Millions of views and the sudden fame that descended upon the professor from the University of Louisville in Kentucky (USA) confirm that the scientist’s message is, at the very least, alarming. And that is exactly what it is.<\/p>\n\n\n\n
Regarding AI, Yampolskiy admitted that for years he had been convinced that building a fully safe artificial intelligence was possible. However, the deeper he analyzed the subject, the more painfully he realized that the idea of “safe superintelligence” contradicts the very principles governing such systems.<\/p>\n\n\n\n
\nI believed it was possible to create safe AI until I understood it was impossible. I was convinced we could create safe Artificial Intelligence, but as I looked at it more and more,<\/strong> I saw that it’s just something that can’t be done \u2013 Yampolskiy said in the podcast.<\/p>\n<\/blockquote>\n\n\n\n
These words do not sound like the thesis of a futurist. Instead, they sound like the confession of a man who has been studying only one problem for over a decade: how to avoid being destroyed by our own technology<\/strong>.<\/p>\n\n\n\n
A World Without Work and 99% Unemployment<\/h2>\n\n\n\n
Although the professor clearly states that a real danger looms, many viewers found it difficult to accept the claim that, within a few years, AI will replace most people in most professions. Indeed, this segment shocked many. Yampolskiy is not talking about the distant future. He is not talking about 2050, 2040, or even 2035. He is talking about now<\/em>. To quote:<\/p>\n\n\n\n
\nIn two years, capabilities will emerge that will allow the replacement of most people in most jobs \u2013 and this will happen very quickly.<\/p>\n<\/blockquote>\n\n\n\n
He continues:<\/strong><\/p>\n\n\n\n
\nWe are heading toward a world with levels of unemployment we’ve never seen. I\u2019m not talking about 10%. I’m talking about 99%.<\/p>\n<\/blockquote>\n\n\n\n
This is not a catastrophic narrative designed for clicks. Rather, it is a logical conclusion derived from observing progress whose pace is accelerating exponentially: models are growing, systems operate autonomously<\/a>, and AI is beginning to self-improve. A human with skill X? By contrast, AI will learn it in a week. A human with experience Y? AI will process colossal archives in minutes.<\/p>\n\n\n\n
Yampolskiy on AI: \u201cThere is No Plan B\u201d<\/h2>\n\n\n\n
Even if someone tried to comfort themselves with the thought that people would “re-skill,” Yampolskiy cuts this thought short, brutally but honestly:<\/p>\n\n\n\n
\nIf I’m telling you that all jobs will be automated, that means there is no Plan B.<\/p>\n<\/blockquote>\n\n\n\n
Re-skilling only works if a target field exists. The problem is that<\/strong> AI is entering every field<\/strong>: programming, design, marketing, data analysis, law, medicine, manufacturing, customer service, architecture. The list is endless, and according to the scientist, robotics lags behind AI itself by only 5 years. Consequently, humanoid robots<\/a> will also replace people in physical labor.<\/p>\n\n\n\n
Where Does Humanity Fit in the World of AI?<\/h2>\n\n\n\n
What strikes the audience most in this interview is not the catastrophic vision of a machine rebellion, but the humanist<\/em> standing behind the words. Yampolskiy is not saying this to scare people. He speaks because he feels a profound responsibility.<\/p>\n\n\n\n
\nMy hope is to lead to a situation where the superintelligence we are currently creating does not kill us all.<\/p>\n<\/blockquote>\n\n\n\n
Yampolskiy reminds us that even without physical danger, the social consequences could be equally devastating: the loss of jobs for billions, the loss of self-worth, the loss of daily structure, purpose, belonging, and social polarization on an unprecedented scale. This is because work is not just an income; it is identity. In brief, AI may take away this identity faster than we are willing to admit.<\/p>\n\n\n\n
Prof. Roman Yampolskiy on AI:<\/h2>\n\n\n\n