THEY KEEP ANNOUNCING MY OLD FRONTIERS
THEY KEEP ANNOUNCING MY OLD FRONTIERS
I Built the Phiverse Before Meta Declared the Metaverse Its Future. I Kept Building When the Servers Died. Five Years and Tens of Billions Later, They’re Unveiling an AI Creature on a Keychain. So Put the Timeline on the Table.
There is a very easy way to settle this.
Stop arguing about who sounds visionary.
Stop counting employees.
Stop counting GPUs.
Stop counting market capitalization.
Put the dates on the table.
Because once you do that, this gets uncomfortable very quickly.
2019: I START PHI NETWORK
I started Phi Network in 2019.
Not a metaverse white paper.
Not a keynote.
Not a speculative render of people wearing headsets someday.
A functioning environment.
Identity.
Value.
Applications.
Creation.
Participation.
An economy.
A world.
I called the larger living environment the Phiverse.
My published account records Phi Network eventually reaching roughly 100,000 users and approximately 200 applications, with a custom EVM-compatible network operating by 2020 and infrastructure I continued operating through 2023.
Remember that date.
Because two years later, one of the largest corporations on Earth was going to announce the future.
OCTOBER 2021: FACEBOOK DISCOVERS “THE METAVERSE”
On October 28, 2021, Facebook renamed itself Meta.
This wasn’t subtle.
This wasn’t reporters imposing a narrative on Zuckerberg.
Read his own words.
He called the metaverse:
“the next frontier”
He said it would “touch every product we build.”
He said he wanted Meta’s “identity” anchored on it.
And then came the extraordinary declaration:
“metaverse-first, not Facebook-first.”
Meta predicted that within a decade its metaverse would reach one billion people, host hundreds of billions of dollars of digital commerce, and support millions of creators and developers.
This was not presented as a side project.
THE COMPANY CHANGED ITS FUCKING NAME.
Meanwhile:
Phiverse: already existed.
I had already spent years building the thing I was describing.
That distinction becomes important.
Because now we get to watch what happens when reality arrives.
FOLLOW THE MONEY
Meta’s own SEC filings let us measure Reality Labs—not purely “metaverse spending,” because the segment also includes VR, AR and later wearables, an important accounting distinction.
Here are its reported operating losses:
Year
Reality Labs operating loss
2019
$4.503B
2020
$6.623B
2021
$10.193B
2022
$13.717B
2023
$16.120B
2024
$17.729B
2025
$19.193B
The 2019–2021 figures come directly from Meta’s 2021 10-K; the later figures are reported in subsequent SEC filings.
Do the arithmetic.
$88.078 BILLION IN CUMULATIVE REALITY LABS OPERATING LOSSES FROM 2019 THROUGH 2025.
And just from the year Meta renamed itself around the metaverse thesis—2021 through 2025:
$76.952 BILLION.
That does not mean Meta “spent $77 billion exclusively on the metaverse”; Reality Labs encompasses additional technologies.
It means something more precise:
the division carrying Meta’s future-platform ambitions produced nearly $77 billion of operating losses during the five years beginning with the company’s metaverse declaration.
Now remember the promise:
“Metaverse-first.”
“The next frontier.”
“It will touch every product we build.”
Keep those words.
We’re going to need them.
2023: MY SERVERS DIE
Here’s where the two stories separate completely.
I ran out of money to keep Phi’s infrastructure alive.
The servers went down.
Applications became unreachable.
Chains went down.
Accounts and balances became inaccessible through the infrastructure that had remembered them.
And instead of announcing:
PHIVERSE 2: NOW WITH AI
I had to confront the actual engineering failure.
My supposed decentralized world could disappear because the machines remembering the world disappeared.
That’s the event from which the subsequent architecture follows in my published record.
That failure gave me a harder question:
WHAT MUST SURVIVE WHEN THE INFRASTRUCTURE DOESN’T?
That question eventually becomes Receiz.
Proof in the object.
Identity in the object.
History in the object.
Authority carried with the object.
Portable state.
Offline verification.
Continuity that doesn’t require the original machine to remain the authority.
The server can help.
The server cannot be the reason reality exists.
That is what happened to the Phiverse.
IT DIDN’T BECOME A DISCARDED BUZZWORD.
Its failure became the engineering requirement for what came next.
THEN THE OBJECT STARTED LIVING
And this is where yesterday’s Meta announcement becomes fucking hilarious.
Because I didn’t stop at proving static files.
The same architecture moved into living subjects and worlds.
By July 15, 2026, the published Wilds V3 release described a persistent world with geography, ecology, bosses, raids, teams, leagues, crafting, lineage, narrative memory and return-player continuity—all built on the Receiz foundation. The release records 552 automated tests passing as part of its qualification.
Wildz is publicly described today as:
“A proof-native creature world”
where creatures fight, evolve, reproduce and become portable beings, with offline-verifiable state and continuity across time.
Then Receiz v120 went further.
On August 17, 2026, I publicly described the creature itself becoming the continuing proof object:
persistent identity;
append-only life;
event-derived memory;
relationships;
autonomous mandates;
ownership transfer without destroying identity;
a subject capable of continuing permitted activity while its owner is away;
and continuity surviving changes in application, device, model provider and owner.
Read that again.
August 17.
Now fast-forward thirty-seven days.
SEPTEMBER 23, 2026: LOOK EVERYBODY, META HAS AN AI CREATURE
Yesterday Zuckerberg walked onstage and unveiled Muse Charm.
A little physical object.
On your keychain.
Containing an animated AI character.
You talk to it.
It talks back in real time.
It accompanies Meta’s new personal AI agent.
The coverage immediately described it as Tamagotchi-like.
And apparently we’re supposed to look at this and say:
WOW. THE FRONTIER.
🤣
No.
Put the fucking timeline back up.
THE TIMELINE IS THE STORY
2019 — Me: Phi Network begins.
2020 — Me: custom EVM-compatible Phi infrastructure operating.
2021 — Meta: changes the corporation’s name and declares the metaverse “the next frontier.”
2021 — Meta: promises to become “metaverse-first.”
2021–2025 — Meta Reality Labs: nearly $77 billion in cumulative operating losses.
2023 — Me: Phi infrastructure becomes financially unsustainable and goes dark.
My response: solve the architectural failure that killed it.
Not another slogan.
Not another headset.
Continuity.
2025–2026 — Me: memory, identity, authority, provenance and state increasingly become portable, locally verifiable objects rather than facts that exist only because my infrastructure remembers them. My September chronology documents public milestones including Memory Crystals in June 2025 and offline operation in October 2025.
July 2026 — Me: publicly released living-world architecture with persistent creatures, ecology, lineage and continuity.
August 17, 2026 — Me: publicly documented proof-native living subjects carrying identity, history, memory, relationships, ownership and continuity across changing infrastructure.
September 23, 2026 — Meta: unveils an animated personal AI companion that fits on a fucking keychain.
Come on.
SO WHERE IS THE METAVERSE?
That’s the question.
Not because Meta literally stopped all VR or metaverse development. It hasn’t.
The company’s filings still describe investment in metaverse, VR, AR, mixed reality and wearables.
But compare the language.
In 2021:
THE NEXT FRONTIER.
OUR IDENTITY.
METAVERSE-FIRST.
ONE BILLION PEOPLE.
HUNDREDS OF BILLIONS IN COMMERCE.
And in September 2026, reporting describes Meta’s new Muse personal AI agent as the centerpiece of Zuckerberg’s AI vision, while his surprise hardware reveal is a pocket-sized AI companion.
I’m not asking you to believe my interpretation.
LOOK AT THEIR FUCKING DATES.
Their words are archived.
Their financial statements are filed with the SEC.
My systems are public.
My repositories are public.
My releases are public.
Run them.
AND THAT’S WHY “CONVERGENCE” IS SUCH A WEAK DESCRIPTION
Convergence makes this sound like two people independently drew similar pictures on a napkin.
That isn’t the interesting observation.
The interesting observation is continuity of work.
When Phi failed, I didn’t throw away the problem.
I inherited the failure.
The failure became a requirement.
The requirement became an architecture.
The architecture became proof objects.
The proof objects acquired portable state.
Portable state acquired memory.
Memory acquired provenance.
Identity acquired succession.
The objects acquired lives.
The lives acquired worlds.
The creatures acquired continuity.
And now other intelligences can reason over those continuing objects.
Receiz’s current developer documentation explicitly describes the same proof object being consumed by different players, profiles, agents and workflows while preserving identity and prior history—and applications being able to recognize identity, objects and verified history without forcing the person to start again.
That is what happened to the Phiverse.
It grew up.
NOW TELL ME WHAT HAPPENED TO “METAVERSE-FIRST.”
That’s the comparison.
One side had vastly more money.
More engineers.
More compute.
More press.
More institutional credibility.
More distribution.
More everything.
And yet the intellectually interesting question isn’t how much infrastructure somebody can purchase.
It’s:
Did the underlying problem survive the fashion cycle?
Mine did.
When my implementation failed, I kept following the problem downward.
When the server failed, I investigated what survives the server.
When the database wasn’t enough, I moved continuity toward the object.
When an object needed history, I made history verifiable.
When intelligence needed memory, I made memory answer to provenance.
When a creature needed to remain itself, I made continuity an object property.
When the machine changed, the subject didn’t have to.
That’s a coherent engineering lineage.
And I don’t even have to tell you that Meta copied me.
That claim is unnecessary.
I can say something much harder for anyone to dismiss:
Here are my dates.
Here are their dates.
Here are my releases.
Here are their promises.
Here are their SEC filings.
Here is what they called “the next frontier.”
Here is what I’m running now.
Run both.
Because after nearly $77 billion of Reality Labs operating losses since the “metaverse-first” declaration, I shouldn’t have to sit here in September 2026 pretending an animated AI creature you can carry around is some incomprehensible new frontier.
I already built the world.
Then I watched my world die.
And instead of changing the fucking buzzword,
I SOLVED WHY IT DIED.
That’s the difference.