Abstract: LLMs are rapidly evolving and handling more and more tasks. They introduce low friction to problem solving and reward with immediate output. This talk dives into what the neuroscience suggests, how excessive LLM usage affects students and knowledge workers, fosters metacognitive laziness and what happens when thinking becomes optional. We also look at how this shift in adoption plays out in software engineering practice. Code is increasingly produced via prompts, creating new bottlenecks: more output than we can meaningfully review, delegating tech debt to a later timeframe and a different cognitive load. Finally, we explore a possible framework to make adoption more sustainable and safer for the individual.

Chair: Hassan Atwi