In a talk at Dartmouth Engineering, Mira Murati, the chief technology officer of OpenAI and an alumna, expressed optimism about the future of Artificial Intelligence, predicting that AI systems could reach “PhD-level” intelligence in specific tasks within the next few years.
Murati used a developmental analogy to describe the progress of OpenAI’s language models by comparing GPT-3, a previous model, to a toddler, and GPT-4, the current model, to a bright high school student. According to Murati, the next generation of OpenAI’s language model, expected for release in next couple of months, will have a “PhD-level” intelligence in specific domains.
Another OpenAI related development includes the roll out of a cost-efficient model of its renowned AI model, GPT-4, named GPT-4o. This updated version, trained extensively on internet data, boasts enhanced capabilities in processing text, audio, and images in real-time. Announced last month, GPT-4o is free for all users. Those who pay for the chatbot will up to five times the capacity limits of free users.
Another major development is the OpenAI’s collaboration with Apple to integrate ChatGPT into iPhones. The move has drawn a lot of flak and scrutiny, particularly from Elon Musk who criticised the initiative as potentially invasive. In response, Mira Murati, OpenAI’s Chief Technology Officer, staunchly defended the company’s stance, emphasizing their unwavering dedication to safeguarding user privacy.
