Contribution networks: who uses your work, and who just rents your resource
Published
The meaningful difference between these networks is not what you get, it is what your contribution becomes. Some rent out your internet connection and resell it. Some hire you to label one company's private dataset. With Tendril, your spare compute and your judgment help build open AI, and the data is a public good the whole network shares in.
| Bandwidth-sharing apps | Paid data-labeling | Tendril | |
|---|---|---|---|
| What you contribute | Your internet connection | Your time annotating | Spare compute, or a few taps |
| Who uses the output | Companies that route traffic or buy the data | A single company's private model | Open AI, built in the open |
| What happens to your work | Resold as bandwidth | Locked in a private dataset | Released as open human-feedback data, a public good |
| Setup | Install an app and create an account | Apply and onboard | One click, no wallet or crypto needed |
| What you get | A token or cash for the resource | Cash per task | TEND Points, proof of contribution, not money |
Bandwidth-sharing apps: you rent out a resource
Bandwidth-sharing apps like Grass, Nodepay, and Honeygain let you share the part of your internet connection you are not using. That connection is then resold and routed to whoever buys it, for example to collect public web data or deliver content. The structural fact is simple: you lend a resource, someone else uses it, and the work leaves nothing behind that you share in. Some networks reward you with a token, some with points; that reward is for the resource, not for anything you helped build.
Paid data-labeling: you work on one company's private data
Paid data-labeling platforms like Scale AI, Appen, and Outlier hire you to annotate data for a single client. You compare answers, draw boxes, or rate outputs, and the result goes straight into that one company's private model. It is real work for cash per task. The structural fact: the dataset you helped create is owned by the buyer and locked away. Nobody outside that company sees it, and you do not share in what it becomes.
Tendril: your work becomes a public good
Tendril does not rent your connection and does not feed a private model. You lend a little of your device's spare computing power in a browser tab, or you tap your phone to judge which of two AI answers is better. Both produce open human-feedback data that helps build open AI. The output is released as a public good the whole network shares in, not a resold resource and not a locked private asset. You also help build AI for languages most models barely speak.
Where Tendril is different
Three structural things set Tendril apart from both columns above. First, the output is open: the human-feedback data the network creates is a public good, not bandwidth for resale and not a private dataset. Second, the on-ramp is zero-wallet: one click and you are in, with no app to install, no account to apply for, no crypto knowledge needed. Third, the reward is proof, not a price tag: TEND Points are your rank, your reputation, and your proof that you helped build open AI. They are not money, and we never promise payouts.
Go deeper on the comparison: how Tendril differs from bandwidth-sharing apps · what a Human-Feedback DePIN is · What is Tendril.
Contribution networks: common questions
- What is the difference between Tendril and bandwidth-sharing apps?
- Bandwidth-sharing apps share your unused internet connection, which is then resold and routed to whoever buys it. Tendril does not touch your connection. You lend spare compute or tap to judge AI answers, and the output is open human-feedback data, a public good the whole network shares in rather than a resold resource.
- Do I get paid to contribute to Tendril?
- No. You earn TEND Points, which are proof of contribution, not money. We never promise a payout.
- Do I need a wallet or crypto?
- No. One click and you are in. No wallet, no crypto, no setup.
Ready to help build open AI? Open a Tab on your computer, or play Tap to Train on your phone. Do one or both.