Open Source Models and Developer Education with Jacob Marks of Voxel51
Dec 21, 2023
auto_awesome
Jacob Marks, Machine Learning Engineer and Developer Evangelist at Voxel51, discusses the hype of AI in computer vision and software development. They highlight the importance of practical learning resources, code reusability, and engaging the developer community. They also talk about building a chat bot with GPT 3.5, the value of foundational models in computer vision, and the potential of using synthetic data. Finally, they explore the role of emoji in emotional expression and mention Andrei Carpathy as an influential figure in science and tech.
Building robust systems that others can interact with is as important as focusing on simulations in algorithmic challenges.
Democratization of developer tooling has made it possible to build systems and applications that were previously unthinkable for many developers.
Deep dives
Jacob Marks' unconventional path into coding and developer relations
Jacob Marks, a machine learning engineer and Developer Evangelist at Voxel51, shares his unconventional journey into coding and Developer Relations. With a background in physics, Jacob gained programming experience during his PhD research, which sparked his interest in algorithmic challenges. Furthermore, his passion for education and interpersonal engagement led him to pursue a career in Developer Relations. He highlights the importance of not only focusing on simulations but also building robust systems that others can interact with.
The differences between scientific research and software development
Jacob discusses the differences between scientific research and software development. In scientific research, the main focus is on proof of concept, testing ideas, and generating results, while software development requires a more rigorous approach to handle edge cases, write reusable code, and make systems accessible to others. He emphasizes the importance of thinking about how others can be enabled by what you are building.
The democratization of developer tooling
Jacob highlights how developer tools and resources have become more accessible, thanks to the democratization of developer tooling. He mentions the availability of open-source models, like GitHub co-pilot and chat TPT, and the ease of fine-tuning language models without extensive machine learning knowledge. These advancements have made it possible to build systems and applications that were previously unthinkable for many developers.
The challenges and benefits of using GPT for documentation search
Jacob explains the process of building a chatbot for documentation search using GPT. He discusses the challenges of ensuring the model understands the specific documentation context, including the introduction of custom prompts and post-processing techniques. Jacob also explores the benefits of documentation search tools, like semantic search and chatbots, in making technical content more accessible and interactive for users.
Open Source Models and Developer Education with Jacob Marks of Voxel51
In this episode, Jon talks to Jacob Marks, Machine Learning Engineer and Developer Evangelist at Voxel51, whose background is rooted in scientific research having formally been a Ph.D. resident at X the moonshot factory.
In this episode, Jacob and Jon unpack the current hype surrounding the implementation of AI in computer vision and software development, assessing its current impact on the industry. Join them as discuss the importance of practical and interactive learning resources and the benefits of contributing to open-source projects.