Fixing the Internet with the Personal Wave Transport Protocol

Is this Internet really the best we can do? Are we forever stuck with irritating ads and invasive data mining?

Now that Facebook has been caught red-handed capturing way too much data about everyone and then carelessly oversharing it, a few more people than usual are wondering what hell we’ve gotten ourselves into.

If, like me, you’re yearning for a better Internet experience, read on as I outline a massive upgrade to the idea of the social network and lay out a vastly superior replacement for our universally despised online advertising model.

The Shared Language Between Humans and Computers

The key to this puzzle is a different way of thinking that resolves the awkward relationship between analog humans and digital computers. It turns out there’s a language in which both humans and computers are highly fluent, a language found right on boundary between analog and digital.

The language? Digitized waves. Behavior waves, to be more specific.

A person’s lifetime of behavior can be viewed as a complex wave, which is itself a large bundle of individual waves. This shouldn’t be a surprise because quantum physics tells us that literally everything can be viewed this way.

Each individual wave represents a unique type of activity that can be isolated and digitized. While we can’t reconstruct the history of the wave, we can begin digitizing any wave at any point in life. Digitization is simple: pick any activity in your life and then record on a computer each time you choose to do it. This results in a timeline (or time series).

Each time series can then be transformed into a wave function by using a linear regression, a Fourier transform, or one of many more complex methods. We are then free to transmit this wave in digital form.

I’ll explain why shortly.

By putting human nature front and center rather than as an inconvenient afterthought, this different way of thinking enables our current technology to be a thousand times, or even a million times, better at making high quality connections and predictions than now.

I’ve distilled this perspective into a new high-level Internet protocol so that we can begin prototyping brand new types of online applications and services.

In case you’re wondering, a protocol simply defines how our computers communicate with each other, such as HTTP for websites and SMTP for E-mail.

Wave Packets

This new protocol, which I’m calling the Personal Wave Transport Protocol (PWTP), allows us to represent, transmit, and receive wave packets. Each wave packet is composed of a wave function (as simple as sin(x)) and details about what the wave represents, who it belongs to, and how it is to be used. You’ll find more detail near the bottom of this piece.

Each of these wave packets represents a timeline of actions you’ve taken in regards to a specific activity. For example, an activity might be “grocery shopping”, and you do it once a week, usually on Sunday. Or the activity might be more detailed, like “buying peanut butter”, which might be every three weeks. The activity could even be ultra-detailed, like “buying 1 KG Adams crunchy peanut butter for $6.99 at the Safeway on Broadway St. in Vancouver BC”. The activity itself and the degree of detail are entirely up to the person constructing the packet.

The underlying idea is that most of what we do each day is part of a cyclical and fairly regular pattern. Breakfast is daily (pancakes on Saturday), grocery shopping might be weekly, oil changes, vacations, and doctor’s visits annually, water-heater replacement every decade, and so on. By merging the timelines of our many thousands of activities, we get the essence of our personalities, thoughts, and life itself.

The Personal Wave Transport Protocol (PWTP) provides a way to share future projections of your past choices at the individual-activity level. This makes the projections surprisingly, even spookily, accurate. And arbitrarily rich, as well, since any combination of individual waves can be either collapsed together or searched for correlations and other patterns.

If you record each time you go grocery shopping, you end up with a series of data points. From those data points, it’s easy to generate an approximate wave function to be transmitted via PWTP (and re-transmit when you do the activity again). The recipient is then able to use that wave function to project into the future when you’re likely to do that activity again.

Yes, you read that right: the purpose of this protocol is to enable other people to accurately predict your future choices. And yes, this is how we fix the Internet for everyone. Allow me to explain.

A Replacement for Online Advertising

While we all despise online advertising in its current form, we nonetheless crave its primary function: making good, timely connections. Whether between buyers and sellers, producers and consumers, or influencers and audiences, we crave high-quality connections.

Even though we like the “free” services that online advertising is funding, we’re becoming a bit wary of the risks of our personal information being surreptitiously captured and shared so freely.

PWTP offers a few critical features that, together, enable it to connect people much more effectively, safely, and efficiently than online advertising as we know it today.

First, PWTP swaps the locus of control from the service provider (e.g. Facebook, Twitter, Snapchat) to you instead. Now you decide which activities you’ll transform into waves, how detailed and accurate they’ll be, who you’re going to share them with, and what the recipients are allowed to do with them. Mostly you’ll choose to share waves because a PWTP service promises to make your life easier or better. Or they simply offer to pay you for your information. Either way, now you expect value in return.

Second, PWTP deals with high quality data because it’s direct from the source: you. The more specific or important your need, the more detail you’ll be motivated to provide and the more care you’ll take in keeping it fresh. This is quite unlike the incomplete, stale, and basically stolen information used by online advertising systems. Have you ever wondered why supposedly targeted ads are so often irrelevant? Low quality data is why.

Third, PWTP masks the details of your past. While you could choose to transmit your timelines directly instead of an approximate wave, that level of detail would reduce the control you’d have over your data after transmission and introduce serious privacy concerns. Wave functions deal only with the future and can’t be meaningfully backtracked. Furthermore, the informational value of a wave naturally decays until you re-transmit it.

And finally, PWTP is service-based like HTTP (Web) and SMTP (Email). That is, there could be any number of PWTP service providers that would accept your wave packets, process them in some useful way, and return value to you. These service providers could be public or private, free or paid, big or small, professional or amateur, permanent or transient. Just like websites and equally as varied.

A PWTP Service Example

Let’s say you buy peanut butter once a month and there’s a PWTP service that helps you find the best price for groceries. This service receives wave packets from many people, processes the packets, and sends notifications when it makes a valuable connection. The service may share the patterns with grocery stores or other organizations (provided you permit it).

Today, after you bought peanut butter, you simply added a new point to your timeline and re-transmitted your wave packet to the grocery service.

The wave you just transmitted tells the service that your peak probability for purchasing a jar of peanut-butter is a month from now. So it waits. In four weeks, the service notices that you are near peak probability. You’re open to buying another jar. After all, you can plainly see that your current jar is nearly empty.

If all the service knows is that you’re open to peanut butter, it may share the data with grocery stores so they can reach out to you (and everyone else who is open to peanut butter right now) directly with an offer. A bulk discount, in your inbox, exactly when you need it? Very nice!

Or the service may send you a well-timed message that Jif peanut butter is on sale for $5.99 in Store A and Kraft for $5.49 in Store B. How might the service know the brand, the price, and the store? Well, they could be receiving pricing data from various local stores or by parsing their sales flyers. Or, one or more other people chose to include those details when they recently bought peanut butter and re-transmitted their wave packet. In any case, whatever the service knows, it can share.

Why would people include those details in their wave packets? Because they wanted better results from the service. The more detail available, the better the connection can be. If you only wanted Adams peanut butter, you’d be sure to include that detail. But if you didn’t care, you’d have no need to provide it. If you wanted to know about a better price, you’d include the price you bought it for last. And so on.

High quality data changes everything; it’s simply never been tried. But that’s the basis of what PWTP offers.

Imagine how much better we could do than our current methods of online advertising when all we have available is high quality data. Imagine what new services we could build atop this protocol.

A Different Kind of Social Network

Like advertising, social networks are all about connections. Search is too but I’ll leave that for another day.

Group Activities

Let’s say that instead of groceries, you maintained a timeline of a more social activity like going out for drinks (or coffee, mountain biking, yoga, running, board gaming, or anything really). So now it’s easy to predict when you’ll next be open to going out for drinks again.

If you shared your probability wave with your friends, they could easily see when you’d most likely accept their invitation. If all your friends shared this activity in the same place, you could all see when most of you are open to going for drinks.

If each friend shared their wave with a special service, the service could automatically recognize when a bunch of you have the same interest at about the same future time and automatically propose a day, time, and location for you all to meet. This would work for any social activity, whether online or offline.

Imagine if that became a well-polished service. Better than Facebook, right?

Actual Status

“How are you doing?”
“Great, thank you. How are you?”
“Never better!”

At least once, you’ve said you’re doing great when really you weren’t, but it wasn’t the right time, the right place, or the right person to open up to. At least once, you’ve posted a damn lie as your Facebook status or latest Instagram photo. At least once, you’ve suffered in silence because you didn’t want to burden your friends or family or you were embarrassed by your problem being too personal, too weird, or too minor.

But with PWTP, you can share information about yourself in a way that lets you get support from your friends without entering the Too Much Information zone.

Pretty much every problem we have is, at the very least, made tolerable if we just do the right thing on a regular basis. If we’re depressed, maybe we have to do better at getting exercise and taking our medication. If we’re getting fat, we’d benefit by swapping out that big bag of Doritos for a small bag or swap that huge pasta dinner for a salad every now and again. Basically, for every problem we have, no matter how large or small, there’s a beneficial activity that we simply need to do regularly and often enough. Sadly, we usually know exactly what that activity is. We need help despite being adults.

Let’s say you’re getting fat and/or old, so you need to eat better, eat less, and exercise regularly. You could maintain three activity timelines then: healthy meals, half-portion (or skipped) meals, and exercise of any kind. For each activity, you’d assign a target frequency such as every other day (3 or 4 times per week). Every time you made one of those choices, you’d add a point to its activity timeline.

Now if it’s been 48 hours since you last exercised, it’s time to do it again. But you don’t feel like it, so you skip it today. Thankfully you’ve shared your probability waves with your friends, so they can see that you’re about to fall off the wagon. They don’t know exactly what you’re doing or not doing, but they do know the activity is important to you and that you shared it with them so they could help.

So your friend, or maybe all your friends, send you a text message to get off your ass and do that thing you’re not doing. And that might be all you needed.

Imagine if that became a well-polished, highly persuasive service. Better than Twitter, right?

A Bit Of Detail

PWTP involves each person in the network transmitting one or more wave-function packets, similar to how we send emails into the SMTP network and for similar reasons.

Like this:

Each packet is as simple as a timestamp, a sine wave, a wave period, and optionally a phase, with no upper complexity limit. Further, each packet is linked to a specific activity that the person does on a regular basis.

The packet might be structured like this:

PersonID: 1234,
ActivityID: 4321,
Timestamp: 1522349778,
Functions: [
Function: "sin(x)",
Period: 86400,
Phase: 0

Each wave function is re-transmitted when the underlying activity re-occurs. The peaks of the wave represent the highest probability of the activity occurring at that time while the troughs represent the lowest probability.

Like this, now with an amplitude decay:

What Comes Next

By now I’m sure you’ve noticed that this protocol is still fairly abstract and leaves many questions unanswered, albeit intentionally.

How do people construct and send their wave packets? How do the packets get to the services? How are the services discovered? How do the services process the waves? How do services send results back to a wave’s owner?

There’s a bunch of unanswered lower-level questions as well. How does a person define who can access their wave packets and what permissions are granted? Who defines the activities and details so that services can make proper connections? How are people identified uniquely while maintaining their privacy? And many more.

All of these questions will be answered as live examples and prototypes are developed and tested.

Right now, my intention is to introduce this different way of thinking and promote the idea that we have a practical way to reshape our corner of the Internet. The current Internet isn’t going to die and social networks will keep doing what they’re doing. Still, I’ve outlined a way forward that can begin as a very small group of people who appreciate this different way of thinking and are itching to get the ball rolling.