AI as a real threat?
Dr. Leszek Bukowski analyzes the challenges related to the development of artificial intelligence, pointing to business and geopolitical aspects and the risk of losing control over the technology.
Quick Look
- Leszek Bukowski in an interview with Radio RMF24 discusses the threats related to AI.
- The expert points to the rivalry between the US and China, the business strategies of technology companies and the risk of losing control over self-reinforcing artificial intelligence models.
AI-generated summary
Why It Matters
The development of AI models is based on training new systems using previous ones, which creates a self-reinforcing process. The US currently controls 75 percent. AI computing power.
Dr. Leszek Bukowski believes that scaring people with artificial intelligence is a business strategy of companies dealing with it.
From the very beginning, companies have portrayed themselves as having the most powerful technology in the world so that we start to feel a little afraid. This is one layer. The second layer is the competition between the United States and China. Both countries see artificial intelligence technology as crucial for further development and technological progress. The third layer is the real issues related to the security of artificial intelligence - notes the guest of Radio RMF24.
The expert on new technologies refers to an essay by computer scientist Jakub Pachocki, who studies the development of models at OpenAI.
He wrote that one of the things to watch is the self-reflexivity of the technology. This is an idea that appeared a long time ago, after World War II. Von Neumann said, "What if there were automata that could recreate themselves"? And today it is so. When training subsequent models, we use previous models. And this is a very dynamic process in which a person can lose control. Pachocki used the phrase that we breed subsequent generations of these models rather than create them. This is probably the first time that humanity has encountered technology that is self-propelling, emphasizes Dr. Leszek Bukowski.
Can artificial intelligence be stopped?
The host Tomasz Terlikowski wonders whether, in the face of the threat of artificial intelligence, it is not enough to turn off the appropriate plug-in, since AI is powered by energy.
This thesis has appeared several times that we press the magic button and turn off artificial intelligence. This is not entirely true, because artificial intelligence is developing most today in areas such as software development and mathematics. Now this is all our modern technology. What if in a dozen or so years we will be in a world where this technology does not function without the use of artificial intelligence? Then you can think a little about whether we actually control the kill switch - replies Tomasz Terlikowski's guest.
Who wins the AI race?
The guest of Radio RMF24 notes that artificial intelligence is currently controlled by several companies in the world, mainly from the United States and China.
The United States is responsible for 75 percent. computing power when it comes to artificial intelligence. Europe is only 5%. – says an expert in new technologies.
Bukowski believes that the rest of the world is just customers. To be in a negotiating position towards these parties, Europe must build its own technological potential, says Dr. Leszek Bukowski.
Open Questions
- Will Europe be able to build its own technological potential?
- How to effectively control self-reinforcing AI models?







