UN Expresses Serious Concern Over AI Development Danger Comparable to Nuclear Weapons

In a report released by the United Nations yesterday (12), it was pointed out that the rapid advancement of technology has heightened the risk of disseminating false information, including the generation of artificial intelligence (AI), posing threats to the integrity and accuracy of various media information, significantly impacting social security. The report specifically mentioned deepfakes.

AI Technology Combined with the Internet Will Enable the Rapid Spread of Fake Information

The United Nations claims in a report that AI-generated fake news, hate speech, or terrorist content will be presented to audiences in a way that is too realistic to doubt, and will be disseminated on various media or social networks. In May of this year, an image generated by AI about an explosion near the Pentagon was widely circulated, triggering panic and causing a brief dip in the S&P 500 index.

Similarly, last month, Jimmy Su, Chief Security Officer of the cryptocurrency exchange Binance, mentioned that fraudsters are increasingly using deepfake tools to forge videos of other people to carry out identity verification (KYC) processes on exchanges, making it difficult for verifiers to distinguish between authenticity and fraud.

In light of this, United Nations Secretary-General António Guterres stated that many preventive and regulatory measures are being discussed, including the EU's proposed Artificial Intelligence Act, and the upcoming UK Artificial Intelligence Security Summit in the fall.

AI = Nuclear Weapon?

According to a report by Magazine by Cointelegraph, it compares AI technology with threats like nuclear weapons and cites a recent warning issued by the Center for the Security of Artificial Intelligence (CAIS), signed by hundreds of scientists.

The article mentions an interesting perspective where OpenAI CEO Sam Altman advocates for the establishment of a global organization to oversee AI technology, similar to the International Atomic Energy Agency (IAEA), possibly exaggerating its dangers to allow leading AI companies to implement authoritarian control and stifle other competitors.

The Real Reason for the Explosive Development of AI

It is well known that AI applications have experienced explosive growth in the past year, from generating text, images, videos, and songs to voice and deepfake technology, which people around the world are using.

Aza Raskin, Co-founder of the Center for Humane Technology and Earth Species, provides a simple explanation for this phenomenon and emphasizes breakthroughs in deep machine learning models: "You can see everything as a language, as images that can be arranged linearly, and then you can predict what is most likely to happen next."

Images can be seen as a language, sounds can be broken down into smaller units, as long as they can be predicted, they become a language. This includes fMRI data and DNA. Therefore, what you see before you, "learning and predicting," will immediately double and enhance the progress of AI technology.