How GOAT Gaming hit 6M players using Telegram, memes, and an AI agent named Amy

Sam kamani

October 18, 2025

What happens when you merge meme culture, AI, and Telegram-native game design?

Simon Davis—founder of GOAT Gaming and Mighty Bear Games—joined the Web3 with Sam Kamani podcast to unpack how their team built one of the most compelling gaming experiences in the Web3 space. With over 6 million players onboarded via Telegram alone, GOAT’s approach is turning heads.

But this isn’t just about numbers. It’s about rethinking game design, AI-powered engagement, and crypto-native infrastructure for a new kind of digital playground.

From QA tester to Web3 game visionary

Simon’s path to GOAT Gaming wasn’t traditional. From a background in music and quality assurance, he eventually co-founded Mighty Bear Games before going all-in on Web3 in 2021. That context matters—his focus isn’t just technical. It’s cultural.

He understands that today’s players don’t just want good games. They want social context, identity, and expression baked into the experience.

Why Telegram?

GOAT didn’t just build a game on Telegram. They built for Telegram.

With over a billion users and frictionless onboarding, Telegram is a sleeper powerhouse for distribution. By embedding their games directly inside the Telegram experience (via bots and mini-apps), GOAT hit 6 million players—without App Store hurdles or traditional UA spend.

Instead of luring users into a new platform, they brought the game to where users already live and chat.

Enter Amy: your AI gaming host

At the heart of GOAT’s experience is Amy, an AI-powered game agent that interacts directly with users.

But she’s not just a glorified chatbot. Amy has a backstory, evolving personality, and mission-driven role. She cheers users on, organizes challenges, and even sparks conversations. She’s an AI character embedded into the game’s narrative—making players feel like they’re not just playing, they’re participating.

Simon’s team iterated Amy from a basic FAQ bot into a dynamic character that keeps players coming back.

“People don’t just come back for gameplay—they return for who they’re playing with,” Simon noted.

And that includes Amy.

Airdrop farmers vs real communities

A standout moment in the episode is Simon’s blunt take on the Web3 engagement trap:

“If they wouldn’t play without rewards, they’re not a community—they’re extractors.”

That shift in mindset—from bribing users to building real bonds—is what’s driving GOAT’s long-term vision. Their meme-fueled game Grand Theft Pepe exemplifies that. It’s absurd, it’s fun, and it builds loyalty through culture—not just incentives.

Designing responsibly: guardrails for AI agents

As AI becomes a core part of gaming experiences, Simon emphasizes the importance of ethical design.

Amy was built with clear guardrails, including content moderation, consent boundaries, and adaptive tone control. The team is intentional about making sure Amy delights, not disturbs.

They’re not just using AI to automate—they’re using it to augment play in thoughtful, human-centered ways.

Why AI + Crypto makes sense

Two standout takeaways:

AI agents can’t open bank accounts. But they can use crypto.
DeFi is the default financial layer for an agent-driven world.

Web3’s programmable money layer makes it the perfect match for autonomous AI agents. From in-game purchases to NFT ownership to coordination across time zones, crypto gives agents like Amy a way to act, transact, and reward—all without human middlemen.

Simon envisions a future where billion-dollar studios could be run by three people and an army of AI agents. It’s not just possible. It’s already happening.

What it takes to build AI-native games

The hardest part? Hiring talent who live and breathe generative AI tools.

Simon’s team is searching for people who don’t just use AI, but think in its language. People who can architect experiences that feel conversational, intuitive, and delightfully unpredictable.

“If you're still building like it’s 2021, you’re already behind,” he says.