Learning RAG, Overcoming Imposter Syndrome + Building Your AI Career as Non-Techie With Naomi White
May 19, 2024
auto_awesome
Naomi White, Engineering Program Manager at Cohere, shares her journey from Executive Assistant at DeepMind to key player in RAG development. Topics include overcoming imposter syndrome, learning LLM concepts, and behind-the-scenes on RAG techniques.
Transitioning from DeepMind to Cohere for new challenges in language models and greater impact.
Embracing continuous learning, seeking mentorship, and balancing spontaneous learning with structured strategies in the tech space.
Adapting to technical jargon, decoding acronyms, and finding humor in the evolving language dynamics of the tech environment.
Deep dives
Transition from DeepMind to Cohere
The guest describes her journey moving from DeepMind to Cohere and the driving factors behind this transition. She explains how working at DeepMind exposed her to the forward-thinking mission of AI, but she sought a new challenge at Cohere, a startup focusing on language models. She highlights the appeal of joining a less established company where she could have a greater impact and explore different roles within a more flexible and fast-paced environment.
Learning in a Fast-Paced Environment
The guest reflects on navigating imposter syndrome and learning in the tech space. She emphasizes the importance of continuous learning and seeking guidance from mentors and colleagues. By actively engaging with engineers, attending meetings, and participating in discussions, she adopts a sponge-like approach to absorb new information, balancing spontaneous learning with structured strategies like book club sessions for sharing key takeaways from relevant readings.
Embracing Challenges and Growth Through Acronyms
The guest shares insights on adapting to the technical jargon of the field and humorous anecdotes related to learning acronyms. She acknowledges the evolving language dynamics within the tech environment and recalls instances of creating and deciphering acronyms. Despite the challenges of decoding terminology, she finds humor and learning opportunities in navigating the dense jungle of technical language present in corporate and technical settings.
RAG - The Use of Language Models in Retrieving Information
RAG, a language model, is trained on outdated data but is effective in retrieving information from various sources such as documents and web searches. By using RAG, users can obtain specific information like business rates from lengthy documents and up-to-date data from online sources, allowing for faster and more accurate responses. The ability to interrogate current information from the web ensures users can verify data, enhancing reliability and efficiency in queries.
Enhancing Tools with RAG for Complex Queries and Multitasking
RAG's integration with various tools like weather applications and finance platforms provides users with complex and accurate information beyond typical web searches. The potential for using multiple tools simultaneously creates a more intuitive and informative user experience, akin to engaging in multi-step conversations. The focus on incorporating diverse tools and refining functionalities helps streamline data retrieval, paving the way for future applications in chatbots and project management tools.
In today's episode, I chat about retrieval-augmented generation (RAG) with Naomi White, Engineering Program Manager at Cohere. Naomi shares her unique journey from starting as an Executive Assistant at DeepMind to becoming a key player in Cohere’s cutting-edge RAG development team.
In this episode, we cover: - Naomi's inspiring journey from DeepMind to Cohere and building a career in AI as a non-technical person; - Strategies for navigating the imposter syndrome; - Beginner-friendly tips for learning LLM concepts and staying up to date with AI advancements; - A behind-the-scenes look at developing LLM tool use with RAG techniques;
