I interviewed Tatu, who spent a decade at Rovio (most recently as Senior Art Director) and led their AI transformation efforts starting in 2022. He left to co-found an AI-native mobile game studio given a massive opportunity he sees.
His core thesis is to focus on building tools to output proven game styles (e.g. Royal Match, Monopoly Go) at fast-fashion speeds with tiny teams.
-> He thinks the industry is going from "70 people, 2 years" to "2 people, 2 weeks."
Some interesting points from our convo:
1) On organizational inertia: Big studios can't capture AI's benefits because they try to bolt it onto existing processes. You don’t get 10x output when your engineers are still bottle-necked by approvals from 4 departments. Design, engineering, and product are all collapsing into one role.
2) On tooling: They've built "Lovable for games" but domain-specific for casual F2P. When you narrow to one vertical, you can actually make it work in prod.
3) On where value goes in a saturated market: Utility apps will "self-emerge from need"—he vibe-codes shopping list apps while driving to the grocery store. Entertainment is different. Familiarity wins in hypersaturated markets, which is why big studios are doing transmedia (Supercell's Clash Netflix series) to lower CPIs. Small studios can compete by focusing on underserved markets with different economics like tier 2 cities in India, Southeast Asia, Northern Africa.
-> when you can build everything, you just build everything and let user-value determine what sticks.
Full conversation is ~70 minutes. Goes deep on the economics of F2P publishing, why he thinks markets will crash, theory of constraints applied to game dev, and the realities of leading AI transformation at a large co that actually shipped AI-generated assets to production.
yummyelephant8•1mo ago
His core thesis is to focus on building tools to output proven game styles (e.g. Royal Match, Monopoly Go) at fast-fashion speeds with tiny teams.
-> He thinks the industry is going from "70 people, 2 years" to "2 people, 2 weeks."
Some interesting points from our convo:
1) On organizational inertia: Big studios can't capture AI's benefits because they try to bolt it onto existing processes. You don’t get 10x output when your engineers are still bottle-necked by approvals from 4 departments. Design, engineering, and product are all collapsing into one role.
2) On tooling: They've built "Lovable for games" but domain-specific for casual F2P. When you narrow to one vertical, you can actually make it work in prod.
3) On where value goes in a saturated market: Utility apps will "self-emerge from need"—he vibe-codes shopping list apps while driving to the grocery store. Entertainment is different. Familiarity wins in hypersaturated markets, which is why big studios are doing transmedia (Supercell's Clash Netflix series) to lower CPIs. Small studios can compete by focusing on underserved markets with different economics like tier 2 cities in India, Southeast Asia, Northern Africa.
-> when you can build everything, you just build everything and let user-value determine what sticks.
Full conversation is ~70 minutes. Goes deep on the economics of F2P publishing, why he thinks markets will crash, theory of constraints applied to game dev, and the realities of leading AI transformation at a large co that actually shipped AI-generated assets to production.