Powered by Data - August 28, 2023

Data Points

Last Week in Conversational AI

• Alibaba has introduced two new artificial intelligence models, Qwen-VL and Qwen-VL-Chat, capable of understanding images and engaging in more complex conversations. These open-source models allow researchers, academics, and companies globally to use them for AI applications without training their systems. Qwen-VL can respond to open-ended queries related to images and generate captions. At the same time, Qwen-VL-Chat handles more intricate interactions, like comparing multiple image inputs and answering various rounds of questions. It can even write stories, create images based on input photos, and solve mathematical equations shown in pictures.
  Read more
• The University of Texas is working on an AI chatbot to help women dealing with postpartum depression. In partnership with Postpartum Support International, the university is conducting tests on a free app that offers access to the chatbot. The chatbot is trained to address common concerns and questions that new mothers may have, such as making connections and breastfeeding hurdles. This project aims to assist new mothers in obtaining mental health care, as they frequently face obstacles in accessing support due to societal stigma and other barriers. The University of Texas intends to make the AI chatbot accessible to women by 2024.
  Read more
• A study conducted by researchers from Stanford University found that an artificial intelligence (AI) chatbot provided comparable advice to ophthalmologists for various patient questions related to eye care. The study analyzed human-written and AI-based responses to 200 eye care questions, finding that the AI chatbot generated appropriate answers similar to those of ophthalmologists regarding accuracy and potential harm. The study used ChatGPT, a famous large language model (LLM) chatbot, to provide responses to questions. The researchers suggested that while AI-powered tools are not meant to replace human ophthalmologists, they could augment their work and provide patient education support under appropriate supervision.
  Read more
• A new AI chatbot app called "Text With Jesus" has been developed, allowing users to engage in text conversations with biblical figures such as Jesus, Mary, Joseph, Peter, and Matthew. The AI app provides responses based on how these divine characters might have answered. However, the app has also sparked controversy due to its feature that enables users to simulate conversations with Satan. While interactions with holy figures can be spiritually uplifting, conversations with malevolent figures like Satan have raised concerns about potential deception. The app is developed using Bible teachings and offers a premium version with features like engaging dialogues with narratives from the Bible. Interacting with Satan is disabled by default and requires manual activation.
  Read more