Microsoft AI, model routing, and the move toward portable workflows

Full Show Notes: https://www.theintelligenceagepodcast.com/845

Mark Smith and Joris de Gruyter, who has spent four years deep in the Microsoft AI ecosystem, discuss where AI has become useful recently and where the rough edges remain.

The conversation centers on model choice, vendor lock-in, local versus cloud inference, and how Markdown and open standards are becoming a portability layer for notes, skills, and AI workflows.

Key topics

• Joris says one of the biggest shifts has been the move from coding agents like Codex and Claude Code into non-coding work for information workers.
• He explains that tools like Microsoft Co-work and Copilot can be powerful, but users need enough technical savvy to get the most out of them.
• Mark and Joris compare cloud-first AI workflows with local, device-based workflows, especially for browser work, Playwright, and tasks requiring local access.
• They discuss Microsoft’s effort to make Co-work feel more like a first-class Microsoft product, with broader model support.
• Both argue that model choice matters less than before; skills, prompts, and workflow design are becoming the real intellectual property.
• They explore Markdown as a portable format for AI-era data, and why Microsoft supporting Markdown in SharePoint matters.
• Joris explains how OneNote, Obsidian, and Loop each solve part of the note-taking problem, but none fully combine syncing, portability, and API access.
• Mark describes using AI to clean up old data, emails, and archives rather than preserving everything indefinitely.
• They discuss vendor lock-in and the need for open standards that let people switch models and services.
• Joris makes the case for local inference as open-weight models improve and consumer hardware gets better.
• They close with a critique of major technology companies, arguing that strong hardware has not yet been matched by software and AI integration that meets user expectations.

If you want to get in touch with me, you can message me here on Linkedin.

Thanks for listening 🚀 - Mark Smith

Te Reo Māori, translation, and using AI without losing human meaning

Full Show Notes: https://www.theintelligenceagepodcast.com/844

Mark Smith speaks with Tane Karamaina, a senior translator working across Te Reo Māori and English, about how language, culture, and AI intersect in real-world translation work.

The conversation explores why translation is never just word-for-word substitution, how AI can support but not replace human judgment, and the risk of homogenizing dialects as language technology improves.

In this episode, Tane shares practical examples from translation work, including punctuation, implicit meaning, and culturally specific phrases like “O for awesome.”

The discussion also covers pronunciation, regional dialects, ancestral intelligence, and how AI in New Zealand could be used responsibly as an assistive tool rather than a substitute for human expertise.

Key topics

• Tane Karamaina introduces himself through a pepeha and explains the importance of naming whakapapa, place, and identity.
• He describes his work as a senior translator with more than 10 years of experience translating between English and Te Reo Māori.
• Translation is framed as cultural mediation, not word-for-word conversion.
• AI is presented as a useful tool for translators, but not a replacement for human understanding, interpretation, and responsibility.
• Tane explains how he uses AI for simple support tasks like punctuation and breaking down source text structure.
• The episode highlights how meaning is often implicit, culturally bound, and not directly transferable across languages.
• A deeper discussion compares Te Reo Māori and English through examples like “language” and “moana,” showing how words carry different worldviews.
• Mark and Tane discuss parallels between Te Reo Māori and Japanese, especially respect, vocal tone, and open syllable structure.
• Tane raises concerns about AI-driven language standardization flattening regional dialects and local variation.
• He argues the best protection against homogenization is for people to keep speaking and transmitting their ancestral dialects.
• The conversation closes on the idea that AI in New Zealand can be positive if it supports, not replaces, human beings and their communities.

If you want to get in touch with me, you can message me here on Linkedin.

Thanks for listening 🚀 - Mark Smith

AI, context, and New Zealand’s data center future

Full Show Notes:
https://www.theintelligenceagepodcast.com/843

Mark Smith speaks with Oliver Hartwich about how AI has changed since January, why agentic tools have become the real shift, and what that means for workflows, context, and cross-model collaboration. They also dig into New Zealand’s position in the global AI and infrastructure landscape, especially around data centers, energy, solar, batteries, and water use.

Key topics

• In this episode: Oliver says the biggest change in 2026 so far has not just been better models, but the rise of agentic AI tools like Claude Code, ChatGPT work modes, and Perplexity Computer.
• He explains why AI now feels less like a novelty and more like a practical, increasingly capable working layer, especially when multiple models are used together and checked against each other.
• Mark raises the concern that New Zealand may be vulnerable because AI compute and data may not stay in-country across the major hyperscalers operating locally.
• Oliver discusses the post Fable moment, when access uncertainty highlighted how dependent users have become on frontier tools and how quickly expectations shift across providers.
• The conversation turns to context management across platforms, including when to preserve long-term memory and when to use a fresh AI for independent review.
• Oliver describes building a personal Claude Skill from his own writing archive, compressing books, reports, articles, and newsletters into markdown so the system can reflect his style and past thinking.
• They discuss using one AI to peer review another AI’s work, including ping ponging a skill between Claude and ChatGPT to improve quality and completeness.
• Oliver shares how Codex can be used for computer control and debugging by going directly into system settings and config files instead of relying on manual UI hunting.
• The discussion moves to MCP-style integrations, including using Site CITE AI to connect large academic literature databases into AI workflows for faster research and self peer review.
• Mark and Oliver compare the old PhD research process with the newer AI-assisted version, where initial literature collation can be compressed from roughly a year to a few weeks.
• They reflect on the tradeoff between speed and thinking time, arguing that the slower, manual research process also created space for reflection, sleep, and deeper synthesis.
• The final section focuses on New Zealand data centers, geothermal energy, grid stability, solar adoption, battery storage, and the challenge of building infrastructure with social license.

Resourcess:
Oliver Hartwich's essay on AI in education: https://oliverhartwich.com/2026/06/25/bildung-and-the-machine/

If you want to get in touch with me, you can message me here on Linkedin.

Thanks for listening 🚀 - Mark Smith