Democratizing AI with Conversation Markets
SEED Use Cases
Use Case: Autonomous Cars
Consider current solutions to customer service. For example, you're driving on the highway when you hit a piece of debris in the road. Your car starts to shake and you pull-over to the side of the road. Most people don't know how to diagnose what's wrong with the car and whether or not it's safe to continue driving—even just to the closest mechanic.
In a world of bots, a driver could call the manufacturer's customer service department and the customer service bot for the manufacturer would engage in dialog to help the driver determine the best course of action. Currently, the solution involves finding and making a phone call or touching a call button somewhere in the cars' control. However, in the very near future, the car, itself, may initiate the call or the car may be it's own bot (conversational user interface). Drivers may talk with their cars and the car may perform diagnostics and communications with other bots on the network (or partner bots) to advise the driver what to do.
If the damage isn't severe the car might automatically diagnose and log the incident, direct the driver to the nearest repair site, and negotiate the repair between the manufacturer and the mechanic (based on the current warranty). All of this would be done through authenticated communications. If the damage is too severe to continue to drive the car, the car could automatically call for roadside assistance, negotiating time, rates, and location and then alerting the driver.
In the case of autonomous vehicles, all of the bots' communications would be automatic, while keeping passengers informed and, possibly, calling another car to pick them up to complete their journey before repairs are made.
In all of these cases, these bots not only share valuable information (and, therefore, need to be secure and trusted) but they must interact appropriately with a range of systems, not the least of which are the drivers or passengers. Many of these bots will live "behind the scenes," in the backend of these systems, but several will interact directly with people and many of these companies will design their bots with appropriate social behavior, as well as character and appearance characteristics that embody their brands and desired customer relationships.
In addition, developers of many kinds will solve challenges at many levels that are universally applicable to other developers and bot-owners. The same social behavior that BMW uses in its 3-series car bot could be used across their many brands, changing only the appearance so that their Mini and Rolls-Royce bots function reliably (allowing for the specific differences in these cars) but with different personalities. Rather than create each bot from the start, all of the components and elements of any bot's personality template can be used in the creation of others. This bot description framework allows more rapid create of sophisticated bots.
Any developer can then sell the various components they've created to any other developer and earn value for each use. While companies like Apple and BMW may want to create custom and proprietary bot components that only they can use, an enterprising developer can solve these challenges and offer them to the entire marketplace of companies building and deploying bots within the same framework. This creates an easy, reliable market for valuable work.
Just like text written in books or available online at places like Wikipedia, or music available through a variety of sources, many of the components needed for sophisticated bots can be used, shared, reused, and recombined from various elements (from specific components to whole bots), repaying the authors for their work. While some authors may share bot code for free (open-sourcing it to anyone), others may share it via a variety of Creative Commons levels or charge (and be compensated) for every use their code is put to.
Because the bot framework is universal, bot code for behaviors, social interactions, use flows, and appearance attributes can freely move between industries and applications. The same analysis routine used in a customer service bot could be used in any customer service bot—or even parts of a healthcare bot.
Lastly, the drivers (or passengers) using these bot-enabled cars are generating data and metadata about their purchases, driving habits, preferences, emotions, and many other attributes. Without a system for protecting their privacy and/or fairly compensating them for sharing this data, the system continues to be an unbalanced and exploitive one, putting people and customers at a disadvantage. This is, of course, extendable to any data and metadata generated by people's interaction with systems, across all uses and industries. The SEED network aims to form the backbone of a transaction system that authenticates, shares (or not, as users define), enables, and fairly compensates people for their interaction with bots (and other) systems.
While there is a platform (Botanic Technologies, Inc) that has been deployed for this use case in Toyota cars currently, there is no network, or marketplace for these components or interactions, making bot development and deployment costly and slow. In this way, these elements democratise and accelerate the bot market and the SEED token becomes the transactional fuel for the market.
It's worth noting that the above example is already in development by some automobile companies. And, it is not only automobiles and transportation systems that are evolving to support (and be supported by) bot implementations. Nearly every industry in the world holds opportunity and interest for bot technology.
Use Case: Buying and Selling Bot Configurations
Jason is a bot developer, working on building the use flow for his company, an insurance provider in the commercial real estate market. This is a large, complex use flow that integrates real estate law with insurance prediction models and appropriately social interfaces. In the bot store, he's found a use flow for a real estate insurance case and bot it as a template to build on top of. He's customized the use flow for the commercial space and plugged-into his company's API to their proprietary prediction system. He's also found a set of use flows for basic social responses (greetings, escalation, etc.) and purchased these to integrate into his bot. He has to add use flows for some major clients with specific needs, but he's considerably further than if he had to start from scratch. He's also more confident in the sections he's built, since they came from highly-rated sources.
His bot also connects to a language translation interface he found in the bot store to handle the many customers his firm has in Mexico and Quebec. While he doesn't need a visual appearance for this bot, it's specified to speak to customers so he's able to find a variety of voices that fit his company's customers, brand, and needs. The voice-to-blockchain functionality automatically available from the SEED network allows his bot to take direction and complete actions with agency and go beyond the simple advice chatbots that their competitors have.
The developers of the components Jason has assembled his bot from have been automatically compensated for their uploads according to the prices and licensing terms they specified. Jason likes one component so much that he leaves a glowing review about the easy of integration and his company is charged a small amount to host the review.
Use Case: Customer Service
Tran just got home to discover her new flat panel television has been delivered via Amazon but the box is pretty dented and when she opens it, sure enough, there's a crack on the corner of the screen. She's disappointed but, "these things happen," and she calls Amazon customer service. Amazon's automated voice system answers and asks her how it can help. She described the problem and through keyword identification and answers to specific questions, it identifies her account, with all of the associated purchase and tracking data.
The conversation is pretty easy and the system authorizes a return as well as starts the process for delivering another one. In this case, UPS will not only deliver the replacement but pick-up the broken one. A specific note within the instructions tells the UPS driver that there was a problem the first time. This is all possible because Amazon's service bot is authenticated, during the transaction, on UPS' system (and to a service bot on its side). Both bots are verified via the SEED Network since they share the same protocols.
This creates an outline for the democratization of AI via conversation markets as well as enables participation by smaller developers within an industry increasingly dominated by large corporations.
Use Case: Branded Personality Bot
Di-an is finally able to plan a vacation for her, her husband, and her young daughter. She's never been to Australia but she and her husband have always discussed it. She wants this trip to be special, as it's the first one in several years, so she wants to be sure it's fun and anything but ordinary. She's always admired the Virgin brand and she knows that Virgin flies out of Shenzhen, where she lives. She has no idea what to do in Australia but she doesn't want to do the same things her friends have done on their trips (none of whom have traveled to Australia).
When we gets to the Virgin Australia website, it automatically serves her pages in Chinese, seeing her IP address (as well as English, just in case). However, the text and images are immediately funny, interesting, and exotic. A pop-up message with a character named Veronica gets her attention by winking with a short whistle. She clicks on it and it launches a conversational character that is just enough respectful, just enough sassy, and even more hip than the students who hang-out at the nearby mall. It asks her about her dream trip and suggests a flight to Brisbane for a few days and then on to Alice Springs for a memorable family trip. It asks her many questions, remembering and referring later to her answers all with an easy, fun personality. It's everything she has come to expect from Virgin from their television ads—and more (easier to use, more informative, more personable, and more customized)!
Before she knows it, she's booked an affordable, unique trip, complete with lodgings, flights, and a tour package of Alice Springs as if she's been talking to her younger sister—only as a travel agent. For once, she feels like the one "in the know" (as opposed to her sister always being “up” on the latest trends). The itinerary, preferences, and even transactions are all handled within the conversation and she quickly receives a notice on her phone with links to all of the details, as well as introductions to other Virgin bots to help her once they arrive in Australia. She feels an excitement she knows she wouldn't have felt if she were flying China Airlines and feels like this is the right trip for her family, as opposed to something more utilitarian that she's sure she would have been suggested to her by a travel agent or customer service agent at any of the other airlines serving Shenzhen.