LLMs and Generative AI: Increasing Productivity with Modern Tooling

This talk explores the potential of Artificial Intelligence in modern development workflows, showcasing practical examples and concepts for harnessing Generative AI and Large Language Models (LLMs) to streamline daily tasks. Learn how to utilize existing tools and create custom solutions to automate cumbersome processes, and discover how AI-powered tooling can enhance your development experience.

Transcript

1. LLMs and Generative AI: Increasing Productivity with Modern Tooling iJS Fullstack Day Max Marschall @MaxOSchulte Consultant / Developer
2. Arti fi cial Intelligence Machine Learning Data Science NLP A Deep Learning Generative AI Large Language Models Image / Video Generation
3. Bla bla bla … Intention
4. Tools & Functions Process Think Display Bla bla bla …
5. Tools & Functions Process Think Display Bla bla bla …
6. Large Language Model • Transformer Neural Network • Context & relation aware • Pattern matching & recognition • “Natural” language understanding • Generating / creating content • Summarisation
7. Large Language Model • It cannot make decisions • It cannot understand non-linear content • It cannot draw conclusions
8. Large Language Model • Not deterministic • Can hallucinate • Prompt-Injection
9. Result Bla bla bla …
10. Result = Bla bla bla …
11. Tools & Functions Process Work Display Bla bla bla …
12. Agents
13. Evolving Landscape Copilot Cline Cursor Webstorm avante.nvim CodeCompanion Obsidian And many more…
14. What to use it for? • Getting started • Automate the “easy” stuff • Bridging a gap
15. • Getting started • Automate the “easy” stuff • Bridging a gap What to use it for?
16. • Getting started • Automate the “easy” stuff • Bridging a gap What to use it for?
17. • Getting started • Automate the “easy” stuff • Bridging a gap What to use it for?
18. • Getting started • Automate the “easy” stuff • Bridging a gap What to use it for?
19. • There’s an integration “for you” • Fast paced • “Wild West” • Better, faster, smaller models weekly What now?
20. It's a wrap! • Slides: https://thinktecture.com/max-marschall • Contact: @MaxOSchulte, max.marschall@thinktecture.com