(Yicai Global) Feb. 23 -- OpenAI's ChatGPT is likely to generate the kind of self-awareness that will threaten the human race in two to three years, according to the founder of Chinese cybersecurity firm 360 Security Technology, also known as Qihoo 360.
The chatbot will achieve capabilities that its designers could not predict via hundreds of billions of system parameters, online news outlet The Paper reported today, citing Zhou Hongyi.
The biggest difference between ChatGPT and other types of artificial intelligence is the feature of being human-like as previous AIs did not actually understand or code knowledge, but simply matched already existing answers based on keywords, Zhou said. Instead, ChatGPT answers questions in a flexible and mature way, he added.
When talking about security and ChatGPT, the focus is not on technical details but on social and ethical implications, per the founder. The chatbot can be used as a powerful tool for cybercrimes, as many hackers have demonstrated using ChatGPT to write phishing e-mails and attacks, he added. Qihoo 360 is studying technologies to establish a knowledge base that automatically monitors and detects AI-generated attacks, per Zhou. It will implement AI to complete such processes.
ChatGPT is a strong tool for productivity and reconstruction and it can impose great impacts on the entire industry, Zhou said, adding that the development of such technologies will determine whether China will fall behind or catch up in science and technology with other nations in the next decade.
The AI gap between Chinese companies versus overseas giants such as Microsoft and Google is not as large as imagined, Zhou explained. China will hopefully boast world-leading capabilities in ChatGPT-like chatbots in two to three years so exploring such technologies should be fully supported, he concluded.
Editor: Emmi Laine, Xiao Yi