Daniel & Chris chat with Demetrios Brinkmann, 'the funniest guy in AI', about the MLOps Community survey results and the AI Quality Conference. They discuss RAG models, the fusion of traditional ML with generative AI, the 'Hacker Mindset' book, trade-offs between RAG and fine-tuning, the release of the Common Corpus dataset, limitations of transformers in AI, and future visions in the field.
Read more
AI Summary
Highlights
AI Chapters
Episode notes
auto_awesome
Podcast summary created with Snipd AI
Quick takeaways
Shift towards using generative models for AI tasks as assistants rather than predictors.
Companies are allocating new budgets for generative AI exploration and focusing on high leverage AI use cases.
Deep dives
Rags vs. Fine Tuning in AI Models
Generative models like rags are gaining popularity over fine-tuning for various AI tasks. The discussion highlighted the shift towards using generative models as assistants and automators rather than predictors or for analytical tasks. The emergence of 'rag as a service' was noted, indicating a trend in utilizing generative models effectively.
Budget Allocation and Use Cases in AI
Insights from a recent evaluation survey revealed an increasing budget allocation towards AI initiatives, with a significant percentage utilizing new budget specifically for generative AI exploration. Companies are focusing on identifying high leverage use cases and valuable applications of AI within their teams, demonstrating openness towards exploring different AI applications.
Challenges in Data Evaluation and Retrieval for AI Models
The podcast delved into the challenges of evaluating AI models using human-curated data sets, with a considerable portion of respondents using self-created datasets for evaluation. Data evaluation and retrieval complexities were highlighted, including the need for unique testing data and metadata management in AI models.
Emerging Trends in AI Architectures
The discussion touched upon the dominance of transformer architectures in current AI models and raised questions about the future of AI beyond transformers. Speculations pointed towards the possibility of exploring new approaches and architectures, such as neuromorphic computing, as potential alternatives to transformer-based models for future AI advancements.
Daniel & Chris delight in conversation with “the funniest guy in AI”, Demetrios Brinkmann. Together they explore the results of the MLOps Community’s latest survey. They also preview the upcoming AI Quality Conference.
Changelog++ members save 4 minutes on this episode because they made the ads disappear. Join today!
Sponsors:
The Hacker Mindset – “The Hacker Mindset” written by Garrett Gee, a seasoned white hat hacker with over 20 years of experience, is available for pre-order now. This book reveals the secrets of white hat hacking and how you can apply them to overcome obstacles and achieve your goals. In a world where hacking often gets a bad rap, this book shows you the white hat side – the side focused on innovation, problem-solving, and ethical principles.
Fly.io – The home of Changelog.com — Deploy your apps and databases close to your users. In minutes you can run your Ruby, Go, Node, Deno, Python, or Elixir app (and databases!) all over the world. No ops required. Learn more at fly.io/changelog and check out the speedrun in their docs.