Thursday, 20 June 2013

Semantics Gives Computing A Whole New Meaning

In my previous article “Rise of Linguistics” I highlighted the increased focus on language technologies by the big IT firms. They predict the future of computing is where humans interact with intelligent machines through the use of natural language. And who would argue with that?  

Linguistics represents an entirely new software category, providing super-intelligent tools to enable a whole new class of applications. It is the door opener to a new era of computing and a money-spinning ticket to the future. This futuristic branch of computing is based on deep scientific principles – from anthropology to neuroscience, and a whole lot more.

Make no mistake; this remarkable engineering achievement will confine modern day computing to the dark ages. Computers will no longer be hamstrung by the restrictions placed upon their own unique, computer “language” – orchestrated exclusively by IT programmers. Instead, human-computer interaction will become the societal-norm, driven by its new interface, the human interface. The very fabric of computing will be changed forever.

The upcoming adoption of Semantics, a sub-discipline of Linguistics, will act as a proof point for this game-changing software category. Semantics, in the context of linguistics, aims to infer the meaning of language. This means that computers will understand human speech and text. This means that machines will be able to respond in kind. This means that Artificial Intelligence is coming…

Context, Context, Context – What Do You Actually Mean?

This means the world will communicate in one universal language thanks to real-time translation capabilities, both in speech and text. The “language barrier” will become all but a distant memory.

This means the internet will evolve from a static, unstructured information repository to a living artificial brain augmenting everyday life – providing you with knowledge and insight whenever you need it most.

This means computers will evolve from “dumb” terminals with one-way communication protocols, to interactive machines with intelligent two-way dialogue. Say hello to your new best friend.

This means everyday appliances will be voice-controlled – buttons, switches, and physical commands will disappear for good. You will tell your appliance what you want it to do. And it will do it.

This means business applications will receive a welcome face-lift as language capabilities provide a new layer of human-like intelligence, the first of its kind. Automation just got automated.

This means social applications will become more instinctive, more interactive, more helpful. More social. Your very own personal assistant awaits.

This means a whole new class of applications will be made possible. Digital forensics.  Counter-terrorism surveillance. Customer (self) service. Expert-as-a-service. Everything-as-a-service…the list is endless.

This means Internet content will be automatically tagged, linked, clustered and categorised – the semantic web will move from concept to reality.

This means marketing will undergo step-change. Customer data will be extracted like a gold mine – empowering highly tailored, highly specific one-to-one communication, operating in real-time. Every customer touch point will be redefined.

This means search engines will be transformed overnight – keyword-based queries will be a thing of the past as meaning-driven search pinpoints your exact intent. Say goodbye to endless pages of irrelevant search results and say hello to precision.

This means multi-lingual search will become the norm. Search queries will crawl data across multiple languages, returning results in the language of your choice, of course. If it exists, it will be found. If it doesn’t, you’ll not miss a thing. 

This means analytics will be taken to a whole new level. Masses of unstructured data will be bought to life – content understanding, fact extraction, evidence gathering, relationship mapping, sentiment analysis, concept search – impossible to achieve through human analysis. Easy to achieve through Artificial Intelligence.

This means, I think I made the point. Welcome to the future.   


  1. This kind of thing is definitely fun to think about, but I don't think that there is any kind of acknowledgement of the amount of research that needs to be happen before we come to these points. To some degree we can even "fake it as we make it" in a lot of these frontiers already, but the focus on automated research methods seems shortsighted to me.

    Then again, I'm pretty much obligated to say that, because I was trained as a Discourse Analyst, a branch of Sociolinguistics that you didn't touch on ;)

    I'm enjoying your blog so far. Keep it up!

  2. good point, there is certainly a lot of research and development that still needs to take place. i've been lucky enough to get some insight into some cool NLP stuff that is not yet available on the market - things are certainly hotting up in this space. in my opinion, the above realisations will be less about the tech, and more about the use case, application layer, design and UI - no easy task!