The Dual Nature of AI: Fascination and Fear

AI triggering both fascination and unease | News | Hamburg Business

More and more people enjoy using ChatGPT. Around one in three people had tried out the chatbot by the U.S. software company, OpenAI, at least once by late 2023 and 53 per cent of users said they enjoyed the interaction with artificial intelligence (AI), a survey commissioned by the Bitkom digital association found. Yet the bot is not really intelligent, according to Andy Witt, Professor of Business Informatics at the Hamburg School of Business Administration (HSBA).

Understanding ChatGPT

"ChatGPT is a large language model based on word statistics. It offers the most likely response to a query or prompt. If I use unusual terms to prompt as precisely as possible, the probability increases that ChatGPT cannot categorise the term - and then gives a wrong answer." Blind trust in the technology is misplaced, Witt stressed, whose research focuses on big data analysis, digitalisation of production and AI.

AI in Education and Business

ChatGPT can be a useful tool when used cautiously. Around 85 per cent of students and 90 per cent of professors use AI to come up with ideas, to do research, or to write texts, a non-representative HSBA survey has found. HSBA trains students in dual course of studies in cooperation with around 300 partner companies as part of its role as a Hamburg-based business university. These companies are at the cutting edge of technology, with 57 per cent using generative AI.

Fascination and Unease

Can the increasing use of ChatGPT lead to fewer reservations about AI? Around 41 per cent of Bitkom respondents, especially in the 65+ age group, said ChatGPT frightened them. "Most fears about AI centre around possible redundancies or its use in military contexts," Witt said. Many people have not understood the process of generating results. A generative AI model makes connections between words, sentences, and facts based on huge amounts of data. Nevertheless, the mere idea of a 'learning' machine causes some people to feel uneasy. This is precisely what fascinates other people.

Regulation and the Future

ChatGPT is practical and offers tangible, everyday support, whether you consider it magical or a complex mathematical system. And it can remove barriers, Witt pointed out, adding: "I can communicate with the bot naturally, which makes AI seem human. The simple version is free, so there are no financial barriers. As with any technology, AI can be used for good or for ill. And AI is definitely here to stay. "We must control it as we see fit." The EU's AI Act took effect on August 1 and is a decisive step in the right direction.

Conclusion

Regulation is crucial and must be reliable as AI is gaining momentum in the industry now that the hype in the media has peaked. "Industry is increasingly using AI for new products and to increase efficiency," said Witt, who uses the technology himself. "We are moving towards a new normality. ChatGPT was a small revolution when it was launched in November 2022," said Witt. But every revolution is followed by disillusionment. "And that's exactly where we are now in terms of the practical application of AI and without lofty visions. We should not fear AI."

ys/pbNOTE: The AI images used in this article were created with DALL-E using Open AI, based on a prompt by Professor Andy Witt.