Digital transformation is based on digital innovation, and digital innovation stems from experimentation at scale — a challenge that has eluded all but the biggest companies.
Know My Company
Tell us about your interaction with smart technologies such as AI and cloud-based analytics platforms.
Our company, Sentient, is at the forefront of applying Artificial Intelligence to optimize digital customer experiences to accelerate revenue growth, and is a leader in the category of AI known as evolutionary computation.
How did you start in this space? What galvanized you to join Sentient?
My career interests were always about being at the forefront of the intersection of technology and marketing, from my early days as publisher of the first CD-ROM-focused magazine and founding CEO of GameSpot, the first and still the largest Internet media property about video games. After stints in gesture recognition and marine technology, I spent a few years as an executive recruiter.
Sentient was a client, and I was immediately drawn by their unique approach to AI, the brilliance of the team…and the clear recognition that AI will be the biggest positive disruptor of the way we work, the way we care for ourselves, and the way we live. I wanted in and joined Sentient initially as CMO.
How do you project the offering from Sentient in the overtly crowded cloud-based Data Science and AI tech landscape?
Our primary commercial offering, Ascend is unique among its field in both its capabilities and the way it achieves them. Ascend accelerates revenue and lead generation growth by accelerating and automating experimentation at scale, using evolutionary algorithms. Our competition isn’t other AI players, it’s companies offering A/B and multivariate testing solutions. The Machine Learning approaches some are just starting to adopt don’t hold a candle to our evolutionary AI in terms of the scale of experimentation we can offer.
Tell us more about Ascend and how your customers benefit from leveraging it.
Business and experimentation leaders from Jeff Bezos to Ronny Kohavi (Microsoft head of Experimentation) to Peter Diamandis have all spoken about the critical importance of experimentation in driving business success. The more you test, the more you learn, and the faster your results will grow. But traditional testing methods such as A/B testing, while valuable in certain contexts, are slow and very manual.
Ascend’s clients are able to run truly massive experiments, trying dozens of potential improvements to a single- or multi-page funnel all at once. Instead of comparing one design to another, as A/B testing does, Ascend uses evolutionary AI to search through thousands or millions of potential combinations of these improvements to rapidly get to the combinations that generate the highest lift. As a result, our customers win more of their experiments (4-6X more than A/B testing’s average), achieve higher lifts per experiment, and save a lot on the resource side as well due to Ascend’s automation.
How do you differentiate between technologies for AI and Machine Learning? Who are you competing within this landscape?
We don’t see other AI and Machine Learning companies as competitors! There are some many different disciplines within AI, and different use cases for each of them. We are proud to be members of the broader AI community and often recommend other AI-based solutions that solve other problems for our customers and prospects.
How do you see the raging trend of including involving AI and Machine learning in a modern CIO/CMO’s stack budget?
Just like with Ascend’s algorithms, this trend is based on survival of the fittest. If you aren’t investing in AI-based solutions to re-make key elements and processes within your marketing and IT efforts, you will not survive to the next generation of businesses. Due to the shortage of AI scientists and the difficulty and cost of recruiting them, the most effective path to bring AI into your business is to work with companies like ours and others that bring solutions to specific pain points or aspirations within your organization.
What is the data engine behind Evolutionary Computation? What data architecture drive your success with EC?
We’ve adapted Evolutionary Computation specifically for the domain of experience optimization. If you look at the way we do evolution, compared to, say, the work at Sentient Investment Management, they are very different applications of the same core technology. In our case, it was very important to achieve the fastest path to better results, since time and traffic are a limited commodity for the digital properties we optimize, whereas in building out intelligent trading desks the run evolutions for hundreds of generations.
In terms of the data we use, the inputs are the ideas to be tested — these are products of marketing, product, UX, agencies and other teams — and the data, all anonymous, is observing the conversion, spending, or engagement behaviors of users exposed to different combinations of those inputs.
What are the biggest challenges and opportunities for businesses in leveraging technology to optimize their customer support and customer success?
The biggest challenge is the wide range of scenarios that exist in both customer support and customer success. There are many different types of customers, and they have different challenges themselves, and different use cases for the product. Fortunately, this also has created a wealth of data to analyze to improve both processes.
Our technology is used more in the customer support area — how we optimize the presentation of support information to reduce phone calls and increase customer satisfaction. There are other technologies that look at how products are being used, to indicate the likelihood of success for a given class of customer.
How should young technology professionals train themselves to work better with automation and AI-based tools?
Learn about the different forms of AI, for starters. My intro tome was “The Master Algorithm” by Pedro Domingos. The book straddles a good line between information for the technologist and for the layman. Knowledge of statistics is also key for many forms of AI. Learn about data pipelines and data hygiene…many AI efforts are thwarted by problems in the core data set, even before the modeling ever begins. Network with your peers.
What is the biggest challenge to digital transformation in 2019? How does Sentient contribute to a successful digital transformation?
Digital transformation is based on digital innovation, and digital innovation stems from experimentation at scale — a challenge that has eluded all but the biggest companies. This is a problem we’ve solved.
Another challenge is the desire by some digital transformation teams to boil the ocean, as it were. The challenges in AI staffing, problems in corporate data sets, and the fixed-purpose nature of most Machine Learning approaches make this, most of the time, a fool’s errand. The better approach is to identify key areas of the business to transform one at a time and apply the best AI solutions to achieve transformation within them, while sketching out a broader in-house plan. In the area of experimentation, we believe we have the best approach, and with our larger customers, we are generally part of their digital transformation efforts.
How potent is the human-machine intelligence for businesses and society? Who owns Machine Learning results?
The advances that are happening, and around the corner, are staggering — in business, IT, communications, healthcare, and most human endeavors. We will solve cancer. We will solve hunger. We will allow humans to flourish creatively and be more productive while improving their quality of life.
Who owns the results? In our case, our clients own their results. In a broader sense, it is important to make sure that privacy remains at the forefront of our conversations, for AI and everything else. Personally, I lament the lack of government investment within our country in the AI sector, compared to other nations, as this will have a leveling effect in terms of bringing the AI opportunities to all concerned.
Where do you see AI/Machine Learning and other smart technologies heading beyond 2020?
It’s hard to predict this future as the shifts will be radical. Certainly — and we see this in our own technology work — there will be a syncretic combination of the various AI disciplines into new and more capable algorithms. And, sure, we will eventually see artificial minds, but I am of the school which thinks this is still a ways off.
One of the best and most alarming books I’ve read on this subject is Yuval Noah Harari’s “Homo Deus.” While he claims not to predict the future, he lays out a number of scenarios which could happen if we don’t steer this ship in the right direction. A lot of the fundamental principles of humanism and equality, initially driven by the need for labor and the need for military forces, could be obviated by AI solutions. Look at how we treat animals, and it’s possible that future humans/cyborgs/AIs could treat current humans that way. Read this book!
What is your opinion on “Weaponization of AI and Automation”? How do you promote your ideas?
I’m not sure whether you’re referring to AI for military applications, or just how we promote our ideas! In terms of the military, we don’t go anywhere near that space.
Our focus in “weaponizing” AI is allowing companies to gain a critical competitive advantage over their competitors by helping them create and continue to enhance their customer experiences — the visual presentation, the recommendations, the content, the mobile apps, and so on. If companies want to weaponize in this way — draft us and we will serve!
The Crystal Gaze
What Cloud Analytics and SaaS start-ups and labs are you keenly following?
What technologies within AI/NLP and Cloud Analytics are you interested in?
Auto-ML solutions, whether for technologists or non-engineers. Data hygiene solutions that apply AI to clean up corporate data.
As a tech leader, what industries you think would be fastest in adopting Analytics and AI/ML with smooth efficiency? What are the new emerging markets for these technology markets?
We’ve found the greatest adoption to be at those companies with the most at stake in a shifting world — retailers are a big sector, consumer financial companies (credit cards, banking, insurance, etc.), travel companies, and players in the affiliate marketing and lead generation space. Emerging sectors include brands and media.
Read More: Interview with Cédric Carbone, CTO at Ogury
What’s your smartest work-related shortcut or productivity hack?
TripIt, given the amount of travel I do
Tag the one person in the industry whose answers to these questions you would love to read:
Thank you, Jonathan! That was fun and hope to see you back on AiThority soon.
Jonathan Epstein is senior vice president of strategy for Sentient Technologies, a leader in the field of autonomous optimization. In that role, he manages all strategy for the company, key business alliances, and oversees the company’s Asian business. Epstein has operated at the vanguard of marketing, media and technology for his entire career, as founding CEO of GameSpot (the world’s leading games site), president of GameSpy, general manager of IGN, president of Omek, and CEO of Double Fusion, among other roles. He holds a degree in physical sciences from Harvard College, and has co-authored three patents in diverse fields including remotely operated underwater vehicles, in-game advertising, and gesture recognition.
Sentient’s mission is to transform how businesses tackle their most complex, mission critical problems by empowering them to make the right decisions faster. Sentient’s technology has patented evolutionary and perceptual capabilities that will provide customers with highly sophisticated solutions, powered by the largest compute grid dedicated to distributed artificial intelligence.