Is it just me or GPT 5 (Thinking) got lobotomized?

Examining Recent Changes in GPT-5’s Cognitive Capabilities: A Personal Observation

In the rapidly evolving landscape of artificial intelligence, large language models like GPT-5 have become invaluable tools for tasks ranging from coding assistance to comprehensive research. However, recent interactions have led me to question whether there have been significant shifts in GPT-5’s reasoning abilities.

Over the past month, I’ve noticed that GPT-5’s performance in tasks requiring creative thinking and complex problem-solving appears to have diminished. Tasks that previously elicited insightful and innovative responses now seem to challenge the model more than before. This perceived decline has prompted me to explore alternative solutions, such as utilizing specialized tools like Grok Expert, which often yield more satisfactory results for complex or creative inquiries.

While these observations are anecdotal, they highlight an important aspect of AI model development: maintaining a balance between safety, usability, and cognitive flexibility. It raises questions about whether recent updates or policy changes may have inadvertently constrained GPT-5’s expressive and reasoning capacities.

As AI enthusiasts and professionals, it’s crucial to monitor these developments closely. Continuous feedback and empirical evaluation help developers optimize models for diverse applications, ensuring they remain powerful, reliable, and adaptable tools for the future.

If you’ve experienced similar observations or have insights into recent updates affecting GPT-5’s capabilities, I invite you to share your thoughts. Together, we can contribute to a deeper understanding of the evolving landscape of AI intelligence.