Unveiling the Capabilities and Risks of GPT-4.

OpenAI's latest AI language model, GPT-4, is the most advanced yet, but the company is keeping its training data a secret, citing competitive and safety considerations. This lack of transparency is concerning to experts who warn that without knowing where the data comes from, it's harder to understand where the worst harms lie. OpenAI has acknowledged that GPT-4 has various biases in its outputs, but researchers outside the company would not know where those biases come from. Critics say OpenAI is willfully ignoring basic risk mitigation strategies and warn that the company is getting too close to becoming just like every other big tech company.