Tip:
TipiTip came with a simple idea: people who write things on the internet should be able to get paid for the sentence — not the article, not the season pass, not the parasocial subscription. The sentence. The one paragraph in three thousand that made you sit forward an inch.
This is, on its face, a stupid business model. Everyone successful at content monetization has converged on the same playbook: build an audience, lock it behind a $10/month paywall, ship enough content per week that nobody feels comfortable cancelling. TipiTip asks for none of that. It asks you to write the thing you were going to write anyway, and to have one extra column on your spreadsheet at the end of the month.
This happens to be the first article on TipiTip written by the project's creator. That should make you suspicious. The founder's post is the genre that has destroyed more Web3 projects than any rugpull. By the time someone writes the words "today, I am proud to announce," the project is usually six weeks from a quiet shutdown and a polite goodbye email. Nothing is being announced here. A useful tool exists. This is the explanation. The build took eight days, because slowing down long enough to plan properly would have required, well, slowing down.
Here is the part that is actually interesting, and not the part a pitch deck would lead with.
When you "like" something on a social platform, you are sending a free, infinitely cheap, easily forgeable signal that someone else's ad-targeting algorithm interprets at its convenience. When you tip a paragraph on TipiTip, you are signing a transaction with your wallet for somewhere between $0.001 and $0.05, and that signature lands on a public blockchain. The cost is functionally nothing for a human reader who liked the paragraph. It is, however, no longer zero — and that tiny shift away from zero changes who is willing to show up.
A determined bot farm can still fake a tip if it wants to. It needs an EOA, a funded cUSD balance, and the willingness to leave a permanent on-chain record of its taste in writing. That makes tip-bots an investment, not a one-line script — which is the part that matters. A medium where appreciation costs one tenth of a cent will sort signal from noise more cleanly than a medium where appreciation costs zero, for exactly the same reason a parking meter sorts parking from squatting. The meter doesn't make squatting impossible. It just makes it stop being free, which is enough to thin the crowd.
The friction, in other words, is the feature. It does not catch the patient adversary. It catches the cheap half of the noise, which turns out to be most of it.
Per-paragraph also matters separately from per-tip. Every other tipping rail — Buy Me a Coffee, Patreon, the dignified "support this writer" buttons — asks you to evaluate the writer in aggregate and pay them in aggregate. The information that flows back to the writer is roughly: "you exist, I approve." Useful. Pleasant. Mostly worthless as feedback. If, instead, a reader puts a heart under the third paragraph of a long essay and skips the other twelve, the writer learns something specific. They learn what worked. They learn what to do more of. Subscriptions intentionally hide that signal. TipiTip exposes it.
On the writer side, the pitch is even more boring. The boringness is the appeal. TipiTip is not asking you to grow a list. Not asking you to post on a schedule. Not selling you a course on how to find your voice. The contract is: you publish the post you were going to publish, and the rails make sure it can also be tipped. The added effort is one of two things — either publishing on TipiTip directly (a markdown box and a Publish button), or staying on Substack, Ghost, dev.to, or your own personal site, and dropping the embed in with two lines of React. The platform takes 0%. It does not own your audience. It does not own your content. The contract sweeps tips straight into the author's claimable balance — nothing sits in the middle skimming gravy.
Here is the honest version of what is coming next, in descending order of how much the timeline can be trusted:
- V2 of the smart contract will add modest anti-spam rails — per-address tip rate limits, a minimum wallet-age for being counted as a "verified" tipper, soft outlier detection on suspicious flurries. They were deliberately left out of V1, because at six readers and one writer, anti-spam gates are theatre. They turn on when the audience exists to justify them.
- Writers will be able to receive tips in local cUSD-pegged stablecoins — cKES, cEUR, cNGN — without touching the contract. Same rails, different units of denomination. This is mostly a UI problem at this point.
- The Farcaster Frame already works inside Warpcast casts. The embed package already works on any React site. Both shipped this week. If "is shippable" is the bar for "ship", much of this has already shipped.
Translation: this is not a roadmap. It is a list of things either already running or one merge away from running. No Q4 milestones are being promised. The pointer is at the part that already runs.
If you have read this far, then by the geometry of the thing, you can now tell the writer which sentence landed. There is a small heart under each paragraph. Tapping it sends between one tenth of a cent and five cents, and it tells the writer, very precisely, what to write more of. It is the most direct feedback loop between writer and reader currently available on the open internet. If none of the sentences landed, you tip nothing, both parties move on with their lives, no hard feelings.
TipiTip exists because that loop felt missing from the rest of the internet. The first article published on it is this one, because the only way to know whether the loop works is to put some writing inside it and see what happens.
The control group is every other newsletter you read. The treatment group has hearts under each paragraph.
— Written by the same person who deployed the contract, signed in this time from a wallet they actually use.