A quick chat with Ines & Matt

2026-06-19 · Source: Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, quick

Summary

Ines Montani, Co-founder and CEO of Explosion and creator of spaCy, along with Matthew Honnibal, discussed "vibrant LP" workflows at their Py4AI session. Their presentation focused on using large language models (LLMs) and agentic coding to develop machine learning systems, contrasting this approach with traditional "software 1.0" explicit coding. They highlighted "software 2.0" as systems where behavior is encoded through data examples rather than explicit code, posing the challenge of how agentic systems can generate this type of software. This topic is central to a new product Explosion is developing, building on their existing suite of natural language processing tools. Montani also shared insights on designing engaging presentations that simplify complex technical content and challenge prevailing trends.

Key takeaway

For Machine Learning Engineers developing new systems, consider integrating agentic coding with large language models to build "software 2.0" that encodes behavior through data examples. This approach moves beyond traditional explicit coding, offering a new paradigm for system creation. Additionally, when presenting your technical work, focus on designing visually engaging slides and simplifying complex arguments to ensure your message is easily digestible and impactful for your audience.

Key insights

Agentic systems can generate "software 2.0" by encoding behavior in data examples, not just explicit code.

Principles

• Machine learning systems are "software 2.0," encoding behavior via data.
• Agentic coding can build ML systems, moving beyond explicit code.
• Effective presentations simplify complex technical arguments.

In practice

• Explore agentic coding with LLMs for ML system development.
• Design custom, visually engaging slides for technical talks.
• Break down complex technical content into digestible arguments.

Topics

• Large Language Models
• Agentic Coding
• Machine Learning Systems
• Natural Language Processing
• spaCy
• Technical Presentations

Best for: NLP Engineer, Machine Learning Engineer, AI Engineer

Related on AIssential

• Vibe coding and agentic engineering are getting closer than I'd like — The increasing reliability of AI coding agents blurs the distinction between casual "vibe coding" and professional "agentic engineering."
• An AI agent coding skeptic tries AI agent coding, in excessive detail — Modern AI agents, guided by explicit `AGENTS.md` files, can generate highly optimized, performant code across multiple languages.
• From agents to world models: What San Francisco revealed about AI’s next phase — AI is transitioning from linguistic models to agentic systems that act, use tools, and comprehend physical and virtual environments.
• Claude Opus 4.8: not ambitious enough? — The model's output, while functional, lacked ambition and failed to push agentic coding boundaries despite explicit prompting.
• 10 Lessons for Agentic Coding — Agentic coding is transforming software development, requiring new approaches and skills.
• How I'm Adopting Agentic Coding — Generative AI tools like ChatGPT can significantly transform coding workflows and information retrieval.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai.