Thoughts about artificial intelligence and society

Artificial Intelligence (AI) nowadays has become a very controversial topic. Some feel threatened by a new technology likely to impact our future, either affecting their jobs, maybe making them obsolete in the workforce, or disrupting society by generating content that is impossible to be distinguished from being manufactured or real. Some regard the development of powerful AI as a tool to solve many of the issues and problems humanity faces right now, from climate change to huge progress in medicine and science. And there are always some, who expect personal gains from using AI, economic or political power.

Since ChatGPT from OpenAI (https://chat.openai.com/) became available to the public, the amount of users has risen to more than 100 million registered users and more than 1 billion visitors on their website per month (ChatGPT Statistics and User Numbers 2023 - OpenAI Chatbot (tooltester.com)). For many people, ChatGPT is their first conscience contact with an AI, their experience ranges from utter excitement to grave concerns about the future of mankind. ChatGPT is not the first AI made available for public usage, but it has had the biggest impact so far.

Benefits of using AI: Before registering for ChatGPT, I already used some very different kinds of AI; WolframAlpa and Microsoft Azure AI. I followed the development of AIs since the first chess programs, notably IBMs' Deep Blue and DeepMinds' AlphaGo. AIs today can help us to be creative like creating images based on a few suggestions, translating foreign languages, summarizing long text passages, assisting in writing CVs and cover letters, creating stories and poems, creating or checking software code, analyzing complex data, or simply telling some jokes. It may assist to develop new and better drugs for treatments, new materials, better models for climate research, breakthroughs in science, and maybe even better solutions to social problems. AIs may find solutions and provide answers to questions, thousands of scientists may not be able to solve.

Dangers in using AI: Besides the worries about an AI taking power and enslaving or genocide humankind, which I regard as very unlikely, there are real dangers in using AI. My main concern is the misuse of AI and its capacities by individuals and groups to gain political and economic power. Another danger is the lethal threat of weapons controlled by AIs, their capabilities, and the lower threshold to deploy them. My main concern today is the possibility to distribute targeted misinformation and the generation of manufactured realities. AIs may boost political campaigns, spread disinformation or disrupt markets for the benefit of a few, and the disadvantage of the society.

Conclusion: A temporary or permanent ban on AIs will be pointless, the benefits will be out of reach of the common people, and ruthless, powerful, and/or wealthy individuals will ignore it anyway, the promise of gaining an advantage over competitors is just too tempting. We need to develop rules using AIs, and sanctions on those breaking the rules. One of my main demands is the labeling of AI-generated content, like gene-modified food. Another is the limitation of AI-controlled processes, like the selection of candidates for a job, the creation of content and execution of political campaigns, or the risk calculation for insurance contracts. In my personal opinion, it's not AI posing a direct threat to us all, it's those who misuse it for personal gain.