Like any major technological innovation, technology can be used for good and bad purposes and can be used with good and bad intentions. That was the case when a finance worker at a multinational firm was tricked into paying out $25 million to fraudsters using deep fake technology to pose as the company’s chief financial officer in a video conference call.
In his article, “The Dark Side of Superhuman Persuasion: Why We Need Education on Generative AI,” Professor Dries Faems, Chair of Entrepreneurship, Innovation and Technological Transformation at WHU – Otto Beisheim School of Management and leading expert in the field of AI, laid out the risks and threats of the new emerging technology.
Takeaways:
- Advanced AI systems can potentially influence human behavior and decision-making to a much greater extent than humans can.
- The GPT Builder application called the Election Persuader, showcases the ability of AI to tailor communications based on individual profiles and preferences, potentially influencing political decisions. The power of AI in crafting messages that resonate on a personal level could have significant implications for the future of advertising, campaigning, and even interpersonal communication.
- Education is essential in the face of rapidly advancing AI technologies. While regulation is necessary, it often lags behind technological innovation, making it difficult to manage the implications of AI globally.
Professor Faems writes that superhuman persuasion is the concept where advanced AI systems can significantly influence human decision-making by utilizing large datasets and algorithms to create highly personalized messages. This capability allows AI to connect with individuals on a deeply personal level, making it more effective than traditional human communication.
While Professor Faems adds that Sam Altman, CEO of Open AI, discussed this as a potential risk in AI development, Professor Faems illustrates that the application called “Election Persuader” was created using GPT Builder. This chatbot can craft customized emails to voters, encouraging them to support a particular political party based on their interests and beliefs. The “Election Persuader” requires the user to upload a political party’s program and a LinkedIn profile to generate a persuasive email. This application demonstrates the potential of AI to tailor persuasive messages that can appeal to individuals’ specific preferences, even if they initially seem unlikely to support the cause.
The conversation on AI is often complex, and the importance of these issues highlight the need for a comprehensive approach. Following up on his article, we had a short interview with Professor Faems:
- Professor Faems, there are voters voting in elections all around the world this year, and we now know that AI is increasingly being used to manipulate voters with tailored content. How can the public and tech companies keep up with this?
Keeping up will be quite challenging as the technology is developing at exponential speed. Nevertheless, we should make sure that citizens are at least aware about what is happening to improve their ability to recognize the difference between content created by humans and content created by AI. In my opinion, educational institutions at all levels of the educational trajectory should start thinking about how they can prepare their students for these challenges.
- How transparent should technology companies be about their use of AI?
At the technological level, transparency is good but will not be sufficient. The problem is that technology companies themselves do not fully understand how these AI technologies exactly work. At the application level, companies should indeed be obliged to clearly disclose which AI technologies have been used for particular applications and how they are used.
What happens next?
While A.I. has taken a giant leap forward — and is evolving so rapidly that it is hard to keep up with the state of play. The big question remains, “How do we actually equip youth, our current generation and future today with AI skills and do it with diversity and inclusion?” and “How do we not leave people and students further behind when much remains unknown about the technology even as its development speeds ahead?”
![](https://cdn.sanity.io/images/32hqilj1/production/02c2da925212ab440561f1403c150c13608a3a91-4200x2800.jpg?w=342&h=228&auto=format)
Save a spot in our bootcamp: Unlocking the Power of ChatGPT and Generative AI for Business Development, in which Professor Faems will give an introduction on how to write effective prompts, combine different AI tools, and develop agents to execute tasks in the context of specific use cases. Registration is open until June 7th, 2024.