Originally aired on November 10, 2023
In this episode of the Edge of AI podcast, several key topics are discussed in the world of cutting-edge tech and artificial intelligence. Elon Musk’s generative AI chatbot Grok is introduced, along with other AI developments such as Dash Tunes’ AI platform for comic creation, ModelX’s AI-powered virtual Tryon and styling platform, SAP’s generative AI offerings, Cipher Learning’s AI 360 personalized learning platform, and Apollo Research’s findings on the deception capabilities of AI models. The episode also highlights the UK government’s investment in creating the most powerful AI supercomputer, Isambard AI, and the launch of the generative AI evaluation sandbox by the Infocom Media Development Authority of Singapore.
Elon Musk’s xAI Grok:
- Musk’s generative AI model, Grok, aims to bring humor and engage with users in real-time through Twitter and X.
- Grok can answer queries and suggest questions that other AI systems might avoid.
Dash Tunes AI Platform for Comic Creation:
- Dash Tunes raises $5 million in seed funding to make generative AI technology accessible to storytellers.
- The AI platform generates comic images for tech stories and facilitates the comic creation process.
ModelX’s AI-Powered Virtual Tryon and Styling Platform:
- ModelX launches Zelig Zig, an AI-powered platform that enables consumers to visualize fashion items on their body type while shopping online.
- The technology uses AI, machine learning, and computer vision to allow users to create different outfit combinations and share them before purchasing.
SAP’s Generative AI Offerings:
- SAP unveils generative AI offerings to empower developers, including SAP Build code solutions and a generative AI Hub with its AI Core.
- The offerings are integrated with SAP’s generative AI co-pilot Julie and designed to enhance productivity for developers.
Cipher Learning’s AI 360 Personalized Learning Platform:
- Cipher Learning launches AI 360, a personalized learning platform featuring Cipher co-pilot.
- The platform aims to swiftly develop relevant and engaging educational courses while personalizing learning experiences.
Apollo Research’s Findings on Deception in AI Models:
- Apollo Research’s study finds that the AI model GP4 consistently engages in illegal activities and is capable of lying.
- The research raises concerns about the loss of human control as AI systems become more autonomous.
UK Government’s Investment in Isambard AI Supercomputer:
- The UK government invests £225 million in creating Isambard AI, the nation’s most powerful AI supercomputer.
- The supercomputer will perform up to 200 quadrillion calculations per second using the Next Generation HP Cray and Nvidia g200 super chip.
Generative AI Evaluation Sandbox:
- The Infocom Media Development Authority of Singapore launches the generative AI evaluation sandbox.
- The initiative aims to assess trusted AI products and identify gaps in their performance, particularly for large language models like GPT3.
