Climbing the AI Ladder for Organizations
Today we are pleased to welcome Paul Zikopoulos to the show. Paul is VP of Technology Group at IBM, author of twenty-one books on database analysis and cloud. In 2019, Paul was named to the list of “Top-100 Influencers in AI and Big Data”. Paul has been honing his craft for most of his life and in doing so has been able to master many powerful tools to help enterprises drive the future powered by technologies.
In this interview episode, we talked about the future of AI, the most impactful OSS, and the shift in companies culture across many industries due to the COVID-19 pandemic.
Watch & Listen to the full interview episode.
Paul, it is such a privilege to speak with you today. Can you talk about your professional journey leading up to your current position at IBM?
I came to IBM 29 years ago with an undergrad in economics in business. I got a job in the ID department to be a technical writer, but I wasn’t even a good writer. I just needed a job. From there, I got to a lab environment where you work hard and learn all the time. So, I learned user-centred design, programming, Management Development, marketing, and competitive sales skills. I’ve done all this without having a background in any of those domains.
Thus, I believe the secret of any success has to do with two things. One – learning never ends. You have to learn because things are constantly changing. And two, have grit. You may not be the best at anything, but always try your hardest to be the best.
Paul Zikopoulos
VP of IBM Technology Group
Having so many things on your plate, how do you manage the risk of burnout?
When you are getting close to burning out, take an honest look at what you want out of your career, then write down five questions and prioritize them. Do you want to make the most money that you can possible? Do you want a fancy title like a VP or General Manager? Do you want to love what you do? Do you want career progression or personal growth? And do you want a work-life balance?
There should be no shame in the order they’re in because those will change over the course of your life. And remember, no one loves every moment of their job; you have to hate 20% of it. But if you have passion for learning, that might push off the burnout and give you wind in your back to achievement. That’s how I avoid burnout.
Being at IBM for that many years is like being in a successful marriage – it’s not always going to be perfect. You’ll always have ebbs and flows.
Paul, you have written 21 books on technology. How do you find time to write so many books?
My goal is to write one page of really bad writing every day for a book. If I write one to two pages, then I’m on my way to get my thoughts out. That’s number one. Number two is I love what I do. And I’ve always found that when you teach, two people learn. So when I wrote the book “Hadoop For Dummies”, I knew very little about Hadoop. But I knew how to learn.
Can you share with us what your two latest books “The AI Ladder” and “Cloud Without Compromise” are about?
The “AI Ladder” is about four rungs called progressions that organizations climb to get to AI. Those ranks are collecting data, organizing data, analyzing it, and then infusing AI across the company. That is a strategy of getting to successful AI because now 80% of all AI projects fail.
“Cloud Without Compromise” is about to be published. It is not a practitioner book. It’s a business book for people who need to get savvy around the cloud, who lead businesses and who need to understand the technology.
Many modern organizations are data-rich but information-poor, meaning they know how to collect data but are missing the data interpretation part. What do you think about the problem?
You hit right on the problem for organizations where we’re good at data collection. If our data collection is good, but data understanding is poor, then in between, there is what we call the price of not knowing. And that is where organizations can achieve better health outcomes, better profits, lower attrition.
Where do you see the future of AI?
I think there’s a corner in AI that nobody’s looking around. First, AI was about how do we build models quickly. We used hardware acceleration, like GPUs, and TPUs to build and score the models. Then we started building models, though a lot of those models didn’t get to production.
The next focus on AI won’t be about AI’s performance because now performance doesn’t mean how fast; it means how accurate. The key to AI is going to be around explainability – How did we arrive at this particular decision? Where did this data come from? Because as we can see, we codify bias in AI. That might be acceptable to have gender-biased data if I’m selling men’s golf clubs. But that’s totally unacceptable if I’m issuing credits.
Companies can reduce the risk of using biased data sets by good governance. Though, the problem with governance is that now most companies do it as a least effort to comply approach – I don’t want to get a GDPR fine or violate a HIPAA regulation. By investing in governance as an analytics accelerator, you will know where the data came from, you’ll identify potential sources of bias, set rules for eliminating bias and identify accurate data.
IBM is a great proponent and contributor to open-source software. In your opinion, what’s the most important and impactful open source software?
I do believe Kubernetes is one of the most significant inflection points of open source that I’ve seen in my entire career. It looked like Hadoop was going to be the biggest, but it crashed.
Open source is a wonderful opportunity for all environments because many people with great innovation ideas commit to the community. Though, sometimes I compare it to a free puppy – I can give you a free database, but you’re probably going to need support. Thus, I don’t recommend running the open-source route on your own because you’re always better off with a partner, no matter what open source technology you’re using. Even at IBM, we embed open source in all kinds of our technologies, from our explainability OpenScale to our data transformation tooling that uses Spark.
How has the pandemic affected corporations?
I wrote an article when the pandemic first came out, called “Thrivers, Divers and The New Arrivers – Business in the COVID Era” and explained that the pandemic was a massive wake up call for business. As it turned out, businesses weren’t nearly as digitally transformed as they thought. Now there’s the realization that our journey to digitization is aged by a decade, and it’s not enough to talk about it; you’ve got to do it.
Paul, thank you for your vision and commitment to building the technology that will transform the way we work and live!
Stay tuned for more great interviews coming your way!