Artificial intelligence and the (possible) downsides
Few things dominate the headlines as much as the topic of artificial intelligence. In both business and personal contexts. OpenAI has unleashed an avalanche with its flagship ChatGPT. Now, any tech giants are racing each other to create ever more powerful AIs. No wonder, because artificial intelligence is already capable of amazing things. But where is the journey heading? We summarize and rank.
Nearly unlimited possibilities
ChatGPT can do quite a lot. With the GPT-4 update released at the end of March, the performance portfolio has been greatly expanded once again. Already the predecessor GPT-3.5 could spit out masses of information, write page-long texts, solve math problems and much more. The newer version solves these tasks more reliably – and a lot of new tasks as well: It can evaluate images, understand humor and passes prestigious exams of American universities as well as the Bavarian Abitur with flying colors. ChatGPT is just one example: Bing AI, YouChat and Co. bring similar capabilities.
AI is also causing a stir in other areas. In music, for example: in April 2023, a song by The Weekend and Drake “appeared” and went viral on TikTok. Hardly anyone realized that this was created with an AI and had to be clarified by the artists’ label. The AI imitated the artists’ voices, which made for a deceptively real result. And also in the field of image processing, artificial intelligence is able to “lead us behind the light”: A photo of the Pope wearing a hip down jacket caused a stir on social networks. This was also created with an AI. Comparable cases also occurred in the political context.
In the field of image processing, OpenAi has its own flagship, which goes by the name of Dall E. The company also has a number of other products. One sentence, e.g. “a cat driving a ferrari”, is enough and you get exactly that sentence as an image.
Chance or peril? Or both?
The capabilities of AI, which most laypeople probably would have thought impossible three years ago, are undeniable. With them come many opportunities. Unfortunately, however, not only; critical questions also arise.
The topic is particularly present in the field of education: ChatGPT is able, for example, to take over essential performances of pupils and students, for which they are graded. Homeworks as examination performances thus become almost witless. A possible solution comes from OpenAI itself: The AI Classifier is supposed to distinguish AI texts from human texts. Currently, however, it is still too unreliable to be used in practice.
This is just one of many examples: Fake news or misuse by propaganda are others. Likewise, data protection poses a major problem. Not only in the area of personal data. AI learns with the data it is “fed”. That’s why companies – Apple, for example – are moving to ban their employees from using ChatGPT. The concern that company internals could leak out is too great. And hanging over everything is the uncertainty of how artificial intelligence will develop. Will it one day be so powerful that it takes over human jobs? Most likely, yes. Or perhaps even so strong that it takes over control and regards humans as an inferior species? Experts believe this is also possible. This is one of the reasons why several tech greats – including Elon Musk – are calling for a halt to the development of artificial intelligence. Our conclusion: Opportunity and danger – depending on what we humans make of it.
One thing is certain: artificial intelligence will change many things in the coming years. That makes it all the more important not to miss out on this development.