Ikigai AI Ventures
AGSM MBA · GUEST SESSION

AI Landscapes

Guest lecture

Lucas Zhu
Founder, Ikigai AI Ventures · Guest lecturer, AGSM UNSW
Thursday 24 September 2026 01 / 06
About me

I build and run AI agent systems inside my own businesses, every day

Ikigai AI Ventures, Sydney. Venture builds, coaching for executive teams, and advisory on where AI creates value.

Lucas Zhu
Lucas Zhu Founder · Ikigai AI Ventures

Builds

Frollie OS, an operating system for an FMCG business, built solo on Claude Code. A live business runs on it, fast growing and profitable, with over 40 business processes on agentic workflows. Open-sourced (MIT).

Before this

Seven years at McKinsey (Senior Engagement Manager). Director of Strategy at HelloFresh ANZ, a $1B+ ARR business.

Teaching

Guest lecturer at AGSM, UNSW on AI strategy. Essays and talks on agentic engineering at ikigaiventures.ai.

Portfolio
ikigaiventures.ai/portfolio
https://www.linkedin.com/in/lucaszhu
Expectations vs results

Expected faster. Measured slower.

Experts, and the developers themselves, predicted AI would make them faster. On real tasks it slowed them down, and they did not notice.

METR chart: economics experts, ML experts and developers all forecast AI would cut task time; the observed result was that tasks took longer

The executive's nightmare

The promised benefit turns out to be a drag on performance, and the team still says it feels faster. Adoption is not the same as benefit. Measure it.

Two curves

Two curves at once

How far have you gone as an individual user? How far has your organisation gone?

TIME AND EFFORT AI MATURITY 1 2 3 4 1 2 3 4 Personal AI maturity Enterprise AI adoption Illustrative only, curve positions are not data.
Personal AI maturity
1L1 Generative AI: photos, email summaries
2L2 Building tools, services using ‘vibe-coding’ platforms
3L3 Chief of staff: knowledge graph, daily briefings, loops
4L4 Product deployments, multi-model agent harnesses
Enterprise AI adoption
1L1 An LLM for everyone
2L2 Specific deployments for functional departments
3L3 Cultural shift to AI-native
4L4 Metamorphosis of the organisation to be an AI stack
The gap: "AI-pilled" vs "human taste is paramount"
The 3x4 framework

Find your level. Map your path.

Personal
L4 Product deployments, multi-model agent harnesses
L3 Chief of staff: knowledge graph, daily briefings, loops
L2 Building tools, services using ‘vibe-coding’ platforms
L1 Generative AI: photos, email summaries
Knowledge teams
L4 Inter-team agents, building towards a company knowledge base with dynamic reporting
L3 Custom agents, a team chief of staff living in Slack
L2 Data and discussions into a team second brain
L1 Meeting notes summarised
Enterprise
L4 Metamorphosis of the organisation to be an AI stack
L3 Cultural shift to AI-native
L2 Specific deployments for functional departments
L1 An LLM for everyone
Now, in the room
1Mark your level on each ladder.
2Where is the widest gap, and who owns closing it?
3Leave today as the lighthouse for AI maturity at work.
Ikigai AI Ventures

Q&A and war stories

Lucas Zhu · Ikigai AI Ventures

Portfolio
AI Landscapes · AGSM guest session · 24 September 2026 06 / 06