Why Nobody Has Built a Modern Second Life (And Probably Won't)
Meta spent over $83 billion trying to build the next Second Life. The result? About 900 daily users, a 30% budget cut, and investors who cheered when the company finally gave up.
Meanwhile, Second Life — launched in 2003, running on an engine older than YouTube — quietly grew to 620,000 monthly active users, paid creators $78 million in a single year, and produced 14 people who became millionaires from selling virtual fashion and real estate.
This is a story about the most expensive failure in gaming history, a platform that refuses to die, and why the answer to "why hasn't someone built a modern Second Life?" might be uncomfortably simple.
Second Life Is Somehow Still Thriving
Let's start with the numbers that surprised me.
Second Life's virtual economy moves roughly $650 million per year. Over its lifetime, Linden Lab has paid out $1.1 billion to creators. Not venture-backed startups. Not corporate partners. Individual people making clothes, buildings, animations, and experiences inside a virtual world.
In 2023 alone:
- 21,152 creators earned income
- 6,446 earned over $1,000/year
- 1,580 earned over $10,000/year
- 139 earned over $100,000/year
- 14 became millionaires — in a single year
The platform takes only a 10% cut of transactions. On Roblox, developers typically take home 25-30% after platform fees and app store cuts.
The average Second Life user has been on the platform for 14 years. Let that sink in. These aren't casual visitors. They're residents. They've built lives, businesses, and identities inside this world.
Second Life has been profitable since 2005 — two years after launch — and has never required its shareholders to take a loss. The company has never taken a dividend either. Every dollar goes back into the platform.
So why can't anyone build this again?
The Graveyard of Contenders
Meta Horizon Worlds: $83 Billion for 900 Users
In October 2021, Facebook renamed itself Meta and declared the metaverse was the future of human interaction. Mark Zuckerberg committed the company's resources to building a virtual world that would replace the internet.
The spending was staggering. Reality Labs, Meta's metaverse division, posted operating losses that grew every single year:
| Year | Operating Loss |
|---|---|
| 2020 | $6.6B |
| 2021 | $10.2B |
| 2022 | $13.7B |
| 2023 | $16.1B |
| 2024 | $17.7B |
| 2025 | $19.2B |
Cumulative: over $83 billion since 2020.
What did they get for it? Horizon Worlds launched to 300,000 monthly users in February 2022. By October, that had dropped below 200,000. A YouTuber investigation found roughly 900 daily users actually inside the platform at one point. An internal Meta memo admitted that even Meta employees had stopped using it.
In December 2025, Meta cut the Reality Labs budget by 30% — a $4-6 billion reduction. Meta's stock jumped 4%, adding $69 billion in market value. Wall Street literally celebrated the retreat.
By February 2026, Horizon Worlds was separated from Quest VR and pivoted to mobile-only. The $70+ billion in annual CapEx was redirected to AI data centers instead. TechCrunch published the epitaph: "Well, there goes the metaverse!"
Sansar: Linden Lab Couldn't Even Compete With Itself
Here's the most damning failure on the list. Linden Lab — the company that built Second Life — poured years of development into Sansar, a VR-native successor to their own platform. If anyone understood what made virtual worlds work, it should have been them.
Sansar launched in public beta in July 2017. It required expensive VR hardware. It offered no way to transfer your Second Life inventory, currency, or identity. It wasn't a persistent shared world — it was a collection of separate VR experiences.
The Second Life community didn't migrate. They resented the resources being diverted from their platform. In February 2020, Linden Lab laid off over 40 Sansar staff. A month later, they sold the entire project to a startup called Wookey Projects.
By December 2021, Sansar went offline without explanation. It eventually came back, but it's functionally abandoned. The company that built Second Life couldn't build Second Life again.
High Fidelity: The Creator Himself Couldn't Do It Twice
Philip Rosedale created Second Life. In 2013, he founded High Fidelity to build the next generation of virtual worlds. He raised over $100 million from investors including Google Ventures, True Ventures, and Breyer Capital.
By 2019, Rosedale admitted publicly: "This model is not working right now." He noted that daily VR headset usage was "only in the tens of thousands" — not enough to sustain a virtual world. He laid off 25% of his staff.
High Fidelity pivoted to enterprise virtual offices. Seventy-five organizations tested it. None adopted it. Another 50% layoff followed. The company pivoted again — this time to spatial audio technology, which is what it does today.
In January 2022, Rosedale invested in Linden Lab. The creator of Second Life, after spending over $100 million trying to build something better, went back to the original.
Decentraland: 38 Users in a $1.3 Billion Ecosystem
Decentraland might be the most absurd entry in this graveyard.
In October 2022, DappRadar reported that this blockchain-based virtual world — valued at $1.3 billion — had 38 daily active users. Thirty-eight. Decentraland disputed the methodology, and DappRadar later revised upward to 650 after expanding contract tracking. Even at 650, the ratio of valuation to usage was staggering.
In Q1 2023, unique active wallets crashed 42% in 30 days. Virtual land prices saw the steepest decline among all metaverse platforms.
This is what happens when you build a virtual world on speculation instead of community.
"But What About Roblox?"
Roblox is the obvious counterargument. With over 150 million daily active users and nearly $5 billion in annual revenue, it's the most successful user-generated content platform in gaming history.
But Roblox is not Second Life. It's a fundamentally different product serving a fundamentally different audience.
Roblox is a game platform where users (predominantly children and teenagers) create and play games within Roblox's engine. Second Life is a virtual civilization where adults create and inhabit a persistent shared world with a real economy.
The economics tell the story. Second Life creators are 100 times more likely to earn $10,000 or more per year than Roblox creators. Roblox paid out $1.5 billion to creators in 2025, but the top 1,000 developers took home roughly 87% of that pool. Second Life earns roughly $190 per active user per year; Roblox earns about $12.
Roblox is a children's entertainment platform with a creator ecosystem. Second Life is a virtual economy with a social world attached. They solve different problems for different people. Roblox's success doesn't answer the question of why nobody has built a modern Second Life — it reinforces it.
VRChat: The Closest Thing (That Still Isn't It)
Of everything on this list, VRChat comes closest to capturing Second Life's spirit. It has roughly 8 million registered users, peaks at nearly 150,000 concurrent on New Year's Eve, and averages about 43,000 concurrent players. It's a genuine social platform where people form communities, attend events, and express themselves through custom avatars.
But VRChat has no economy. No virtual land ownership. No persistent world-building. No creator marketplace generating real income. It's a VR social hangout — closer to a 3D chat room than a virtual civilization.
VRChat proves that the social desire for virtual worlds exists. It also proves that the social layer alone isn't enough to build what Second Life built. The economy is the thing.
Could AI Change the Equation?
Here's where the conversation gets interesting — and where I need to be honest about what's real versus what's potential.
AI isn't ready to build a Second Life competitor today. The technology for procedurally generating persistent, coherent virtual worlds at the scale and fidelity users would expect doesn't exist yet. AI-generated 3D assets are improving rapidly but still can't match what human creators produce in Second Life's marketplace. And no AI system currently understands how to bootstrap the kind of emergent social economy that makes Second Life work.
But if anything is going to break through where $83 billion in brute-force corporate spending failed, AI is the most plausible path.
Consider what made Second Life's content creation so revolutionary in 2003: anyone could build inside the world without being a professional developer. The barrier was low enough that a fashion designer could create virtual clothing and sell it. A musician could build a venue and host live shows. An architect could design buildings and sell them as prefabs.
That barrier, while low for 2003, is still substantial by modern standards. Second Life's building tools require significant learning. Its scripting language (LSL) requires programming knowledge. The reason only 21,000 out of 620,000 monthly users create content for income is that creation is still hard.
AI agents could change this equation — eventually. Imagine a virtual world where you describe what you want to build and an AI constructs it. Where NPCs aren't scripted automatons but genuinely responsive characters that make the world feel alive even when other players aren't online. Where the tools for building game worlds become accessible enough that the creator-to-consumer ratio shifts dramatically.
This isn't science fiction — these capabilities are developing in game engines right now. Procedural generation tools are already producing terrain, vegetation, and building layouts that would have taken level designers weeks to handcraft. AI coding assistants can generate interactive scripts from natural language. The building blocks exist.
But there's a critical gap between "AI can generate a 3D model" and "AI can sustain a virtual economy where people invest years of their lives." The former is a technology problem being solved. The latter is a social and economic problem that no one has solved outside of Second Life itself.
There's also an economic angle worth considering. The same labor cost deflation reshaping indie game development could eventually make persistent virtual worlds viable for small teams — not through billion-dollar budgets, but through AI labor that makes the maintenance and content creation affordable.
AI might eventually lower the creation barrier enough to bootstrap a new virtual world. But the technology needs to mature considerably before it can deliver on that promise. We're watching the foundation being laid, not the building going up.
The Uncomfortable Truth
Every failed competitor made the same mistake: they assumed Second Life's magic was technological.
Meta thought better graphics and VR immersion would do it. Sansar thought VR-native design would do it. High Fidelity thought next-gen infrastructure would do it. Decentraland thought blockchain ownership would do it.
None of it worked because Second Life's moat isn't technology. It's time.
Twenty-three years of accumulated culture. Twenty-three years of economic infrastructure. Twenty-three years of relationships, businesses, and identities built inside a shared world. An average user tenure of 14 years. A creator economy where people have literal careers.
You can't speedrun two decades of emergent civilization. You can't throw $83 billion at community and expect it to appear on a quarterly earnings timeline. You can't announce a "metaverse" and expect millions of people to abandon the virtual lives they've already built somewhere else.
The answer to "why hasn't anyone built a modern Second Life?" might just be: because Second Life is already it. It's 23 years old, it runs on aging infrastructure, it looks dated by modern standards, and it's still the only virtual world that actually works as a virtual world.
Maybe that's not a technology problem waiting to be solved. Maybe it's just what building a civilization looks like — slow, messy, organic, and impossible to replicate on a roadmap.
The next Second Life won't be built by a corporation with a ten-figure budget and a three-year timeline. If it happens at all, it'll grow the way the original did: quietly, from the bottom up, one creator at a time, over decades that nobody with a quarterly earnings call has the patience to wait for.