#1 WTF is Metaverse? ft. Tanmay Bhat, Umang Bedi & Aprameya Radhakrishna

The very first episode of WTF: Nikhil Kamath, Tanmay Bhat, Umang Bedi (Dailyhunt and Josh co-founder, former Meta India head) and Aprameya Radhakrishna (Koo co-founder, TaxiForSure founder) try to answer a question the internet could not: what actually is the metaverse - with call-ins from web3 fashion founder Marcos, gamer Prashant, and Himanshu, creator of India's first virtual influencer Kyra. FULL COVERAGE: notes span the entire episode. Sponsor: none stated. All stock opinions, company claims and numbers are the speakers' own, from March 2023, and several are dated (guests even flag John Carmack's Meta exit as 'last week'); nothing here is investment advice - the hosts disclose on air that they may hold shares in companies discussed. Definitions get a workout: the word comes from the novel Snow Crash, via Ready Player One and The Matrix, but Tanmay argues we already log into universes - Instagram is the lifestyle universe, Twitter or Koo the universe of thoughts and opinions - just in two dimensions. Umang pushes democratization over headsets: an Oculus is bulky, costs $400-500 and serves elite experiences; Moore's law, cheap cloud compute and 5G (500 Mbps-plus) should shrink the headset to a display while processing moves to the cloud. Marcos distinguishes blockchain metaverses (Decentraland, Sandbox - NFT land, creator royalties of 85%, weaker graphics because every asset is traced on-chain) from pure simulations running Unreal Engine 5. Prashant frames it as gaming's natural endpoint - interconnected 3D spaces, haptics digitizing touch after sight and hearing - and bets sports (IPL, FIFA) will drive critical mass, with game-engine owners holding the real advantage; VRChat today caps rooms at 30-40 concurrent users. The spiciest segment is on who wins: Microsoft over Meta on enterprise 'by miles', Nvidia as the infrastructure play, no consumer leader until one or two years after Apple ships a device, and the claim that Meta has only ever won by acquiring - 'good times create weak men', with little real innovation between 2010 and 2022. Tanmay's evidence the behavior already exists: GTA roleplay servers where a thousand players hold jobs, run businesses and marry in character, streamed ten hours a day - 'the closest thing to the metaverse that already exists' - plus an $800,000 Counter-Strike skin as proof people pay to signal status in worlds they inhabit. Himanshu explains Kyra: a 200k-follower virtual influencer with a crafted origin story, brand deals with MG Motors and boAt, and the pitch that characters from Harry Potter to Sherlock were never real either. The episode closes on AI (prompt engineering as a coming profession, ChatGPT at $20 called too cheap) and climate (compute emissions versus the plane ride), landing on Umang's conclusion: experiences are goosebumps, technology is just the enabler, and nothing replaces physical human interaction.

Executive Summary (TL;DR)

WTF's first episode defines the metaverse by behavior, not hardware: we already live in app universes, and GTA roleplay servers, VRChat and virtual influencer Kyra show the demand exists. The panel's bets - Microsoft over Meta, Nvidia as the infrastructure play, game engines as the real moats, sports as the killer app, Apple as the trigger - while headsets stay heavy and expensive until 5G and cheap compute catch up. Umang's closing line frames the whole debate: technology enables goosebumps, but nothing replaces being in the room.

Key Takeaways

You already live in proto-metaverses

Tanmay's framing: Instagram is the lifestyle universe, Twitter and Koo are the universe of thoughts and opinions - we log out of reality into app worlds for hours a day, just in two dimensions. The 'metaversal habit' predates the hardware; the word itself comes from the novel Snow Crash, popularized by Ready Player One and The Matrix.

The hardware problem is an economics problem

Umang's case against Oculus: bulky, $400-500, built for elite experiences. The fix is Moore's law plus 5G at 500 Mbps-plus - separate the processor, GPU and battery from the display, stream from the cloud, and a headset becomes a pair of glasses. A billion people already have 5G; back-end compute cost is the bottleneck.

Bet on engines and infrastructure, not destinations

Prashant: game-engine owners (John Carmack's lineage from Doom to Unreal Engine) hold the real advantage because 30-40 users collapse today's VR rooms. The panel's picks, with on-air stock disclosures: Microsoft over Meta on enterprise 'by miles', Nvidia as infrastructure, and no consumer leader until a year or two after Apple ships. Meta's record is acquiring wins, not building them.

The closest metaverse already runs on GTA

GTA roleplay servers host a thousand players who choose occupations, run businesses with bank accounts, speak in character accents, get married in-game and stream it ten hours a day - booted out if they break character. Add an $800,000 Counter-Strike skin and Marcos's 85% creator royalties in Sandbox: signaling and commerce in virtual worlds are proven behaviors.

Sports will be the killer app

Not one binary moment but a slow grind of million-user apps - and the one that cracks mass engagement is sports, because it is tribal and watched together. Whoever simulates the stadium for IPL or FIFA at home first wins the high-frequency use case, and high frequency owns all the low-frequency ones.

Virtual influencers are character IP, not people

Kyra - 200k followers, a crafted 'lost memories' origin story, campaigns with MG Motors and boAt, meme-page marketing that shifted her audience from 90:10 to 70:30 male-female - monetizes like a media character. Himanshu's analogy: Harry Potter and Sherlock Holmes were never real either. Lil Miquela (3M followers, music videos at 5-10M views) is the template.

The honest ending: tech enables, presence wins

After weighing Zoom against the Saudi Arabia flight and AI's promise (prompt engineering as a profession, hospitals' data leverage), the panel lands on a human conclusion. Tanmay: only the most important things will stay real. Umang: experiences are goosebumps, technology is the enabler, and nothing replaces being a social being in a physical room.

The Conversation Breakdown

Untimed topic

⏱ Untimed topic

Banter and business: sleep protocols, Tanmay's TaxiForSure stress weight, Koo's Brazil hack ('Koo means ass in Portuguese'), and Dailyhunt's family of apps

Untimed topic

⏱ Untimed topic

Is old media dead: S&P lifespans, paywalls, IP-based TV, entertainment fragmenting into 30-second stars, and Tanmay's OTT distribution thesis

Untimed topic

⏱ Untimed topic

Defining the metaverse: Snow Crash to Ready Player One, app universes we already inhabit, and whether a phone counts

Untimed topic

⏱ Untimed topic

The hardware debate: Oculus neck pain and counterweights, 5G plus cloud compute as the fix, Marcos on blockchain versus simulation metaverses

Untimed topic

⏱ Untimed topic

Who wins: Microsoft versus Meta, Carmack's exit, Nvidia as infrastructure, Apple's trigger, and game engines as the real moats

Core Frameworks & Ideas

The behavior-first test

Judge the metaverse by existing behavior, not shipped hardware: 10-hour roleplay streams, $800,000 skins, and 200k people following a fictional influencer all signal demand. If people already pay to signal status and live stories in virtual spaces, immersion depth is a when, not an if.

High frequency owns low frequency

Umang's platform rule for the metaverse: whoever cracks the highest-frequency use case - daily communication, weekly sport - earns the right to own every lower-frequency one, from concerts to commerce. Frequency, not fidelity, determines the winner.

Democratization beats experience quality

Every viral consumer technology won by running on any device at any price, not by being best: TikTok, Facebook, Koo. A $500 headset for elite experiences inverts that logic - so the metaverse tips only when Moore's law and 5G make immersion nearly free.

Actionable Next Steps

✓Before buying any 'metaverse' thesis, list which existing user behavior it extends - if you cannot name the GTA-roleplay equivalent, be skeptical.
✓Track the three real bottlenecks the panel names - headset form factor, back-end compute cost, and 5G bandwidth - as your adoption dashboard instead of token prices.
✓If you build consumer products, study the high-frequency rule: anchor on a daily or weekly habit before bolting on occasional experiences.
✓Watching AI's rise next? Treat prompting as a learnable skill now - the panel called prompt engineering as a profession before it was mainstream.

Notable Quotes

“this is the closest thing to the metaverse that already exists that I've seen”
— Tanmay Bhat, on GTA roleplay servers

“only the most important things in life will be real”
— Tanmay Bhat

“experiences at the end of the day are goosebumps”
— Umang Bedi

The One Thing to Remember

Stop asking whether the metaverse is real and look at GTA roleplay servers - people already hold jobs, run businesses and get married inside virtual worlds for ten streamed hours a day; the behavior arrived before the hardware.