Context and meaning matter now, more than ever
At this year’s World Economic Forum, Ginni Rometty, CEO, IBM sat down with Fareed Zakaria for a fireside chat on IBM Watson’s race against cancer and three points stood out for me.
Firstly, it’s never about just chucking data into a machine. Second, it should not be a black box. Results need to be explainable, credible and real. Hence the idea of confidence levels is critical. Last but not least, while Watson can consume, at last count, 18 million pages of medical content, 200 medical textbooks and 300 million medical journals, it is still the collaboration with the best oncologists and oncology centers in the world to the tune of tens of thousands of hours that actually got the Watson Q & A system off the ground.
It is the art of usability that is tied to context and meaning of the domain.
In our world, the recent debacle with YouTube that saw leading advertisers pull advertising from the platform when their brand sat alongside questionable content, is a clarion call that context and meaning matter now, more than ever, in an age where reliance on algorithms and machine learning will only grow.
Last November, I had the opportunity to have coffee with Professor Alessandra Russo, Professor in Applied Computational Logic at Imperial College London. She has spent 25 years and counting on her passion to “teach” computers context and meaning, as she believes that this is key to how machines can interact with us to create meaningful and productive outcomes.
At the Dentsu Aegis Network Global Data Innovation Centre, data scientists, engineers and technologists motivated by this same passion, work tirelessly to encode the meaning and context of building successful brands into algorithms and software that is now being developed.
For adcelerate, Dentsu Aegis Network’s advanced digital analytics and effectiveness platform, extensive work is ongoing to extract the subtle nuances that would improve digital ROI and more importantly, provide better meaning and context for forward planning.
With Project Infinity, the opportunity is to support strategic planners to operate at a higher level by encoding the wisdom of DAN’s strategic planners, and timeless marketing frameworks, to provide the meaning and context to algorithms being developed to work with the data.
Learn more about how context matters @ Dentsu Aegis Network as we work with algorithms, deep learning models, bots and A.I.
At the end of our coffee chat, Professor Russo left me with a provocative thought. She said,
“Today, machines can identify images in a photograph as well, if not better, than a person.
However, it is still unable to tell the story of the photograph. And that is my dream — I want a machine to, one day, be able to tell the story in the photograph.”
I have to say, I’d like to see that too.
