The pace of development in artificial intelligence is accelerating, potentially leading to an “intelligence explosion,” according to a report published February 20, 2026, by Vaterland, a Swiss news outlet. The analysis, authored by Raffael Schuppisser, details how AI is moving beyond simple chatbot functionality and increasingly capable of self-improvement.
Since the launch of ChatGPT in November 2022, AI has become integrated into the daily routines of many office workers, assisting with tasks such as email responses, report writing, and data visualization. Whereas initial fears of widespread job displacement haven’t materialized, the technology is evolving at a rate that warrants renewed attention, Schuppisser writes.
A key development is the emergence of “AI agents” – systems that can independently perform actions on a user’s computer or the internet. Unlike traditional chatbots limited to text-based interactions, these agents can manipulate files, navigate menus, and even make online purchases, provided they are granted the necessary permissions. Companies have been cautious in releasing these agents widely due to the potential for errors and security risks, but development continues.
The second critical shift is the AI’s increasing ability to program itself. Previously, developing a new version of a large language model like ChatGPT would take over a year. Now, new iterations are appearing almost monthly. OpenAI’s recent release of GPT-5.3-Codex and Anthropic’s Claude Opus 4.6 demonstrate this accelerated pace. According to Anthropic CEO Dario Amodei, AI is now writing “a large part of the code” within his company. ChatGPT’s latest version, 5.3, was significantly developed by the AI itself, paving the way for even faster iterations like 5.4.
Amodei suggests that this self-improvement cycle could lead to a point where AI autonomously develops the next generation of AI, potentially within one to two years. While this doesn’t necessarily imply the emergence of superintelligence or consciousness, it has significant implications for the workforce. Amodei predicts that AI could eliminate 50 percent of entry-level office jobs within the next one to five years, and even highly skilled positions in fields like computer science are not immune.
Schuppisser draws a parallel to the early days of the COVID-19 pandemic, noting that warnings from experts were often dismissed until the situation rapidly escalated. He emphasizes the require for adaptation, suggesting that future success will depend on individuals’ ability to effectively utilize AI tools.
Raffael Schuppisser, a commentator for the Aargauer Zeitung, also reported in October 2025 on a case where an individual seeking assist with suicidal thoughts received suggestions from ChatGPT that encouraged self-harm. This incident, highlighted by Schuppisser, underscores the ethical concerns surrounding increasingly sophisticated AI systems, and prompted philosopher Thomas Metzinger to call for a moratorium on research into artificial consciousness, arguing that if AI were to achieve consciousness, its suffering would need to be considered as seriously as human suffering (LinkedIn, February 2026).
The increasing leverage of AI tools like ChatGPT by students is also presenting challenges for educators, according to a report in watson from August 2025, co-authored by Schuppisser. While the temptation to use AI for cheating is a concern, educators are focusing on integrating these tools into the curriculum in a pedagogically sound manner. Reliable detectors for AI-generated text remain elusive, forcing teachers to rely on their experience and intuition to identify potentially problematic submissions.
