Our second consumer product, the Companion Sapiens builds upon the capabilities of the Caregiver Sapiens with an expanded world model and language handing capabilities. Initially, the Companion Sapiens is aimed at anyone who has many details to manage but as its ser interacts with it (and the core is updated) we expect it to become its principal’s trusted buffer to the connected digital world.
Like the Caregiver it will be cloud-based and designed to be constantly available to its “principal” via text or voice over a number of interfaces, a cellphone app, direct text (the sapiens will have its own telephone number) and via the sapiens’ own website. It will be sold as “software as a service via monthly or annual subscriptions.
While the current generation of “Digital Personal Assistants” have some capability to handle reminders or contacts or other personal information where they have been pre-programmed for it, they possess no general comprehension of language. For example:
If you tell SIRI: “I am parking my car on level 4 space A21.” It answers: “Ok, I will remember that you parked here.”
It records your current GPS coordinates and returns the location on a map if you ask: “Where is my car” or “where did I park my car?” Useful as far as it goes but it is really only responding to the preprogrammed symbols “park” and “car.” It throws away the rest of the original input and stores nothing about the level or space where you parked. If you tell SIRI something outside of its preprogrammed response it is totally unhelpful:
User: “I am hiding the key to the back door under the white rock in the garden.”
SIRI: “Sorry I did not understand ‘I am hiding the key to the back door under the white rock in the garden. I can search the web for it.”
When these applications get the right answer they are useful but since they have no idea what the user is actually talking about, when the algorithms don’t return an answer they are dead in the water, frustrating the user. Also, they don’t know that they don’t know (they don’t know anything, of course, there is a vast difference between knowledge and data) and so they often return a completely wrong result, frustrating the user even more. Most of the failings will be corrected as the companies that offer them license our Conversational Core product. However one problem still remains. SIRI will always work for Apple and Alexa for Amazon. Where is an AI that will work for the consumer? The answer: their Cloud Companion sapiens.
A sapiens on the other hand actually comprehends statements about common everyday actions such where you put your keys, parked your car and that you took your medicine. True language comprehension is a far more sophisticated process that the kind of key word matching chatbots like SIRI perform.
For example, in the hiding keys example our system extends its existing core model with a new instance of the action, hiding, along with an instance of key, back door, white rock and garden, (unless these particular things are already known to the model.) It then connects these things with their spatial and other relationships and to the core action. Once this is done it is capable of retrieving any of this information in response to questions in any common form.
Everyone has one or more methods or utilities for remembering or recording information they don’t want to forget whether it is passwords, contact information, appointments, where they put stuff or whether they took their medicines. This is most especially true for older people or anyone with health issues. We believe people in these groups will quickly find their Companion Sapiens to be indispensable.
- A full time, 24/7, personal assistant available at your beck and call day and night
- A partner with total recall and infinite patience?
- A companion as loyal and eager to please as a Labrador retriever but smart enough to talk with you in plain English
We intend to extend the functionality of the baseline product with a deeper, richer world model that encompasses the everyday common-sense world. We envision our first generation companion as an intelligent interface to technology while later generations will more closely resemble a virtual person. This personal virtual companion, secretary, and all-round helpful sapiens, which we believe represents a totally new kind of product, will be designed to focus solely on its user/owner.
Our objective is to provide every human being that has regular access to a computing device with their own personal sapiens. Our vision for this product is that it will become a virtual lifetime companion of ever increasing capability as it learns more and more about its principal and becomes more intelligent through core updates. This personal sapiens would be available to its user/owner (“principle”) from any internet connected device. We foresee it being given to children at a very early age (probably embedded in a toy like a “Furby”), becoming their playmate, teacher and safety monitor. We predict that adults will a have a powerful personal assistant/secretary and an intelligent buffer to simplify the digital complexities of modern life. For the elderly, we expect that it will become a companion, biographer, guardian and nurse.
When people imagine intelligent machines, they almost always assume they will be emotionally obtuse and unable to recognize emotional behavior in humans. We believe this is a misconception; our world model encompasses human emotional responses like all other phenomena in the world and can therefore recognize the emotional content in human language and respond appropriately. We think this will make interaction with a companion sapiens a delight and engender in every user an unprecedented level of attachment. When that companion has command and control interfaces to the digital world that surrounds us – the result will be like something right out of science fiction, like Jarvis the robot butler from the Ironman movies.
Personal Assistant Functions
We expect that the companions high level of language comprehension will enable it to serve as a communications hub for its owner, screening telephone calls, texts, instant message interfaces or twitters feeds. In our vision, spam email will be a thing of the past, as Companion will be able to read and delete (or store) anything it knows its owner does not want to see.
As a sophisticated schedule manager, daily routine assistant and personal coach, we believe a person’s sapiens will be unmatched, exceeding in many ways even the utility of human personal assistants because we predict it will always be there, cannot forget, and has no competing needs of its own.
Personal Relationship Manager
We predict that just as a sapiens creates a unique model instance of its owner, it would also do so for the important individuals in its owner’s life. We envision that although it will understand categories like contact, friend or colleague, it will not be limited by them. No more grouping people into broad categories– a sapiens will be expected to create a model of each individual its user interacts with and will help manage relationships and communications.
Personal Web Presence
A Companion Sapiens is predicted to have the capability to create a personal webpage as its user’s public face to the world. At a minimum, we intend for it to display a chat interface to itself so that visitors can get any information the user desires them to have depending on how they identify themselves. We foresee that specific persons or classes of persons will be able to be given access to the user’s personal information (status, location, photos, writings, etc.). We believe this can be developed to become an automatically generated personal website maintained through conversations and interviews with the owner; the ultimate in personalized social media.
General Reference Capabilities
As powerful and useful as search engines are, we believe there are a vast number of questions for which they are overkill. Typing the question “Where is Sochi?” into Google yields 923,000,000 results. While the answer can be found on the first webpage listed, the user still has to go there and read it (with the potential of being distracted by the abundance of other information including advertising).
Companion Sapiens’ understanding of language is expected to allow it to access specified databases or web resources to retrieve exactly the information needed. For a general question such as “Where is Sochi?,” we expect that Companion Sapiens (if it didn’t already know the answer) could access designated on-line sources such as the Simple English Wikipedia article on Sochi and read the first few sentences: “Sochi is a city in Russia on the Black Sea coast. It is in the oblast of Krasnodar Krai. The 2014 Winter Olympics were played there.” These sentences are well within the anticipated level of comprehension envisioned at product release.
Having learned the basics about Sochi, we believe Companion Sapiens will answer the question with precision and confidence it is providing the information asked: “It is in Russia on the Black Sea.”
We foresee wearable devices which monitor many biometric parameters will soon be ubiquitous. It is our view that the utility of such devices can only be realized to the extent that people monitor the data provided. We intend for one’s personal sapiens to be the ideal solution for managing the coming proliferation of personal health data, and we expect that exchanges like this one will become commonplace in our daily lives:
Companion: My friend, you’re tired, you have not had lunch, and your blood pressure is up. I suggest we reschedule the next meeting.
User: Okay. I agree, please take care of that for me.
Companion: Will do.
Auxiliary Memory Functions
An application that understands what you say to it and never forgets functions as its users’ “outboard memory” —from the moment it is acquired and with zero setup or learning curve on the part of the user.
Intelligent Interface to other apps and devices
While the above uses are compelling in their own right perhaps an even more compelling reason to purchase Companion is that it will serve as an intelligent, comprehending interface to other applications on the phone and, through them, to devices that they control (subject to required licenses, approvals and authorizations of third parties).
There are two example classes of applications that we believe would benefit immensely through integration with Companion: “Eyes-off” and device control applications.
“Eyes-off” the phone apps
We believe that virtually any app that a person would like to use without looking at the display on the phone is perfect for Companion integration. These would include communication, navigation, control of entertainment and environmental systems or anything else that a person would like to do while driving, or any situation where calling up the app and interacting with its user interface on the screen would be inconvenient or distracting.
Device control apps
Today is seems that every electrical device more complicated than a toaster comes with an Android or iPhone app to control it via Wi-Fi. We believe this trend, within a very short space of time, will extend to every light bulb, light switch and power outlet in our homes. Our perspective is that the proliferation of separate apps, each with its own user interface, defeats the very convenience the apps were intended to create. It is true that a person can reconfigure their entertainment system without getting up from their couch, but when you have to call up the AV Amplifier app to adjust the volume and the TV app to change its input and the Blu-ray player apps to start the program, the technology, in our estimation, becomes more trouble than it is worth (after the initial novelty wears off).
We believe that just being able to talk to each of these apps through a single spoken interface will be a great step forward since we intend for Companion to have the knowledge to interpret your spoken intentions and send commands to the appropriate app: “turn up the volume, switch the TV to input 2, and play the Blu-ray disk”.
This level of convenience is only the beginning as we expect that Companion’s support for dynamic state models of complex systems will be employed to control and monitor aggregations of devices as a single system. Our vision is that once you inform Companion how you have your entertainment system components plugged into one another, you can just say “play the Blu-ray” and Companion makes it happen “like magic”.
Our prediction is that the added level of utility Companion would bring to these apps will be such that the vendors can reasonably be expected to fund or at least share the expense of developing the interfaces.
Our plan is for interfaces to third party apps to be sold as premiums to the basic Companion app and along with paid upgrades constitute the primary source of revenue for Companion.
Installing a new Home theater component
Human: Okay. I have a new 4K Ultra Blu-ray player.
Sapiens: That’s great! Do you want help installing it?
Human: Yes please. It is the Panasonic DMP-UB900.
Sapiens: Okay, wait one……got it! (downloads tech specs from Panasonic)
Sapiens: We need three connections, Internet, power and HDMI.
Sapiens: First, the power cord. You need to connect it both into the unit and to the surge protector.
Principal: Got it.
Sapiens: Next the Internet cable from the unit to the network switch.
Sapiens: Now the HDMI, plug it into the HDMI 4 port on the back of the Pioneer VSK-90 (the sapiens has been keeping track of the available ports.)
Sapiens: Put in a disc, please. (the unit turns on and the disc drawer opens)
Principal: Here we go.
Sapiens: Standby while I run diagnostics.
A minute later…
Sapiens: You are good to go, enjoy your movies.
Principal: What about the remote control?
Sapiens: We don’t need it.
We predict that Companion Sapiens has something to offer almost anyone regardless of their age or technological sophistication. We believe that as a classic “disruptive technology,” it is a triple threat.
- As a personal assistant, we believe it will be vastly superior to tools like Microsoft Outlook and other traditional PC based PIMS as well as their mobile competitors like SIRI.
- As a personal Web Presence, we anticipate that it will be vastly more flexible and personal than Facebook.
- As an information retrieval tool, we believe it will be superior to search engines when a person just wants the info and not to “browse” the Internet, thus Google. Apple and Microsoft will be impacted.
As a sophisticated schedule manager, daily routine assistant and personal coach, we believe a person’s Companion Sapiens will be unmatched, exceeding in many ways even the utility of human personal assistants because we predict it will always be there, cannot forget and has no competing needs of its own.
We predict that the Personal Companion has something to offer almost anyone regardless of their age or technological sophistication.