Ten percent, he says no research

The stamp on this IT Home page is 2026/10/24 10:56:44. It retells an AFP piece from October 22: at a company meeting in Berlin, NVIDIA chief executive Jensen Huang again attacked talk that AI will destroy humanity. He treats it as nonsense, and as fabrication and cheap hype meant to make headlines. The translation on the page says: when you hear that AI will end human civilization, do not believe it. Those lines win headlines by fabrication and cheap hype, and they are not fact. Not fact is his judgment. The page marks the lines below as IT Home's translation of Huang's remarks. A translation on this page is a retelling. It is not a word-for-word check against the AFP English. AFP's date is October 22. The stamp on this page is October 24. The two days should not be collapsed into one.

The same page also records an earlier occasion. It should not be collapsed into the Berlin day. In a New York Times interview in September this year, he also said panic about AI in the industry and among the public had gone too far, and that current AI will not necessarily replace occupations, but will replace tasks first. Not necessarily occupations, and tasks first, belong to that September interview. This Berlin page does not include a list of jobs. The page also says he thought panic about AI, in the industry and among the public, had already gone too far. Too far is the judgment in that interview. This page does not attach a separate survey of how afraid people were.

The ten percent stays inside the sentence he is rejecting. The page says that, answering scientific experts who think AI will destroy humanity within ten years, Huang says all of their predictions have been wrong, that the so-called 10 percent probability has neither a scientific basis nor research behind it, and that these predictions only hurt people. Ten years and 10 percent are figures inside those views. This page does not give a probability of his own, and it does not attach a study. The so-called 10 percent cannot be written as a risk that has been measured.

His positive line on this page is an attitude. Huang calls for a responsibly optimistic view of AI, and says that if people care about the future after us and about sustainable development, now is the time to build artificial intelligence. Now is the time is a call. It is not a schedule. The page's word is the best time, not a year. Responsible optimism also comes without a list of steps. Hurt people is the result he says those predictions cause. This page does not say who was harmed, or how. A separate Berlin link about software companies, and the piece about ChangXin Memory, stay at the bottom and are not folded in. Virus-making and attack methods are not on this page, and this story does not write them.