Stack Overflow’s Industry Guide to AI
2024
Introduction
Our team worked with Stack’s Demand Generation and Content Marketing teams to develop a new, comprehensive microsite on how to understand the AI landscape and upskill teams for the GenAI era.
Our goal was to create a dynamic and focused piece of content that further pushed Stack Overflow into the AI space as a thought leader and source of knowledge.
UI Design
We began by exploring inspirations for similar content and creating a layout that would support a long-form, multi-chapter guide. We wanted to prioritize focus and clarity, meaning a pared-back design approach with only the most necessary CTAs and exit opportunities.
This brought us to a basic wireframe, demonstrating for other stakeholders in our project what our goals were and how far from our typical approach we would like to go. This helped us demonstrate early into the project the kind of user experience we want to build, especially as the content and copy were being developed concurrently.
Art Direction
As we finalized copy, and our wireframes took shape, I began iterating on a visual language for this project.
I wanted something representative of the recursive and exponential properties of AI growth — and it had to be a scaleable approach that could live as a low-weight vector in the final product. I developed a practice of warping geometric shapes according to specific rules — point (x) and point (y) will move some unit of distance with each successive shape, or the entire shape will shift (x) unit left and grow proportionally by (y) percent. After repeating, these shapes create an entirely new form.
Illustration
Illustrations within chapters, compared to the art direction for hero graphics, had to be much less abstract. I focused on helping readers understand the complex topics detailed throughout the guide.
Fine Tuning
Retrieval-Augmented Generation (RAG)
Knowledge Management and Answer Quality
Invasive Data Collection
Generative Adversarial Networks (GANs)
HarveyAI: Research Processing for Legal
Momentum of Neural Network Growth
Iterating and Improving
Customization in IDEs
Debugging
Semantic Search
AI Impact on DEI
From Chatbots to LLMs
Providing Chatbot Feedback
Intellectual Property (IP) Concerns
Tailoring Your Chatbot
Explainability
HarveyAI: Data Ingestion for Medical
Debiasing
Open-Source Models
Model Drift
Khan Academy: Personal AI Tutors for Socratic Learning
Model Size and Performance
Bloomberg: NLP Tasks With BloombergGPT
Ethics and AI
Query Tokenizing
Cost of AI
Upskilling Teams
Limitations of Tailored Chatbots
Conclusion
This project unified our brand, content, and demand-gen teams to create a type of content we hadn’t seen before. This really helped me push my skills in project leadership, owning the design and iteration phase in the latter half of this process, and all of our hard work brought this to an extremely polished, high-impact launch.