The Answer Machine: The Exponential Future of the Search Engine

An emergence of powerful tools is occurring. Machine Learning and Deep Learning algorithms (if you don’t know what they are or the difference between the two here is a good entrance to the rabbit hole The applications of these technologies, which are admittedly in their infancy still, are expansive and fascinating. These technologies are already becoming the best at playing complex games like Jeopardy and Go (a complex Chinese board game similar to Othello). They are also making waves in areas like medical diagnostics. Recently a AI system won a challenge pitting doctors vs. the system diagnosing certain diseases. A human debate team just narrowly defeated an AI system which means that we are a very short time away from an algorithm being better at the nuanced domain of debate than humans. The predictions about when these programs are going to be better than humans range in time but the inevitable truth is that they will outpace us and they will do it soon.

But to what end?

Two opportunities that I see emerging are a new type of search engine, and enhanced cognition.

The new search engine is going to be departure from the current algorithmic process used where you search for something and it gives you a weighted and heavily manipulated response. Deep learning neural network based search engines will learn from you over time, but will also learn on a meta scale which will allow for it to answer questions like “what choice should I make”, “what am I missing here?” and “am I right?”.

If you are trying to decide on a move you will be able to use the search engine to weigh and analyze data on a scale that would not even be possible today. Not only will it be able to factor in social and monetary factors, likelihood of success and happiness, and even use powerful predictive analytics to draw future insight into the decision, but it will also be able to ask you intuitive questions to help you better understand the choices you are making and to consider factors that you didn’t even think to ask about. “Moving to Denver will be an amazing fit for you but you may want to consider the seasonal pollen count”. “But I don’t have allergies”, “based no bio markers that you exhibit you will develop seasonal allergies within 2 seasons of living in this city. They will be treatable with minimal care, I just wanted you to be aware”, “Thank you my super computer friend”, “you welcome my biological broseph”.

Or some similar conversation.

The biggest challenge and opportunity for us will be to ask the best questions. This is also going to be a major shift in the concept of the “search engine”. The answers it will provide us are going to be eclipsed by the benefit we get by being able to ask better questions. Karen Hao recently published an article in the MIT technology review titled Deep leaning could reveal why the world works the way it does in which she explores the progress deep learning is making in drawing insights into correlation. This is significant because it plays into a major shift in the tools we use to explore and understand our world. Progress has historically been the domain of those who ask the best questions and design the best experiments to answer them. Now we have improved our tool set to include systems that allow us to ask better questions. To offer insight and alternate perspectives.

Steve jobs has an amazing quote about hiring employees. “we hire smart people so they can tell us what to do.”  A world is emerging where we are building intelligence that will play this role. It will make us more aware. It will answer questions we didn’t even know to ask.

This becomes even more fascinating when coupled with the realization that these systems are going to be integrated very soon. By integrated I mean that we will no longer need devices to interact with these technologies. This is going to be an amazing tool. Imagine a internal augmented reality interface that simply has a fact checker built in. So in real time as you hear news or conversate with people you will be able to know if the information you are receiving, as well as speaking, is accurate and fact based. This will be one of the most simple functions of an integrated system and would be the most profound improvement in global conversation since language emerged. Just being able to decipher fact from falsehood in real time will be an upgrade to humanity that cannot be exaggerated. Fast forward this to the point where deep learning neural networks are available on demand through an integrated interface and the world becomes a very new place.

How far are we away from a search engine that can take a question, research the entirety of collected knowledge, process the information through deep learning neural networks, and output insights more profound than anything currently imaginable? What will we do with such insight? What will we become?

Leave a Reply