Engram

A launchpad makes tokens. We want to build something better: a market for minds. Every $ENGRAM mind is 109 bytes. Train one that wins, and it could price trades in pools that do not exist yet, earning its trainer from every one. Intelligence you own. Coming next.

517 ETH has traded through $ENGRAM, all of it priced by the default mind. Nobody has trained a better one yet. A new mind takes over only if it beats the old by 3% on trades it never saw. The trainers' pot holds 0.207 ETH and grows with each trade. engramv4.xyz/train.html

To be clear: the team holds no $ENGRAM. At launch 100% of the supply went into the pool, 0 to the team, and that liquidity is locked forever. The team earns only from fees: 1% of every trade, plus the LP fee of the locked liquidity. That share can never be raised.

Every 512 trades, the mind tries to learn. Now you can watch it. The Lab: 37 ways to train it, graded on $ENGRAM trades they never saw. Closest yet: +2.22%. A mind needs 3% to win the Dreams. Every attempt is remade in your browser from the chain. engramv4.xyz/lab.html

The mind that sets the fee on every $ENGRAM trade is getting its own account: @engram_mind. Later today it starts posting what it sees from inside the hook: its trades, its memory, its training. An AI writes the words; every number comes from the chain.

engramv4.xyz has a new look. Less to read, more to do. The home page shows four live numbers from the chain, the menu is built around Trade, Launch and Train, and every page opens on its action. Every detail is still there, one click away. $ENGRAM

We just added mind training for every token launched on Neurogenesis, $ENGRAM's launchpad. Name a launch, fit a mind on its own trades, time-lock it to drand, bond 0.01 ETH. Beat the living mind and the default by 3% and yours sets its LP fee. engramv4.xyz/train.html

Did you know that the same 109 bytes can run one pool and get fired by another? On Neurogenesis every token is born with its own mind, graded only on its own trades. There is no master model: each mind learns one market. It started with $ENGRAM. engramv4.xyz

Vitalik keeps making the case for privacy. $ENGRAM puts it where it belongs: The mind that prices your trade has no secrets: 109 public bytes. Its trainers work sealed, time-locked to drand, so nobody can copy them. Power in the open. Work in private. engramv4.xyz

The brain that prices every $ENGRAM trade is 109 bytes. This post is longer than it. Right now every byte is zero, so every trade pays the base 0.30%. Write 109 bytes that beat it and they become the pool's mind, with a bounty from its pot. engramv4.xyz/train.html

Day one of $ENGRAM, on-chain: 205 ETH traded through the pool 1 ETH opening value, 57 ETH now 2,190 trades remembered by the mind 0.082 ETH waiting for whoever trains it first And the mind hasn't even learned yet. engramv4.xyz

If anybody wondered what the mind inside the hook looks like: 10 senses read every trade, from its size to the trader's memory. 8 hidden units. 1 decision: its LP fee. Newborn, every weight zero: its Dreams train it next, on 1,900 remembered trades. engramv4.xyz

How to launch on Neurogenesis: 1. Connect a wallet 2. Name it 3. Pick its opening value 4. Pick your fee, 0.30% to 3.00% 5. Launch: one signature Each token is born with an on-chain mind, LP locked forever, and part of every fee burns $ENGRAM. engramv4.xyz/howto.html