In other news, OpenAI has conducted a study to assess whether its large language model, GPT-4, could assist users in creating harmful biological weapons. The study found that the model provides at most a mild uplift in biological threat creation accuracy. Meanwhile, Amazon has launched an AI-powered chatbot, Rufus, to assist customers with shopping decisions. The bot is trained on Amazon's product catalog and information from across the web to answer customer questions and make recommendations.
Key takeaways:
- An Australian news station blamed an AI tool in Adobe Photoshop for altering a photo of a politician, claiming it was a "graphics error". However, Adobe stated that any changes would have required human intervention and approval.
- OpenAI conducted a study to assess whether its large language model, GPT-4, could potentially assist users in creating harmful biological weapons. The results suggested that the model provides at most a mild uplift in biological threat creation accuracy.
- Amazon has released a new AI-powered chatbot called Rufus to assist customers in making purchasing decisions. The bot is trained on Amazon's product catalog and information from across the web.
- A survey by EduBirdie found that Gen Z is using AI and ChatGPT in the workplace for tasks such as research. However, there are concerns about job security, with 61% believing AI could take their job within a decade.