GPT-5: Proceed with Caution

OpenAI's GPT-5 demonstrates advancements over previous models but still exhibits a tendency to 'hallucinate' information. Despite improvements, the model can generate incorrect or fabricated details in approximately 10% of its responses. This tendency raises concerns about its reliability as a sole source of information.

Nick Turley, Head of ChatGPT, advises users to utilise GPT-5 as a supplementary tool rather than a primary reference. He suggests cross-referencing its outputs with trusted sources to ensure accuracy. While GPT-5 shows promise, its imperfection necessitates a cautious approach, particularly in critical decision-making contexts. The AI model is better when paired with something that provides a better grasp of the facts, like a traditional search engine or a company's specific internal data.