Futura, a scientific journalism outlet, recently put ChatGPT, an artificial intelligence created by OpenAI, to the test. The results were mixed, with ChatGPT demonstrating the ability to adapt to language and provide accurate information on scientific topics, but also producing repetitive and lengthy responses.

During the test, the team at Futura asked ChatGPT to introduce itself as if it were a science journalist named “Futur.IA.” The AI responded with a brief, but satisfactory introduction that highlighted its capabilities as a source of information and problem-solver. However, when prompted to expand on its introduction, ChatGPT’s response became repetitive and overly verbose.

Despite these limitations, the team at Futura was impressed with ChatGPT’s potential as a tool for scientific journalism and its ability to understand and answer a wide range of questions on various scientific subjects. The team believes that with further development and refinement, ChatGPT could become a valuable asset for providing accurate and up-to-date information on a variety of scientific topics.

The interface with ChatGPT. © OpenAI