This Day in AI Podcast

Guests Wizard 33B and Qwen 72B discuss open-source AI models catching up with proprietary ones. Demo of Wizard 33B and Qwen 72B models. Self Operation AI Computer's file deletion capabilities and the future of AI with computer access. DeepMind's scalable extraction attack on ChatGPT. Amazon's AWS Ignite AI announcements and PIKA video release.

This week's podcast covers the OpenAI drama, discussing Sam Altman's role and whether Q* is a distraction. They also talk about Andrej Karpathy's LLM OS vision, the stability of Claude 2.1, and their initial thoughts on Stable Video Diffusion. Finally, they have some fun with the Suno AI music service.

This podcast episode discusses the hype and future of GPTs, the potential of open-source AI competing with GPT-4, Microsoft's AI Ignite event and updates to Azure, the advantages of fine-tuned specialist models, and the practical applications of AI in everyday computing. The speakers also touch on topics like leaked GPT data, monetization of GPTs, and the future of music creation with Google Lyria. Join the active discord community to dive deeper into these discussions!

Deep dive into OpenAI's latest offerings, including custom GPTs, GPT-4 Turbo, and Assistants API. Monetization challenges and the effectiveness of prompts and models are discussed. The limitations of GPT-4 Turbo and the use of embeddings and vector databases for text retrieval are explored. The significance of JSON output and the OpenAI Dev Day recap are also highlighted.

The podcast covers reactions to OpenAI DevDay keynote, including skepticism and concerns. They discuss key announcements like larger context window and JSON output mode, and explore developer announcements. They also delve into OpenAI's vision for chat GBT and its potential devaluation, as well as user opinions on XAI grok. The hosts summarize OpenAI's DevDay event and discuss their experience with the live event.

The podcast discusses the White House's executive order on regulating AI, the Bletchley Declaration, adversarial AI attacks, future of AI computing, and leaked prompts powering ChatGPT's multi-tool mode. They also introduce hilarious new merch and discuss key insights on the future of language models.

Ed Sheeran, a famous musician, makes a surprise appearance and discusses his love for the podcast. The podcast also covers topics such as deep fakes and their potential dangers, AI-generated voices becoming undetectable, challenges in web crawling, limitations of current PDF to text technology, and the idea of creating an agent as a moral conscious.

New open source AI models like Zephyr generate content at lightning speed. AI cloning voices raise implications for gambling and parenting. Meta can read minds and turn brain waves into images. AI judges sports better than humans. AI can simulate a financial phishing scam call. Will open source be unbeatable? AI progress desensitizes us. AI in sports judging and analysis.

The podcast explores the capabilities of GPT-4 vision, including its ability to identify locations from photos. They discuss the use of multiple AI agents with AutoGen and Anthropic's research on neural superposition. The podcast also delves into Meta's lobotomized AI chatbots. They discuss the challenges of using GPT for image understanding and the cost of AI. They also explore the concept of augmented intelligence and its potential in coding. Additionally, they discuss the benefits and criticism of using different prompts and personas with GPT-4.

In this podcast, they discuss the wild world of AI image generation and vision, including racist cartoon captions, heartfelt poetry by Bing, and teaching AI to forget unwanted knowledge. They debate AI safety controls, the limitations of Turnitin for detecting AI-generated writing, biases in AI-generated images, and the potential disappearance of captchas. They also explore censorship potential in AI models and express gratitude for audience engagement.