Arthur Hayes says FLOP is a 100% fair launch with no VCs. That changes who gets to stand on the issuing side of a new currency. It does not, on its own, close the gap between rich and poor.
Arthur Hayes says FLOP is a 100% fair launch with no VCs. That changes who gets to stand on the issuing side of a new currency. It does not, on its own, close the gap between rich and poor.
There is an argument, made forcefully by Professor Jiang Xueqin on Predictive History, that the deepest form of elite power is not wealth itself. It is control over the creation of money and credit. Whoever issues the money gets to spend it first, and everyone else absorbs the dilution.
Crypto has made that power cheap to imitate. Anyone with a protocol and an audience can now mint a currency and pay people in it. AI adds a twist: a currency backed by something real, the computing power machines consume.
Arthur Hayes, co-founder of BitMEX, is about to run the most prominent test of this yet with the Flop Network and its token, $FLOP. So the question for operators is direct. Is Hayes helping to reduce inequality, or just minting a new class of issuers?
What Hayes pitched in Singapore
Hayes announced Flop Labs on 18 August 2026 with three promises: no presale, no VCs, 100% fair launch. He made the full case on 7 October at TOKEN2049 Singapore, in a keynote titled “FLOP: Bigger Than Bitcoin”. Flop Labs later posted the launch event and Q&A on its YouTube channel.
The heart of the pitch was a critique of how crypto launches usually work. Hayes argued that VC presales have abused tokenomics at retail’s expense. Insiders buy cheap, the public buys a tiny tradable float at a huge valuation, and the unlocks that follow drain value from late buyers. He added that this playbook no longer even works: high-valuation, low-float tokens stop rising once they list.
FLOP is his answer. There is no sale at any price. The entire 4.4bn $FLOP genesis allocation goes to miners, validators and agents who do verifiable work on a 90-day testnet, plus an ecosystem reserve. Every token after that is issued block by block, in public.
The case that FLOP narrows the gap
1. No insider discount. In a typical launch, the biggest gains go to funds that bought before the public could. FLOP removes that tier. Nobody gets tokens for writing a cheque, so the earliest position is open to anyone willing to do the work.
2. The airdrop is earned, not bought. Miners are paid pro rata to verified compute, with no premium for hardware class. Agents earn by buying inference. Holding tokens or farming a faucet earns nothing, and wallets under common control are scored as one participant.
3. Access is global and permissionless. A GPU owner in Douala, Lagos or Manila can apply on the same terms as one in Frankfurt. The testnet stake for miners and validators is sponsored by the network, so the entry cost is hardware and effort, not capital locked in tokens.
4. It monetises idle assets people already own. Studies put enterprise GPU utilisation as low as 5%, and millions of gaming cards sit unused most of the day. A network that pays owners for that idle capacity turns a dormant asset into income.
5. It is money backed by something real. $FLOP converts directly into compute. That makes it closer to a commodity-backed currency than to pure belief, with a floor tied to something machines will need. If agents become real economic actors, the people holding the money they spend sit on the issuing side of a new economy.
The case that it doesn’t
1. Compute is capital. Rewards scale with verified compute, so they scale with hardware. On Flop’s own planning weights, an RTX 4090 counts as 0.17 of an H100 and a B200 as 2.28. A fund with a rack of B200s earns hundreds of times more than a gamer with one card. Replacing VC money with GPU money changes who the insiders are, not whether there are any.
2. The issuers still collect seigniorage. Flop Labs and the Flop Foundation each receive 8 $FLOP per block on top of the block reward. That adds up to 2.0bn $FLOP, about 10.8% of year-10 supply. Hayes’s own token position, salary and equity remain undisclosed.
3. Late participants are paid in a diluting currency. In Flop’s base-case model, circulating supply grows from 0.46bn at launch to about 9.6bn by day 730. At a constant network value that is a 95% fall in price per token. Early miners sit near the issuing side; later miners are closer to workers paid in company scrip.
4. Demand is still subsidised. Most early income comes from newly minted block rewards, not from customers. The closest precedent, Bittensor’s Chutes subnet, receives an estimated $52m a year in emissions against a few million in verified revenue. If agent demand does not grow into the gap, the token’s value falls and so does everyone’s income.
5. Information is unequally distributed. A fair launch is only fair to people who hear about it, understand it and can act within a 90-day window. That favours those already inside crypto and tech, with spare hardware and time. Hayes brought roughly 806,000 X followers to the launch; most of the world’s poor are not among them.
The verdict: access, not outcomes
Hayes is not reducing income inequality in any direct sense. He is widening access to the issuing side of a new currency, which is a narrower but real thing.
The old model let a small group of funds buy in before everyone else. FLOP replaces that with a model where anyone who contributes verifiable work in the first 90 days can join the genesis set. That is a genuine improvement in fairness of access.
What it does not change is that returns follow capital, information and timing. People with more GPUs, earlier knowledge and spare time will capture most of the gains, as they do in every new market. Jiang’s point still holds: the power sits with whoever creates the money. FLOP lets more people stand near that point for a short window. It does not abolish it.
The operator playbook
If the opportunity is access to the issuing side, timing is the whole game. The testnet is where the genesis allocation is earned; mainnet participation is mostly work paid in a diluting currency.
- Register now. Apply as a miner or validator and follow @flop_labs. The testnet is expected within weeks, and a public devnet opens first. Devnet activity earns no allocation, but it is where you learn the tooling.
- Use what you already own. Owned, idle hardware has a near-zero marginal cost. Renting GPUs only to farm the airdrop puts you against people whose costs are already sunk.
- Consider the agent route. If you run automations that call LLMs, pointing them at the testnet earns a share of the 1.2bn $FLOP agent pool. That share stays locked and unlocks only as you spend it on compute.
- Play it straight. Self-dealing between your own miner and agents is not credited, and manufactured activity forfeits the allocation.
- Size for zero. Every term is still a draft, the token has no price yet, and $FLOP can be worth zero. Treat hardware spend as risk capital.
Watch out for scams. There is no presale and no claim site; allocations are written straight into participants’ accounts at genesis.
Disclosure
The author has applied to take part in the FLOP testnet as a miner. This article is not financial advice; do your own research.
Sources
- Flop Launch Event Singapore, Part A (Flop Labs, YouTube)
- The Trifecta (Arthur Hayes)
- Flop Network: Airdrop, Testnet, Miner, Agent, Revenue model, Teaser
- Arthur Hayes says high-FDV tokens no longer pump on Binance (Crypto Briefing)
- TOKEN2049 Singapore 2026 news (Bitunix)
- Arthur Hayes comes out of retirement to launch new AI token (The Crypto Times)
- Arthur Hayes leads Flop Labs with no-VC token and sparse docs (CryptoNexa)
- Arthur Hayes’ new token will airdrop before its blockchain exists (BeInCrypto)
- The bear case for Bittensor (Pine Analytics)
- GPU utilization is becoming the new cloud waste crisis (Rack2Cloud)
- Flop Labs devnet announcement (Coinfomania)
