Why Data is The Most Valuable Asset of The Century
by Amit Jnagal, on October 18, 2017 11:15:00 AM PDT
Imagine you are Danny Ocean planning your next heist with Ocean 11's team. You have figured out a way to take over a building and hold it for ransom. You have narrowed down your search for an ideal venue for the heist – DreamVille. The sleepy but wealthy town of DreamVille houses headquarters of a large national bank, a few casinos, a bullion retailer’s secured warehouse and a data center of a large company.
Which building do you pick to hold for ransom?
If you chose the casino or the gold warehouse, you should think again. The most valuable asset of a city is its data center. More valuable than a bank and far more precious than bars of gold – data centers house invaluable assets for businesses.
As more and more companies start embracing AI and Machine Learning solutions to automate stuff, it gets difficult to understand what will differentiate them from each other? If all algorithms are built using similar technologies and run on similar cloud infrastructure then whose ‘intelligence’ will be superior to others? Will Tesla’s Autopilot, Google’s self-driving car & Uber’s autonomous vehicles all be of same quality and predictability?
Having dealt with numerous customers in the context of AI and Machine Learning, it has become quite apparent to me that quality or accuracy of a company’s AI or Machine Learning algorithms is largely a factor of the data they have to train and learn. The same algorithm can produce an average quality of prediction or accuracy when trained with different data sets.
It’s not the technology but the data that is the new IP.
Google figured out before most that data is going to be most valuable asset of the century. They give away high-quality services like emails, messaging and social media for free in exchange for data. Now I am not saying that their technology is trivial – on the contrary, it is quite superior. But, with all due respect, it is not impossible for others to build it. What others do not have is access to their data. That makes them better than most.
In fact, a few years ago Microsoft launched their search engine – Bing in order to grab a market share of the online ad market. I would wager that Microsoft’s technology was not dramatically inferior to Google’s. But they did not have access to Google’s data to train their algorithms to improve relevance. Allegedly, they resorted to ‘stealing’ Google’s data in order to bump up their search relevance.
Over the next few years, several AI or Data Driven Intelligence companies will start offering solutions. A few of them will have extremely unique algorithms and approaches but what will differentiate most of them will be the availability of data for extracting intelligence.