Toward the "Answer Engine" - 2 of 2

 

Part 2 - The Peril

Also published on the Atlantic Council's FutureSource Blog

In my previous blog, I outlined what living in an Answer Engine-powered Internet might look like. There is much to be excited for as we move toward the future where the Internet is woven into every facet of life. Think how far we’ve come already and how quickly it happened. Think about your relationship with your mobile device. This merging of Internet and life is inevitable; in fact it is already almost a fait accompli. There is also an array of dangers to be concerned about. The most worrisome negatives must be understood and addressed if we are to have any hope of mitigating them. 

In the future Internet, privacy as we have known it is gone. The global mesh of cameras and sensors combined with a powerful Answer Engine AI will ensure that the most information that can be known will be known, although a secrecy arms race will likely still be boiling under the surface. Every human and every artifact he or she owns can be tracked and accounted for in real time. 

Yet I do not think we are headed for an uncontrollable transparent society. Instead I foresee social contracts of the future that are immensely more nuanced and codified than they are today, especially at the individual level.

Lovers might grant each other full privacy permissions, while new acquaintances will be insulated from each other with myriad permissions doors. Every piece of your networked experience will be sharable, but you choose who is allowed to share it, for how long, and even at what price. Access to cameras, health data, location, even inferred mood will be available to share with others. Everything you do and say, but also increasingly, everything you think or even feel will be accessible to anyone with the key to these doors, or a good crowbar.  Perhaps governments won’t even need a key; they may weasel their way in at a design level, requiring guaranteed access as the cost of selling computing products within their borders. Every user’s most intimate data will potentially be accessible to every other user - but the individual will control access to much of this information. 

As well, the Answer Engine paradigm will mostly forgo providing huge swaths of broad information, instead providing small, calculated snippets of densely relevant information. As I forecast in my previous post, such a capability will be a boon for inquiry. However, it also presents a dark problem. Even modern Google responds to queries by placing advertisement links at the top and links prioritized by more dispassionate calculated relevance below. While the ad link might be a fine answer, it is also a terrible answer; it got the top spot not from pure relevance but by buying its way there.

 
 

The Answer Engine world could see people ferried about their daily lives not by the best answers, but by the answers that paid the most to be delivered. As we continue to offload internal memory to our networked world so too will we offload some amount of inference and even decision-making. Existing in and relying on an Answer Engine Internet opens your mind up to a powerful new form of direct manipulation: the answers to your questions delivered by the top bidder, the solutions to your life’s daily problems offered by the entity with the biggest wallet. Who is in control then? 

With huge swaths of society being literally controlled by advertisers, the potential for dystopia is alarming. More worrying still, there are efforts currently underway by Facebook and others to gather data and synthesize inference about people so nuanced that it begins to form a picture of who they actually are, not just what they like. Advertisements taking the form of help, of solutions. Ads tricking people into thinking they are the best answer by being a great answer. Not actually the best answer mind you, merely a great answer, one that makes you spend money and feel happy.

Consider this hypothetical: The Answer Engine knows you’re feeling sad based off body sensor data, the wistful social posts you’ve been making, and the fact that you just got a digi-paper from your boss telling you that you’re fired. It also knows you like sweets. All on its own and at a perfectly calculated moment, it proposes that you stop worrying about your future and buy a treat for yourself from Cand-Corp instead. Wow, thanks Answer Engine, you always know how to cheer me up! Given this level of context and insight, ads might move from annoyance to friend. But of course, this friend only wants one thing from you, and it will use every trick imaginable to get it, all while giving you vapid nothings dressed up as wisdom in return.

Hopefully, the future will be characterized by seeking balance, by being aware of what one might gain by accepting a technology and what one might lose. Understanding the forces at play will be instrumental to retaining power and agency. To accept blindly is to be powerless. The majority of Earth’s population could end up as mere cyber-serfs who have no knowledge of how the technologies that define their existence function and thus are basically at the mercy of whatever entities they have trusted with their data - and their lives. They could become "ignorant masses" trapped inside the information pool, being ferried about by a handful of mega-corps and savvy individuals, blindly following while given a shallow illusion of control. Yet to swim completely upstream and escape the pool will become increasingly difficult. Some trends are unavoidable, at least not without forgoing integration with the rest of society. There will simply be fewer and fewer places left to hide. Joining this potentially dystopian future could become less of a choice and more of a silent oppression; either join or be left behind.

Billions more humans are still waiting to be connected to the Internet. These next decades will see the completion of the transition from the Internet as a tool, to the Internet as part of existence itself. This transition will be rocky, it will change who we are, and it will open up cans of worms previously unimaginable. Many already use the Internet for the most nefarious of ends, and they have abilities at their disposal that were once the sole purview of only the world’s most powerful governments. 

Still, regardless of risks and perils, this is what we humans do. We’ve been absorbing our technologies into our lives since before recorded history. Now we cook food, we wear clothes, we live in structures, we use electricity, and we are connected to the Internet. This network of networks is a basic utility now, something that all are entitled to. It is driving the biggest change that humanity has ever undergone, a growth in knowledge and power of truly exponential proportions. Yet knowledge and power are just tools to forward ambition. Good and evil both will be empowered. The opportunities for failure are high, but then so too are the opportunities for success.

Personally, I am optimistic. Being aware of potential future pitfalls helps us avoid them, or at least know where to look and to try to mitigate their impact. Now is not the time for pessimism in the face of massive change.  Rather, it is the time to find the balance, to assert control, to be proactive in shaping our future. I believe optimism and pessimism alike create self-fulfilling prophecies, and I prefer to pick the better prophecy. 

- JH

 
 

Toward the "Answer Engine" - 1 of 2

 

Part 1 - The Promise

Also published on Atlantic Council's FutureSource Blog

 

The Internet is moving beyond a portal to another world and becoming part of the very world itself, meshing with and augmenting physical reality. The idea of the “Alive Web” sees the Internet becoming an increasingly real-time affair.  We are in the beginning. Apps like Snapchat and Twitter already represent the new pace and face of the new web: it’s nearing real-time, and it involves us directly. No longer about stale ‘feeds’ and static data, but experiences and emotions in the present. Less about user names and pseudonyms and more about identity strewn across the world, ready to interact with anything. Not snippets of text and culled media, but a nuanced and raucous interactive conversation happening across the globe. Forget myriad disparate websites and links; imagine an ever-growing pool of everything being fed by a never-ending deluge of data. This is the Internet of tomorrow. 

Currently, when you have a question that cannot be answered within your brain, you must take it upon yourself to seek the answer. This has become so ubiquitous and easy that it merely takes a few taps on a device in your pocket and dozens of answers appear before you. Yet the burden is still on the user, to format the question within guidelines, to cull the answer from myriad sources and arrive at a conclusion. It is still very much a process of actively seeking, which often constitutes a break in flow, if only just a brief one. 

Imagine if you didn’t have to seek. Imagine if you didn’t have to cull. Imagine if the answer to your question was delivered right as you desired it, maybe even before. No break in flow, just ubiquitous information when you need it.

The future will be powerfully defined by the emergence of servile computing, enabled by powerful contextual information. No more asking for help, future computing will simply provide it -from sensors on the device itself, sensors on your body reporting your bodily state, and the coming wealth of ambient information available from the “Internet of Things” along with, of course, the huge amounts of data exhaust we all leave behind in living our cyber lives. Combined with clever and powerful processing, these huge swaths of data can transcend noise and become salient real-time answers. The push toward ubiquitous mobile computing all but demands the emergence of the next generation of search engines, or more accurately, Answer Engines. No more clunky queries, no more digging through links, no more breaking flow, just the answer you want when you need it.

I use the term “Answer Engine” as a catch-all phrase for powerful artificial semi-intelligences—think IBM’s Watson Jeopardy champion— fed by immense sources of data from individuals and environment, coupled with the fluidity of interaction that will characterize the coming wearable modalities of computing. Already the search engine-powered Internet has become part of working cognition, and the Answer Engine-powered Internet likely will see this role increase to the limits of our current imagination.

Interacting with the Answer Engine Internet will be characterized primarily by very little actual interaction. The Answer Engine -- if you choose to enable its features -- can listen to your conversations and do searches on key terms or concepts automatically. The Answer Engine hears you humming along to some tune and can find it for you for later listening. The Answer Engine knows where you are and knows what you’re interested in. It can make suggestions about where you might like to be, or where your schedule says you’re supposed to be. The Answer Engine might even have access to where your gaze is pointed and can infer interest, querying relevant details automatically based on objects, people, locations – anything it thinks you’re interested in. It might even have access to detailed information about your bodily state and help guide you in regulating your health. If you allow it, it might even have access to do this via small wearable medication injection systems or more exotic Nano-machines, biological or otherwise. This is no longer merely about search and ‘social’ applications; this is an extension of self, of mind.

Aside from answers to daily problems and questions, the Answer Engine’s powers will scale exponentially for providing more complex answers. It will have increasing access to networked artifacts and sensors via the “Internet of Things.” The Answer Engine can be used to provide real-time answers to queries about the current state of almost the entire Earth, scaling elegantly from a single room to the entire planet and beyond. It will also have access to networks of connected cameras, terrestrial, and orbital. On a lark, you could have it deliver answers to queries that are almost impossible to address today.

For example, combining its access to sensors, location data for every connected human and artifact, computer vision, and the global mesh of cameras, you could ask it–in natural language–something like, “when it next rains in Paris, count the number of yellow umbrellas seen from beginning of rainfall out 5 minutes; do this for 5 rain days and report the average number of Parisians with yellow umbrellas for this period. Correlate observed location of umbrella with individuals’ locations and ignore duplicates. Report margin of uncertainty.” Useful? Perhaps not. Representative of statistical inference abilities unheard of at any cost today? Very much yes. Of course the data that a badly worded query like this would generate would be full of problems, but think what it represents. Imagine what kinds of questions you might like to ask it, and imagine if you could ask as many as you would like; think of the questions such answers would lead to!

The Answer Engine is about smashing together data from any and all sources available, and future sources of data are set to increase exponentially. Its powers of inference will only grow, and every question it answers or query it receives will be catalogued and recombined to provide better answers in the future, to hone its intelligence further.

While the Answer Engine Internet itself is a powerful amalgamation of artificial intelligence (AI), it also knows humans who might have an answer for you. If you require complex assistance beyond its scope, it might facilitate a connection to someone who can help. The Internet of the future knows where everyone is and can even coordinate them in "meat-space" if so desired. This is the longer-term future of the Answer Engine Internet, presenting all users with a seamless river of relevant information built from unifying human and software abilities and knowledge into one useful whole. 

For example, a little marker might appear in your field of view pointing you to a human near you who shares your interest on a topic or can help you with a 3D-space project. Or you could instantly give an expert access to your current view and have her help you through the problem, likely for a small fee. If today’s networked youth are any indication, many users will opt-in to this system. Providing a way to seamlessly engage with new people anywhere on Earth, requiring only the desire to do so; perhaps only coming together to solve a problem or provide an answer, never to see each other again. As the Internet and physical reality merge, we’ll see online paradigms and modalities of community and interaction rub off on the physical world and vice versa.

Collaboration as fluid and dynamic as meat-space interaction, but no longer limited by physical barriers or confines. The online/offline distinction is dying and in its place a new kind of hybrid reality is emerging. No longer just running parallel to the physical world, but intricately woven into it; responsive and aware, but also affective. The Internet of the future is ubiquitous. The fluidity of querying your own mind applied to the global mind. Software and human voices aggregated together into a seamless whole. As search engines enabled the indexing of data, Answer Engines will enable us to do something with it.

Continue to Part 2: The Peril