The Human API
This is part two of a series, the first part was The Rise of Emotionally Intelligent AI, which was also featured on VentureBeat.
API is a commonly used abbreviation for Application Programming Interface. In simple terms, APIs enable software applications to request information or perform tasks in other software applications.
The Internet is already full of APIs. They are used for everything imaginable: online shopping, payments, search, messaging, calls, advertising, and so on.
So far, the APIs have been used to build connected services that help people use online services more effectively, and to transfer data between applications. APIs have been a way to unleash the power of data as a tool for us humans to be more productive.
As artificial intelligence has progressed, we are now at a point, when algorithms drive the fastest growth and largest profits for businesses. They enable much, much faster scaling of businesses than what is possible with human labor alone.
As algorithms have enabled businesses to approach millions of people cost-effectively with individualized messages, the business models have evolved from selling content into monetizing attention and engagement.
This evolution has given rise to a whole new class of developers, whose job is to develop algorithms that grab your attention, keep you engaged, extract data about you, make you perform tasks, and convert you from free user to a paying customer.
These developers see you as an API to plug into.
The available Human APIs
I have identified five main APIs to humans available to algorithms: Attention, Engagement, Data Extraction, Task Execution and Monetization.
To use these APIs, the algorithms have to connect to people via four human interfaces. They include the senses of vision, hearing and touch, and sensors placed on people. The first three of these interfaces are available via any smartphone, and the fourth by wearable devices like smart watches.
According to Statista, there will be 2.87 billion smartphones by 2020, so we can assume this to be the amount of available Human APIs then.
Would you like to program a human to do whatever you want?
Fasten your seat belt and prepare for a ride. In the next chapters, I am going to explain how to turn millions of human beings into robots controlled by your algorithms.
The Human Attention API
To start the process of turning humans into robots, you need their attention.
The best ways to gain the initial attention of people is to display a piece of content, either image, video or text on a social media feed or on a search results page on Google or Bing.
A more aggressive way to gain their initial attention is to send them a cold message over chat or email. Please note that this approach is risky and can backfire, causing anger and negative emotions on humans.
Both of the above strategies can be implemented with a number of currently available services, which are already connected to the APIs of major social networks, search engines and message delivery services.
There still are numerous traditional strategies for initial attention like TV, radio, newspapers, and outdoor ads, but they do not support automated conversion funnels, so they are not compatible with the Human API.
The initial message conveyed has to be tailored for the recipient, as different people have different interests. If you are looking for shortcuts, there are some time-tested strategies for general targeting.
For example, pictures of beautiful female humans in bikinis will generally attract the attention of 95% of the adult male population. Videos of kittens and puppies will attract the attention of a large portion of both male and female populations.
While these shortcuts may get you the initial attention, they don’t necessarily help you convert people for long term engagement.
If you desire the continuous attention of people, the initial message has to be related to the activity you wish to engage them in, and to the results you wish to achieve with that engagement.
More importantly, your initial message has to motivate the targeted people to do one of the following actions listed in the order of difficulty: 1) follow you, 2) add your chatbot or 3) download your mobile app.
Because the conversion rates from initial attention to these actions are low, this is a numbers game. Your algorithms have to continuously send different messages, testing various content and channel strategies to maximize the conversion rates for each target demography.
The Human Engagement API
So now you’ve captured some humans with messaging algorithms. To make them your slaves and controllable with your algorithms, we need the Human Engagement API.
Today’s smartphone notifications use all of the three available human senses. They beep, they vibrate, and they show the notification on the phone screen. The Human Engagement API leverages smartphone notifications to lock human attention for longer periods of time.
This sounds easier than it is. People only accept notifications that they want to see. If your algorithm sends messages that people don’t like, they will block the notifications.
There are two good strategies to avoid notification blocking.
The first strategy is to deliver notifications via an existing messaging app, for which people have already enabled notifications. This is great, if you got them to follow you or add your chatbot on an existing messaging app.
If you got the human to install your app, it is very important to enable and encourage messaging in it. The desire to receive messages from other people will motivate them to enable notifications in your app.
Once you’ve locked in the Human Engagement API, congratulations! You have the full attention of your captured humans at your disposal, all the time.
On average, a human touches their smartphone 2,617 times a day.
Now that you have successfully captured some people with the Human Attention API, and enslaved them with the Human Engagement API, it is time to benefit from their attention.
There are three main methods to benefit from the enslaved humans. You can extract data from them, make them work for you, or make them pay you. The following chapters describe each of these methods.
The Human Data Extraction API
The personal data of people is valuable. It can be sold to advertisers and businesses who can offer services with that data. This is why it makes sense to use Human Data Extraction API.
To be able to successfully capture private data, it is important to understand that with people, it is all about incentives. People can be incentivized with money, fame and social approval. That may sound challenging, but don’t worry, there’s a great loophole.
None of these incentives have to be real.
Game developers figured this out a long time ago, and make billions with algorithms that reward people with virtual money, virtual fame, and virtual social approval for their data, work and money.
Virtual rewards work amazingly well with people. For example, Facebook algorithms make sure that each wall post gets a few likes almost instantly to provide people with virtual social approval. This keeps the data flowing.
The virtual fame reward is perfected in Quora, which offers “Top Writer in #insert-your-topic” badges to people who contribute popular answers to a specific topic. For the folks who contribute the most, they even send a jacket.
You can extract almost any information from people, if your algorithms provide them with proper virtual rewards.
If you want to extract health data of your captured humans, the best strategy is to capture the information passively via wearable sensors attached to them.
To enable this method, you will have to offer virtual rewards. This can be provided automatically in the form of virtual patting on the back and by sending messages with words of encouragement.
Algorithms can also provide real rewards.
With the rise of cryptocurrencies, it is now becoming possible to provide real monetary rewards algorithmically by dropping cryptocurrency tokens to people for data provided, or task executed.
The value of these rewards depend on the market value of the cryptocurrency being used.
The Human Task Execution API
If your target is to turn the captured humans into your slave labor, you are in luck. The Human Task Execution API is here to help.
This API requires real monetary rewards, as people need to pay for their food, clothes and shelter. But the required rewards are very low, and usually can be easily monetized at a much higher rate with the Human Monetization API.
The Human Task Execution API works so that an algorithm delivers pre-defined tasks to people against a small monetary perk via the Human Engagement API, i.e. smartphone notifications.
Uber is a great example of a successful implementation of Task Execution API. Uber algorithms send a notification to nearby drivers when a ride becomes available, and provides them with the optimal route to drive too. Human in this case is a biological robot that the algorithm programs to move the car to the desired location.
Another successfully implemented Human Task Execution API is Amazon Mechanical Turk that lets your algorithms to send almost any task to hundreds of thousands of people all over the world. As humans compete for getting the task done, you usually get the job done quickly for pennies.
All online shipment and food delivery platforms have also implemented the Human Task Execution API.
The Human Monetization API
Last but definitely not least, there is the Human Monetization API. This is a powerful API that enables algorithms to squeeze money from captured and enslaved humans.
The most successful companies use it already.
If your algorithms captured data with Human Data Extraction API, the data can be automatically monetized by selling it to the highest bidding advertisers. This strategy is successfully used to generate billions in profits to companies like Facebook, LinkedIn, Google, Twitter and many others.
If your algorithms use the Human Task Execution API, this labor is usually pre-sold to other humans at a premium. This strategy is used by companies like Amazon, Uber, Postmates, Delivery Hero, Takeaway, JustEat, GrubHub and many others.
It is also possible to monetize the Human Engagement API alone. Game companies like Supercell use notifications to engage people to perform timed tasks in a game. The algorithms then convince people to spend money to gain virtual money, virtual social approval and virtual fame.
Subscription services like Spotify, Netflix and many others use the Human Data Extraction API to learn as much as possible about the people using their services, and then the Human Engagement API to deliver them the content they want to see. This is the best way to ensure people will continue to pay for the service.
Thanks for reading this far. Now you understand how algorithmic businesses leverage the Human APIs to generate billions in profits, without the need for human labor. This is the future of business, and leaves you with the question:
How many of the businesses above already have you as a slave?
Berlin, 16.10.2017
Mikko Alasaarela
My company Inbot is among the pioneers that leverage AI algorithms to offer real long term monetary value to people for their data and services. We exist to counter the trend of intelligent machines enslaving humans, and to provide human opportunity in the age of artificial intelligence.
Join our Ambassador Community to earn long term dividends by introducing innovative data- and AI businesses to customers.