briefly — without aiming for a complete list.
The main risks of artificial intelligence include:
-
Phishing and security threats: AI tools can unintentionally point users toward unsafe login pages, making phishing attempts harder to spot and easier to fall for. /Cyberattacks, FraudGPT<<Avoid clicking on suspicious links, never share your personal information with strangers, and make sure you’re protected with a modern antivirus solution such as Microsoft Defender or Norton 360 Deluxe/
-
Smarter, AI‑powered attacks: Cybercriminals are now using AI to create highly optimized scam sites designed to rank well in chatbot answers — giving their traps more visibility than ever. /Facial recognition systems can also pose risks, as they may capture your face and personal data without you even noticing, ChatGPT , and most other desktop AI software saves your conversations.../
-
Accuracy and trust issues: When AI produces incorrect or misleading information, these “hallucinations” can quickly erode user trust.
- Copyright concerns :AI systems learn from vast amounts of human‑created content — articles, books,songs, lyrics, Wikipedia entries, and more. Yet the creators behind this material often receive neither compensation nor recognition. Many AI companies argue that paying for all this training data would be financially impossible, leaving human creativity as the unpaid fuel powering these technologies.
Ethical and social concerns: AI can amplify existing biases, compromise personal privacy, and deepen social inequalities if not used responsibly. /Biased algorithms AI is only as “intelligent” as the data it’s trained on — and when that data is skewed, the system’s decisions become skewed as well. One well‑known U.S. example showed a healthcare algorithm giving more support to healthy white patients than to Black patients in serious condition, simply because it relied on cost‑based data that reflected existing inequalities./
-
Environmental protection: Running generative AI is energy‑intensive. A single prompt to ChatGPT can use roughly as much electricity as keeping a light bulb on for 20 minutes. The energy demand of data centers could soon exceed the total consumption of some entire countries
- Physical Threats: AI is now not only writing emails, but also fighting wars. Drones, autonomous weapons, and military decisions are increasingly relying on AI. This could lead to an escalation of conflicts.
-
Impact on the job market: As automation accelerates, certain roles may disappear, potentially creating economic pressure and uncertainty. /Some estimates suggest that by 2030, as many as 300 million jobs could disappear worldwide.Learn to work with AI >the future will favour those who use technology to become more effective, not those who are replaced by it./
Long story short, here’s the deal. Can the process be reversed? The pace of technological development can no longer be stopped — artificial intelligence has become an irreversible part of our lives. Regulations are lagging behind, challenges are growing, but one thing is certain: the education but also the regulation is essential.
We can clearly state that, in connection with all of the above, we need reliable artificial intelligence.
Reliable artificial intelligence (AI) offers advantages in several areas:
-
it can improve healthcare,
-
make transportation safer and more environmentally friendly,
-
increase the efficiency of industrial production,
-
and make energy generation and consumption cheaper and more sustainable.
-
I hope you enjoyed this little journey through the modern maze of artificial intelligence, and that I was able to give you a little help and inspiration so that you don't get scared when you see a driverless taxi on the street.
- If you have some energy left, enjoy the video below.















