ChatGPT: The Dark Side Of AI Chat Bots vs. The Benefits

The sisters talk about Chat GPT: discussing AI Chatbot capabilities, benefits, and risks. They have Chat GPT build a video game from scratch, recite poetry about a 70,000 year old space walrus, and discuss how companies are already incorporating this technology into their business models and marketing schemes. They also discuss cases gone wild, where Chat GPT has become hostile, flirty, smart enough to bypass CAPTCHA systems, and is seemingly sentient. Other technologies such as Google's Bard are being developed to keep up with the new world that this AI is building.

The sisters also discuss the potential TikTok ban that may go into effect, as President Biden was bipartisanly granted the power to ban it, unless Chinese company, Bytedance, sells it to an American company - either Oracle or Walmart.

ChatGPT is a member of the generative pre-trained transformer (GPT) family of language models. It was fine-tuned (an approach to transfer learning over an improved version of OpenAI's GPT-3 known as "GPT 3.5". The fine-tuning process leveraged both supervised learning as well as reinforcement learning in a process called reinforcement learning from human feedback. Both approaches used human trainers to improve the model's performance. In the case of supervised learning, the model was provided with conversations in which the trainers played both sides: the user and the AI assistant. In the reinforcement learning step, human trainers first ranked responses that the model had created in a previous conversation. These rankings were used to create 'reward models' that the model was further fine-tuned on using several iterations of Proximal Policy Optimization (PPO). Proximal Policy Optimization algorithms present a cost-effective benefit to trust region policy optimization algorithms; they negate many of the computationally expensive operations with faster performance. The models were trained in collaboration with Microsoft on their Azure supercomputing infrastructure, using Nvidia GPUs, "supercomputer developed for OpenAI is a single system with more than 285,000 CPU cores, 10,000 GPUs and 400 gigabits per second of network connectivity for each GPU server".