René Magritte, “The Son of a Man”
No user interface.
Base on user’s circumstances.
Intelligent operating system.
Devices that can hear and see you.
Today we live surrounded by software products to a much greater extent than surrounded by people. It is difficult to imagine any activity which is not connected with software at any rate. Even those who don’t use smartphones or laptops also face interfaces against their own will.
Over the past two decades with the effort of designers and developers, interfaces of software products have become more user-friendly, so the users have learned their basic language and now can use the same app on different devices. However, the major issues of software didn’t go away. Every product still makes users perform a certain algorithm of actions to reach their goal. Still software programs are created for the
If you open Google and search for “Java” then links to websites connected with a programming language will occupy most of the first page. You might also see a link to a restaurant with the same name but you probably have nothing to do with the programming language and was looking for coffee or an island in the Pacific. Google showed you the most popular links for the query, relevant to the requests of the majority but not you personally. Google, as well as other software products, does not take into consideration who you are, your current circumstances, and what had happened before you opened Google.
Can a software product be targeted not at the majority of users but for me? Can it take my current circumstances into account at a particular moment in time? Can I use a product that looks different for me rather than for you?
Not so many early adopter companies today are working on the creation of products, which will manage to do it all, new high-quality products, software of the new generation — software based on artificial intelligence technologies.
These kind of products work more like our brain does. We don’t think in algorithms as today’s software demands us to do. It is too energy-consuming, we think in ready-made patterns — get into a car — insert a key, it’s gloomy outside — take an umbrella. Our brain does not have any desire or any capacity to analyse such patterns.
We don’t pay attention to how we make the right decisions when the time is right, how we receive information without asking any questions in our heads. Today’s software makes us formulate queries, execute algorithms step by step.
The next-generation products must provide certain information to fully meet its user’s needs
- information for the specific user,
- information for the particular moment in time, certain circumstances,
- the information in the needed volume at the particular moment.
These principles are the base for the new high-quality software which delivers information not based on use cases and target group analysis but based on the needs of each user individually.
Today we already have quite enough data on every user behaviour in different situations, what they prefer and what they try to avoid. If software products can access this information, analyse it, then their work will become more specific user-oriented and we won’t need to perform a huge number of actions, which we perform today to achieve a result. Another important principle of the new software functioning is access to all user data which already exists or can be collected.
There are a lot of people who are doing the same job as I am, looking for the same information, those who have achieved more in the sphere than me. Access to their knowledge will let new products not only to be based on my personal experience or qualifications but will be able to take all world experience into account when performing any task for which I have “hired” the product. And this data is already there, it is updated and is modified every day. Every day some post in blogs, some upload photos, some host online courses, some read articles which I must read to be successful in what I’m currently working on.
We need to reconsider our attitude to data privacy for the future products. Software must have free access to all data, human gathered experience that should be available to everyone.
Each second information is not only created but is also modified, it is evaluated. The intelligent system must take into consideration a user’s interaction with itself and adjust its behaviour accordingly, that is, continuously studying.
All of the above will be quite difficult to achieve having such discrete level of Internet access. Internet availability depends on the country, mobile operator, Wi-Fi access point, paid bills, too many conditions. For future products, it will be vital to have constant internet access, it must be as easy as a socket access today.
Since the future intelligence systems will need to process huge pools of data, which have all humankind experience, taking into consideration its modifications and results of its work, making it impossible to say that future software is a set of separate products that live their own lives as it is happening now. Any data must have public access including the work results of the intelligence system itself, not only people’s. It is possible to be achieved only if we have a common global platform that will connect all AI products into one single system — an intelligent operating system.
Today we see operating systems as a set of applications that we use to do our “job”. We use applications that we visually like more, which work faster, can do more or are easier to understand, but in the end, its work result is to complete a particular task. When I want to send a text, I want to deliver some information so for me there’s no difference if I use WhatsApp or Facebook Messenger, the result will always be the same. So I just choose an application that I like more.
An important difference between an intelligent operating system and those that we have now is that intelligent OS is condition-based and not app-based as today. Applications how we understand them today will cease to exist and will become a set of skills that software suppliers “teach” the intelligence system, like a teacher at school.
To sum up, future software based on the following principles:
- Providing the required information meeting the user’s needs at a particular point in time;
- complete openness and accessibility of any ever created data from every user, excluding the data which can identify a specific user;
- continuous self-education;
- existence of a single global platform (intelligent operating system), which will connect all products based on AI and will be based on circumstances of each particular user, not on applications;
- universal free access to the Internet with no restrictions.
Products based on artificial intelligence are about a different way of user-device interaction. They demand a high-quality new interface, unlike the one we use today.
Interface is always a change indicator. If any drastic improvement is made in technology, then firstly, it affects the visual side of it.
The last global breakthrough was 36 years ago when Apple introduced Lisa with a graphic interface. It was something monumental in comparison to other computers at that time which interfaces looked more or less like this:
But 36 years passed and today we have quite decent artificial intelligence works. Our mobile phones know our fingerprints, they can execute voice commands and recognise a person looking at them. And today our interfaces look like this:
But 36 years ago the first Lisa interface looked like this:
So nearly nothing has changed at all, apart from the visual part and several functions performed.
A new type of software demands a new approach in the interface. Our current interfaces are morally outdated and that their dullness limits the development of technology is the worst. We need a pathbreaking approach in interface architecture, which won’t make us study it but work for us.
In this particular case, the idea is simple — there will be no interface. I mean traditional interface concept which we are used to — windows, icons, controls, menus, forms, tabs, etc. All this junk will stay only in a computer science museum and in retro films. Content will be the only thing left for us, the reason why we use our devices, applications, and websites for.
There are 3 global areas where we use our gadgets:
- receiving information
- creating information
If the system is fully based on the principles from Part 1, then we don’t need an interface in today’s understanding to receive information. In this case, content becomes interface and occupies the whole screen.
Knowing about the circumstances about me and what I prefer, the system will provide me with the content at the right time and right form when I need it. The need for a section like “Settings” disappears automatically. We will no longer need to register anywhere, our information will be in free access anyway. The device sees that I am in front of it and, it is a considerably more secure way of authentication.
Devices will be able to see and hear, distinguish what’s happening around them, they will hear what is said, what is happening, consider and analyse the information, that is, continually study. Having “eyes” and “ears” the system will be able to accurately respond to the changes in circumstances and deliver relevant content by minute. If me and friend of mine are discussing that we are hungry and it would be nice to order a pizza, and then I take my phone and see a pre-order on the screen. It is my favourite pizza and I’m likely going to order it, and its price will be taken into consideration according to my credit card balance. The pizza will be delivered from a reliable place approved by other users and the delivery guy will already have my current address and I will only need to accept the offer. I’m sure it’s going to be the best offer and, I won’t even want to double-check it.
If the device can see, then it will easily distinguish a person in front of it. So when I take my friend’s phone, I will see the same things that I would see on my phone if I took it instead. And such a thing will happen with every device, everywhere. What I won’t be able to see is my friend’s content.
This principle must work on every device. If I stand in front of a display at the airport, the system will distinguish me and will show information about my flight, where I should go. At McDonald’s, the display will automatically make my order and I won’t have to take my credit card out because the system already knows its information. Same as in the situation with the pizza I will only have to agree.
Our whole device usage experience is dictated by the goals we have: we look for flowers to please a person we love (short term goal), we write a report to our boss to continue getting our salary and live with comfort (everyday goal), we are studying a new discipline to start our own business and change our lives (global aim). The problem is that by “users”, use cases of whom designers consider, today we understand the target audience of a business or the majority. And today’s product is designed in a way to satisfy the desires of the majority of profitable users, not individuals.
The main task of the future intelligence systems is to understand user goals and aims at every particular moment. And here by a user, we mean a particular person who is interacting with the interface.
If I have just arrived in a new city and went out of the airport, then my goal is to get to the booked hotel, which complies with my habits and my bank account balance. If I went to bed late in the evening and took my phone, what I regularly do, then it is the right time to offer me something not connected with my everyday tasks but connected with my global aim (e.g. find something useful about artificial intelligence, taking into consideration the time of my usual falling asleep). And this is the moment when global access to data about every person and his activities work for the progress of mankind. Because there are people with similar aims to mine and they have already achieved more in the area than I did, so I can use their experience and knowledge.
Mobile phones’ popularity has outrun personal computers and even laptops. You might not have a laptop but you have a phone whose main task is communication. Phone, of course, was initially designed as a communication device but with smartphone advent, the word was reconsidered. Today, communication is less about phone calls and more about text and graphic communication. Communication is a separate type of user-software interaction. It can neither be labeled consumption nor information creation, it has signs of both of them.
We communicate to receive and send information.
Today, despite the huge number of messengers, we still have unresolved major problems in this type of interaction:
- inability to find the right time for communication;
- inability to reproduce the real mood of the message;
- lack of instant feedback from the recipient (if understood or didn’t understand, is embarrassed, angry, surprised, etc);
- language barrier.
Some of these problems are easily solved by verbal communication (audio/video) but the majority of problems remain unsolved.
On the one hand, having access to user information, it’s not so difficult to notify me that my message is not likely to be read any time soon because the recipient is currently on the plane. On the other hand, if I’m attending a very important meeting, I do not want to be distracted by my device, showing a text from my mom.
Language translation is another field of work for the intelligent OS. We already have enough information about languages, their pronunciation, and constructions. Language is a dynamic system and, if we have global access to all the data, changes in languages can be noticed instantly, so we won’t have to wait for linguistic institutions to make a decision. The profession of a translator or an interpreter is just another atavism which shall bite the dust in history books or an article like “50 extinct professions”. A Greenland Eskimo and a Carribean fisherman can talk to each other only by having a device, which will do the job of an interpreter, at hand.
If the intellectual system knows our habits well, listens to and studies what we say, knows our circumstances and plans, then it can notify people we communicate with without our participation. Just imagine yourself flying to visit a friend in another country, you’ve already got on a plane and has turned all your devices off when your friend suddenly realises he doesn’t even know your flight number so he sends you a voice message which you will be able to answer only in a few hours. But your OS, having this information, can answer your friend and even have some small talk with him, using a pattern of how you usually speak. And this is a completely new approach in communication, it can be conducted without the direct involvement of the user. Worried parents can hear your voice or receive a text message answering their “Did you get safe?” without waiting for you to finish an important meeting. A customer will be able to receive information which the supplier has forgotten to provide even when the supplier can’t reply for whatever the reason. In theory, it might give us possibility to communicate with those who are no longer here. And this will become the most lifelike realisation of the idea that a human being is what they have created, said, learned and decided because the information remains forever and can’t be changed.
Our current work experience with software fully depends on what applications we install on our device. Applications give extra functionality, in addition to the basic within the operating system. The more I install, the more my device can do. This approach goes absolutely against the fast development of new generation technologies, for which it is vital to receive information about the user’s aspects of life in full. Without it, the idea of access to any kind of information is beyond attainment.
This is the reason why the future system shall be global as the Internet. For the system, devices will only be entry points, tools for getting information, kind of “portals”. Every user will be able to use any device to acquire any information they need, regardless of whether the device belongs to them or not. The device will identify the user and will show them what they need at that particular moment, however, it will not let them acquire any information that is not intended for them. At first, it might sound dangerous but actually, it is the most effective and logical private data protection ever — your private information will be available only to the physical you and no-one else whoever managed to get to your device or even your password, which happens quite frequently today.
This approach abolishes the understanding of a device as localised closed information storage. Information from all the users must be available in full. It doesn’t mean that anyone can get access to everything that other users have written or taken pictures of. All of it must be anonymised, deprived of any opportunity to personify the author, except when the author posts something under his name. That is to say that in comparison to today’s systems, which protect access to information, new systems will protect the identity of the information owner.
Indeed, does it matter who created this piece of knowledge if I received what I needed? I can get valuable advice before getting on a trip to the South Pole from someone who has already been there. Or advice which other users considered useful, not even knowing it was a 78-year old lady from Australia.
The new generation of technologies and interfaces inevitably creates a new type of developers, designers, and other IT specialists. The disappearance of old professions and the creation of new ones is inevitable. For example, when graphic interface emerged, professions like interface graphic designer, front-end developer, back-end developer appeared. Product manager, customer support specialist and many other professions became very popular with the development of online services.
The new generation of interfaces ruins the understanding of interface design as we have understood it for the past 30 years. This profession will no longer be needed because the traditional concept of the interface will no longer exist. We will no longer need to analyse use cases, create convenient registration forms, identify essential and minor controls, etc. Instead of this, designers of the new generation will have to focus on content design (which is largely neglected today, I must say, and interface processes are currently given the priority). I mean that the role of the future software designer will be, as it’s clearly understood by now, to deliver the content concept to the user simply and sufficiently.
Today it is largely left to the mercy of so-called content managers because not so many designers care about photo quality, infographic accessibility, text clarity and correct emphasis in it.
Designers take off this responsibility by their ingenuity to hide technology imperfections and maintain the interface, which will lead to a great improvement in content, which is the final aim of every user.
However, the idea of total absence of interface puts great responsibility on developers. The future of the new technology highly depends on them now, they have the major role in creating a new reality of software. Today developers are mainly engaged in creating applications. Users install applications to add a new “skill” to their device, they kind of hire this application to do a particular job. As a result, the user’s ability to get one or another result from their device depends only on how well they prepared it, stuffing with applications, setting them up correctly and going through all the humiliating stages of registration.
But the point is that we don’t need applications to interact with them, we need them to execute specific tasks. Today we chose an application from a great variety of those, which perform the same functionality, we choose the ones which we like visually, which have a simpler interface, have more functions or just differ from the rest. However, the idea stays the same, we use applications for the same tasks, which in the new generation systems will be executed without traditional interface thus making the design no more than just a gold plating.
The work of an artificial intelligence developer will be to “teach” the system new skills. The work will be not to create algorithms but to “train” the system, to make it acquire new skills as if it were a human being. Basic functionality has to be incorporated into the system in the first place — identifying what is important for a particular user at a particular point in time, studying their behaviour patterns, defining their goals and aims. Several other skills will have to be integrated and that is what the new generation developers will do. Instead of creating Vivino application developers will teach the system to understand the wine preferences of each user and give them advice from important users upon a bottle of wine, to the label of which they bring the camera of their device. Messenger developers will teach the system to understand when and which message to show and when to notify my talk partner that I’m currently busy and can’t talk.
Developers of the new generation will slightly resemble today’s ones — they will create much fewer algorithms and deal more with collecting, analysing and structuring data, which will be fed to the system, determining the standard behaviour of the system, and optimising its learning process. The role of data and knowledge engineers will grow; their objective will be knowledge collecting and structuring, as well as controlling its validation. We can say that these two professions are going to assume major roles in the new generation business and be the most wanted. New professions will appear, but now it’s not so easy to define them. However, we will need someone like a “mental physician” who will deal with causes of the situations when the system would fail to make a decision or a specific conclusion in those situations when a person can easily comprehend. A list of new professions must be much longer than we can imagine now.
New generation software is approaching day by day. Today we are not even at the beginning of this path, we have come a decent way and we are somewhere around the moment of execution.
However, the execution demands particular reforms in software, revolutionary in some cases, and perhaps it might sound crazy for today’s user but that is what the new generation is about. It is something we have never faced before. If one single company created the second generation of interfaces, then the third generation being more global will require the effort of not even one or ten companies, but it is about a consolidated effort of thousands and dozens of companies altogether. Many areas in IT from legislation to device production will need to be reformed.
The campaign roadmap for most significant changes is as follows:
- User data worldwide collection. This stage has started a long time ago and currently is developing in a variety of modern products. The future of software products is collection of user data, their behaviour, knowledge, behaviour patterns, and preferences. Data will become fuel, the future “black gold” for software products. It will be impossible to create a new type of product without data from a sufficient number of users, thus making this resource commercially viable. Major companies like Google or Facebook are actively accumulating data almost in every aspect of a modern person’s life, which will let them become leaders in the development of a global intellectual system in the future.
- Data privacy revision. Today’s understanding of private information is nearly the only obstacle to software development, like censorship in literature. Now everything that a user did not allow direct access to other users is considered to be private. The concept of privacy should be changed, it should protect the user’s identity based on his data and not protect the access to the data itself.
Providing widespread access to the Internet. This process has also been launched a long time ago and is developing very successfully.
One way or another, we will end up thinking that the absence of the Internet in any place on the planet is as wild as the absence of electricity today.
Future operating system base functionality design.
This step is most likely intended for those, out of thousands of companies working in AI, who will manage to, following the example of the first UNIX team, develop the core of the future operating system. The core which other companies will develop bringing in their experience until it becomes ready to use product. This is a defining task, which, unfortunately, is still not even at the stage of awareness of its necessity.
- Device improvement. Devices must be able to hear and see what is around them. It seems that since the first computer, now we are in the situation when device development outruns the development of software. Today’s devices already perform these functions quite well, however, in the future, they are expected to perform even more.
- Transition to a new generation interface. It is important to understand that first of all, the improvement of technology should be aimed at changing every user’s life. There’s no point in wasting so much energy to create new types of information that would still be not easy to reach without going through a set of algorithms and not understanding the patterns of a particular operating system. The future of software must be focused on every user, as it has never been before.
A few years ago “AlphaGo” movie was shot. It tells us about a program created by DeepMind company. It was a program that played Go and which outgaimed the world champion. While playing a game the program made move #37 after which all the Go experts in attendance clutched their heads being sure that you could never make such a move, it was totally stupid and disastrous, and that the game was lost. However, that exact move let the player win, making a turning point at the tournament, because a human being, the world champion could not expect such a “lack of logic”. Some say that this lack of logic is ingenuity. In my opinion, move #37 shows the main object of future software — look for subtle patterns. An intelligence system managed to do it. It put emotional, traditional, cultural barriers aside, those things that we — humans can’t do as a rule.
November 9, 2019. Oslo, Norway.