We hear a great deal about how artificial intelligence (AI) is going to change our future – some say it’s for the better, others are worried about Terminator becoming a reality.
I was called in to help with the deployment of a new data collection and analysis solution for a local business. It’s exciting to see small businesses begin to wake up and realized that investment in data can help them outmaneuver larger, more established competitors.
In the process of drafting their project proposal, I was forced to face the realities of a limited budget. Thankfully, SAP offers a number of solutions catered to smaller businesses – taking a load of advanced analytics off of the company’s labor budget.
We ended up leveraging SAP’s suite of Business Intelligence products. Most of my time was spent teaching the team how to extract data from their existing technology. Then, once curated, we ran through a few exercises – importing excel, google sheets and SQL databases into the cloud-based systems.
Unexpected Sources of Data
The name of the game is data. And most small businesses can’t afford to reinvent the wheel. Instead, small businesses should focus on collecting their existing data and placing it in an easily accessible format. And then, move onto collecting that data that’s just passing through our fingers every day.
One source of data that many companies forget they even have is diagnostic data. Every time a new device is added to the network, or the infrastructure is upgraded, comprehensive tests are run to assure network reliability – many of these are automated, but some organizations deploy analytical hardware to help systems operate at peak efficiency.
But speed tests and packet loss readings are just part of the story. Information about how a network is being used and could be improved is critical data that can help companies expand – whether that means supporting more employees or delivering a smoother digital experience to consumers.
According to Ron Nersesian, CEO of KeySight Technologies, “we are investing to make sure that we continue to be market leaders in high technology products but also we want to make sure that we give very cost-effective solutions for mainstream products. We are constantly looking at the balance between those two investment cycles – investing for the very long-term breakthrough while also providing cost-effective solutions for today’s problems.”
He shared with me that their organization is seeing an increased request for equipment that constantly measures network communication quality and congestion. This is important because it allows companies to understand how their hardware is being used. And AI technology is increasingly being deployed to help with load balancing. For examples of this, you need to look no further than the US power grid. More than 70 million smart meters have been installed in the United States alone.
How much data already exists for AI to leverage?
It’s challenging to find a current statistic for how much digital data exists in the world today. There are just so many sources of data – pictures being taken on smartphones, posts on social media, podcasts, YouTube uploads – the list is endless. A 2014 BBC article estimates that 2.5 exabytes of data were generated every single day in 2012. The ways that we create, transmit and compress data have changed a lot since then.
For efforts to introduce AI into new use-cases, the zettabytes of data that already exist are a welcome site. AI leverages data in order to understand the world within which it operates. Even the most minute detail can be used to give context to the supercomputers of tomorrow.
AI uses data different than a standard analytical platform. Creating a bunch of if/then commands is powerful enough to handle routine tasks. But to understand the subtle nuances of human speech, or drive a car through a busy intersection, you need artificial intelligence. AI takes the data you give it, accepts a mission, and then unleashes itself to achieve that goal – leveraging existing data, gathering new observations and learning from how well its efforts were executed previously.
That’s a lot more than a couple of algorithms combing data. And that’s why even the most minute detail can prove useful. There are many unexpected sources of data that is powering today’s advances in AI.
Data Sources You May Have Forgotten About Include:
Feed these to your AI platform and you’ll discover even more contextualized insights.
Did you know that the government makes some of the data it collects open to the public? The U.S. Small Business Administration created a list of available government data sources to help fledgling startups succeed.
You have access to the web, right?
Beyond the government databases, you can access an endless amount of publicly available data on the web. Every time someone updates their profile or makes a post on social media, you’ve got an opportunity to intercept that data. And then there are the forums, oh those glorious places where people share their most intimate problems and request solutions.
I once worked with a company that built a crawler to pour over every new post in Quora. The data they mined ended up allowing them to refocus an entire product category. Why? They figured out that the majority of posts related to their market involved specific pain points. And they adapted to solve those pain points.
AI can use information from the web to build profiles on potential leads (a.k.a. Lead Generation). And based on what someone shares online, and when they share it, you can time your sales team’s outreach to coincide with their mood. For example, did they just post that they’re back from a trip? You could have your sales team emphasize that they could go on vacation more often and experience less stress at the office by purchasing your solution.
In conclusion, Artificial Intelligence doesn’t require a massive investment in new data. The information is already out there. By compiling easily accessible information in a format that can be utilized by an AI platform, companies can leapfrog the competition with data-driven insights.
And by offloading the heavy-lifting of analysis to a SAS platform, like SAP, companies can reduce their data analysis overhead. The growing popularity of the cloud also means that corporate data can be stored offsite – benefiting from geographic redundancy and encrypted accessibility from anywhere.