When I say IBM, you say … cutting edge machine learning? When they sponsored my trip to the 2016 World of Watson conference as a perk of my involvement in the IBM Predictioneers Social Influencer Program, I wasn’t totally sure what to expect, but I was pleasantly surprised with how advanced Watson and Watson Analytics have become, and the data accessibility that they’re focusing on across many different product lines. As it turns out, IBM is a contender for future analytics implementations that I will do. Here are my 4 key takeaways from the conference.
The most powerful feature for moderately technical to business side users is the ability to ask – literally in text – “what factors contribute to sales” and the combination of Watson’s natural language processing, and its Watson Analytics data engine will return a chart for you with the primary drivers of sales from all of the data sets that you have access to. Since IBM is still a large-enterprise focus, they have relatively sophisticated data governance capabilities, but for a less strict environment (like most of mine), you can likely surf through your entire corporate data set with relative ease.
There’s even a free trial and a relatively inexpensive option for a single user to experiment with. Not something that I expected from IBM, but frankly, I was excited to be able to work with a new tool like this. I’ll be interested to see how I might work it into my own workflows over the next year. There’s a ton of technical capability there, and Watson Analytics does a relatively good job automatically choosing visualizations for different data types, while allowing more knowledgeable users to customize as needed.
The most noteworthy announcement from the conference for me was the introduction of the Watson Data Platform. This is an AI powered tool that can ingest data at 100gb/second and help you analyze it alone, or collaboratively. Need to feed insight back to your developers? You can do that. It’s theoretically only several clicks from ingesting data, running some machine learning algorithms, and sharing those results with your team. I haven’t worked with it yet, but needless to say, if IBM can deliver on that promise, there’s nothing else currently out there that can compete.
Check out the explorer video:
Or hop over to fellow Analytics Expert Jen Underwood’s first take on the Data Platform.
Most people are familiar with IBM Watson from the amazing feat on Jeopardy several years ago. By beating Ken Jennings at the game he had mastered, Watson showed that language processing was becoming strong enough to rely on. The really incredible part was that Watson could recognize a play on words. While this feat, in particular, isn’t all that important for a business user, its implications are. Allowing the functionality that I mentioned in point 1, watson can help your business user get access to the right data whether it’s labeled sales, revenue, order_price or something else.
It’s clear that IBM is trying hard to encourage folks to consider cognitive machines as a complement to human intelligence and work, but even with that effort, I’m not convinced. We saw several demos where the combination of new technology (IoT, drone cameras, etc) will simply replace humans where applicable: faster, cheaper, better. Need photos and damage assessment from a hailstorm? Queue up that drone, feed the images to Watson, be done in hours, not days or weeks when an inspector is available. Have Cancer? When Watson can ingest every medical paper ever published and can 99% match, and 30% improve upon the diagnoses of a trained cancer review board, why wouldn’t I just go straight to the computer? Need a tutor for the SAT? With Watson, a personal digital tutor is available to everyone, not just those who can afford $100/hour personal sessions. When we make a mistake driving, we learn from it. When a computer AI makes a mistake driving, every car running that software learns from it. After a few generations, why would I every trust a fallible human driver again?
While IBM is making a concerted effort to push Cognitive computing as human assistance, I sure see obvious human replacement opportunities. Be sure that as you develop your career, you don’t become stagnant. Lifelong learning will help you maintain opportunities to grow into areas that still need human creativity and critical thinking while offloading repetitive tasks, even complicated ones, to cognitive machines!