An AI safety talk at Yıldız Technical University
Together with MATE, we talked at Yıldız Technical University about where AI stands today, its risks, and where to start in AI safety.
Within a few years, AI systems have gone from tools that complete text to systems that can make decisions on their own. The speed of this progress is plain to see; where it will lead is discussed far less.
On Friday, 23 October, together with MATE (Modern Smart Technologies and Information Club), we took up this question at Yıldız Technical University. In the first part, we looked at where things stand today: What can these systems actually do, and how did we get here? Then we moved on to the part that gets talked about less. Often, even the people who build a model cannot fully explain why it behaves the way it does. If systems that can act on their own, and whose decision-making we don’t understand, keep advancing at this pace, where will they take us? Starting from this question, we talked about the risks, what AI safety is, and where people who want to work in the field can start.
The speakers were Berke Çelik (co-founder of AI Safety Türkiye) and Ege Uğur Amasya (PhD student at Koç University, SPAR Research Fellow).