OpenAI just released a powerful deep research. But hallucinations remain a challenge. If it’s just numbers, the errors might be easier to spot. But as AI tackles more complex queries, verifying its outputs becomes harder.
What happens when the human user lacks the expertise to