Between enthusiasm and disillusionment
Advertisement
The year 2024 was marked by rapid progress in the field of artificial intelligence (AI). But alongside excitement about new possibilities, there was also disillusionment about the limits of technology. In this last deep dive of the year, the experts from heise Magazine and The Decoder take stock and provide assessments of the most important developments, challenges and trends.
“What has moved me most over the last year of AI is the back and forth between enthusiasm and total despair,” sums up Wolfgang Stieler from Technology Review. On the one hand, AI tools like ChatGPT make the analysis of large amounts of data much easier. On the other hand, the models quickly reach their limits when it comes to reliability and reproducibility. “Transformer models are fascinating, but not yet the final solution,” says Stieler.
Hardware boom with downsides
The AI hype is fueling demand for specialized hardware, reports Christoph Windeck from the c't department: “The market value of this former graphics card company [Nvidia] is now higher than that of Apple.” But the boom also has its downsides: “The environment falls by the wayside, concerns about electricity and resource consumption fly overboard.” The question of profitable and sustainable business models in the area of generative AI also becomes one represent a central challenge for the coming year.
AI in everyday life
Smart glasses could establish themselves as a new device category, believes Eva-Maria Weiß from heise online. After Google's data glasses “Glass” failed spectacularly, Meta's Ray-Ban glasses are enjoying great popularity – even though the AI functions are not yet available in Europe. “Glasses could bring real added value,” says Weiß. Meaningful applications such as navigation and the integration of an AI model are crucial.
André Kramer from c't observed rapid development in the area of image and video generators: “Image generators like Midjourney and DALL-E are getting better and better, so that it is now really difficult to distinguish whether you have a photo in front of you or a .” generated image.” This development also raises ethical questions, particularly with regard to the proliferation of deepfakes and the increasing difficulty of distinguishing between reality and fakery.
Dorothee Wiegand from c't highlights the positive aspects of AI in the education sector. She talks about tools like Fiete.ai, which can help teachers with individual learning support: “The program praises the things that the students have already done well and makes sensible suggestions for improvement in the event of weak points.” Given the shortage of skilled workers in schools Such tools represent a valuable addition to teaching.
How intelligent is artificial intelligence? What consequences does generative AI have for our work, our leisure time and society? In Heise's “AI update” we, together with The Decoder, bring you updates on the most important AI developments every weekday. On Fridays we examine the different aspects of the AI revolution with experts.
Small view
Generative AI has made tremendous progress in 2024, but many unanswered questions remain. The experts agree: There needs to be a conscious approach to the technology and a discussion about how we want to use AI in the future. Because artificial intelligence will penetrate and shape more and more areas of life in the next few years. The AI update will continue to help us remain informed humans in this age of intelligent machines in the new year. New episodes will be available again every weekday from January 6, 2025.
(igr)
