When the latest version of ChatGPT was released between 2022 and 2023, it was the most impressive advance to date in terms of questions answered with AI. Since then, there has been a real hype surrounding AI. On the one hand, extreme efficiency gains are expected in the economy - on the other hand, there are warnings that many jobs will be lost. Many people are also calling for regulations and laws on the use of artificial intelligence because they see it as dangerous. We don't want to get involved in socio-political debates at this point - we are focusing on areas that are of interest to associations in order to save resources and implement projects more quickly. That is why today we are looking at areas where AI (artificial intelligence) is useful for associations.
What is artificial intelligence?
AI attempts to transfer human learning and thinking to the computer and thus give it intelligence. Instead of being dependent on individual programming for every purpose or work step, AI can find answers independently and solve problems on its own. However, the difference between this vision and the current possibilities is still quite large. However, there are significant further developments that make it difficult to tell the difference between human and machine task completion.
Programs in the field of artificial intelligence
ChatGPT is certainly the best-known AI program on the market. However, Google already has a promising competitor in the form of “Bard”. ChatGPT works like a chat bot, allowing users to ask a question and receive answers. This means that users can request information or knowledge, obtain recommendations and formulation aids or have content summarized. The program has been trained over many years with countless data and calculates the most likely suitable answer after a question has been entered. But AI has also been able to take decisive steps forward in image processing: ChatGPT developed a tool called “DALL-E” for this purpose, which is more advanced than the competitor product “Midjourney AI”.
Important: The AI does not yet “think” independently, but is dependent on information fed in by humans. This means that the underlying sources and data may be incorrect or problematic (racist, discriminatory or incomplete). In order to obtain satisfactory answers, users must ask very good questions or give very good instructions. These instructions are called “prompts” in the AI environment and influence the result. ChatGPT has a built-in blocker to prevent racist content from being published. In addition, sites with trustworthy and verified content (e.g. Wikipedia or Reddit) have been given a higher ranking. However, users always find ways to circumvent the blockers by using cleverly written prompts.