Baxnet Ideas · Founder note
The First Customer for Personal Intelligence Should Be Mimoto
TL;DR: Before Personal Intelligence becomes useful to outside researchers or businesses, Mimoto can test the model itself: ask for one defined contribution, offer a fixed reward and return a better benchmark.
Suppose Mimoto calculates that your typical response time in a long-running relationship has changed.
The number may be useful on its own. The next question is still obvious: compared with what?
One private conversation can show how a pattern changed for one person. It cannot tell that person whether the change is common across similar relationships, unusual for that stage of a relationship, or mostly an artefact of the way the measurement was calculated. For that, Mimoto needs a population benchmark built from measurements produced in the same way.
That creates an interesting problem. The benchmark would make Mimoto better, but users would create the raw material for it.
I think Mimoto should treat that contribution as worth rewarding. Mimoto would become the first customer for the aggregate insight, with Baxnet funding the exchange.
This is a secondary, experimental route to value—not the main economic case for Personal Intelligence. Mimoto’s first job is to make a person’s own history more useful to them through better recall, context, and decisions. A contribution economy should be tested only where a narrowly defined aggregate can improve that private product and the user can see exactly what leaves the device.
A comparison needs more than a large pile of numbers
A credible benchmark cannot be assembled by collecting a few vaguely similar statistics and averaging them together.
The relationships have to be meaningfully comparable, the observation periods clear and response time defined consistently. A measurement based on a few weeks of messages should not quietly sit beside one based on ten years of conversation as though they describe the same thing.
This is where the idea becomes more concrete than saying personal data has value.
The first experiment we are designing asks a narrow question about a user-confirmed, long-running romantic conversation. It would compare the participating user’s median self-response time in the first and tenth complete observed years of that chat. Median is a better primary measure than a simple average here because sleep, work, travel and long gaps between conversations can distort an average rather badly.
Mimoto would determine eligibility locally. Before anything was shared, the user would see the purpose, the method, the exact derived fields leaving the device and the fixed Mimoto Credits offered in return.
No messages. No chat title. No names, topics, identities or exact timestamps. The first version would contribute the participating user’s response distribution, not the other person’s. Having somebody’s messages on your device does not automatically give you a right to sell intelligence about them.
The reward would also stay the same whether the result was fast, slow, surprising, boring or inconvenient to whatever theory Baxnet hoped to test. We would be paying for a valid contribution to a declared question, not a desirable answer.
We have not shipped this feature. I am writing about it now because these boundaries become much harder to change once personal data and rewards are already moving through a product.
Baxnet can become the first buyer
Usually, discussions about people earning value from data jump straight to a marketplace filled with advertisers, insurers, researchers and mysterious buyers.
That is too large a first step for information drawn from private relationships.
Baxnet has a simpler reason to fund the first contribution opportunity: Mimoto needs a trustworthy benchmark to produce a more useful comparison. Baxnet can state the question, prefund the reward, publish the method and accept only contributions that meet the same contract.
That makes the exchange visible. The user is not asked to donate an archive in the hope that it becomes useful later. They are offered a specific opportunity with a defined purpose, an exact outbound packet, a fixed reward and a permanent receipt showing what happened.
The contribution can then join a versioned benchmark library. When a compatible benchmark is ready, Mimoto can download the aggregate distribution and compare another user’s locally calculated result against it. That person should not have to upload their own private archive merely to learn where their measurement sits.
The benchmark comes back into the product.
That last part matters. Users would not only receive Credits for accepted contributions. They would help create a feature that gives people better context about their own results. More compatible contributions could improve coverage across relationship duration, messaging source, region and carefully chosen optional demographic bands. Better coverage would make the comparison more useful, which would make contributing to the next bounded question more worthwhile.
If that loop worked repeatedly, it could create a network effect. The defensible product advantage would not be the raw contributions alone. It would come from the combination of permission, consistent measurement, provenance, cohort quality, transparent rewards and a product capable of returning the aggregate value to users.
The wider market can wait until the first exchange works
If Mimoto can validate this model with Baxnet as the first buyer, a longer-term possibility becomes worth testing.
A researcher or another approved organisation might eventually commission a tightly defined aggregate question. Existing benchmark releases might support public research, reports or carefully governed insight products. A valuable reuse could even justify another form of contributor reward later.
None of that should happen because a user once pressed a general-purpose consent button.
A new buyer, purpose, field, licence or reuse rule should create a new decision for the user. Outside organisations should normally receive an approved aggregate output or controlled result, not the underlying contribution store. Uses that score named people for employment, credit, insurance or other consequential decisions do not belong in this model.
That makes the first internal experiment important. It gives us somewhere small enough to discover whether the exchange is genuinely useful, whether the reward feels fair, whether the method survives scrutiny and whether people understand what they are agreeing to.
I have argued before that personal data should become a productive resource for the person. This is the first practical version of that belief I can see Mimoto testing.
Control begins with the ability to refuse access. It should also include the ability to approve one defined use, inspect what left, receive something in return and remain free to refuse the next request.
If Mimoto can make that small loop work, it will have done more than improve a response-time comparison. It will have shown what it means for Personal Intelligence to create value for the person it came from.