GPT-6 Astra is creating debate about whether AI is getting closer to human-level intelligence. OpenAI’s new AI model can use computers, browse websites, do research, write and test software, analyse data and complete several tasks with limited human help. It has also performed strongly in tests related to mathematics, cybersecurity and computer use. However, these results do not mean that AI has already reached human-level intelligence. Some of the test results depend on how the tests were conducted, and Astra does not perform better than every other AI model in all areas. Its biggest improvement is its ability to work more independently. It can set out to complete a task, use different tools, learn from mistakes and continue working. AI is already better than humans at some specialized tasks, but human intelligence involves many abilities, including common sense, adaptability, social understanding and handling unexpected situations.
GPT-6 Astra: Can AI Really Surpass Human Intelligence?
