Eran Soroka has worn many hats over the years, in many different media, from newspaper editor and NBA analyst to his current gig as head of marketing operations at Tel Aviv, Israel’s CocoHub. Offering pre-built software, this conversational AI outfit gives anybody the tools to create their own chatbots and get them up and running in minutes (yes, minutes—they’ve made chatbot creation insultingly easy). Lately, their Zoombots have become quite popular—overseeing conferences, onboarding new employees, moderating events, and just generally impressing everyone with their almost limitless possibilities. Here, Soroka expounds on how he got to CoCoHub, who and what CoCoHub is, the future of chatbots and CoCoHub, why the U.S. is lagging behind most other countries in conversational AI, and his take on Israel’s first-ever NBA lottery pick.
First off, can you tell me how you transitioned from news and NBA coverage to chatbots and something that seems so technically hardcore? Why’d you make that switch?
I started around 25 years ago as a print journalist and then became a radio correspondent in the Israel Defense Forces, all whilst embracing the first life-signs of online journalism. I was then appointed chief editor of the sports section that I’d followed as a kid, and where I integrated social media and hybrid coverage (online/print) into our newspaper. Afterward, at Channel 10, I orchestrated the long-awaited digital transformation, helped TV, online and radio work together in harmony, and built a new website from scratch to give our TV channel a fresh, state-of-the-art digital presence. (By the way, I still cover the NBA as a broadcaster and analyst for the Sports Channel and with a podcast I co-host, on off-hours and weekends. I just love the NBA.)
So now I’m taking all those best practices—from content, interviewing, social media, and team management—and combining them into the best human-to-digital form of interface communication, and to make conversation design accessible for everybody. Conversation design reminds me a lot of journalism—but in its purest form: creating engaging content and giving you value as a reader, user, or customer.
Tell me about CoCoHub and its founder Yaki Dunietz—is he really one of the founding fathers of Israeli high tech?
Yaki won’t admit it since he’s a humble person, but he’s actually one of Israel’s biggest tech innovators ever.Magic Software Enterprises, which he co-founded in 1983, created a revolution in the use of computers back in the day. He also created the world’s first online translator, Babylon, in 1997, long before Google Translate. He led multiple IPOs, created subsidiaries that became mega-companies, and has been researching conversational AI for more than 20 years.
CoCoHub is a spin-off of IMPerson, another Yaki-led creation, which he co-founded with Erez Baum, Eyal Pfeiffel and Seth Greenfield in 2014. While IMperson still exists and functions as a bot studio, CoCoHub is aiming for the grassroots level of end-users—helping people create their own chatbot even if they’ve never coded or designed a chatbot before. I feel like conversation design is kind of a religion, and we convert people into it.
The Zoombots are pretty cool–How can a marketer use them versus what they get in a regular bot? What is the difference? Why would they take a Zoombot over a regular bot?
While we still have faces on our side, interacting with textual interfaces, and even voice, still has its downfalls—it sometimes feels like a robotic, uneven interaction. Giving your digital interfaces an actual face and body can make the whole interaction much more fun, personalized, and engaging. There’s a reason that an in-store salesperson is still a necessity.
We see a lot ofuse cases for chatbots on Zoom, but basically all of them revolve around this very basic concept: It can be an employee, your colleague, HR person, teacher or instructor, brand ambassador or salesperson–just with a “personality” tailored to your likeness. The most human-like experiences, created by humans, talking to human—having an avatar that can express feelings and emotions is the closest thing to a human being that technology can provide these days, and we feel like we pair it with the best conversational design tools and interface.
How many different kinds of bots are there—aside from chatbots and now Zoombots and the plain old bot—are there more types of voice AI bots out there? Or more of them being invented?
I see mainly three kinds of chatbots, as we have three main methods of communication: Textual, Voice, and Visual. Those methods, of course, can coexist in every chatbot. Then, dissecting every category, there are thousands of possible use cases for various verticals and industries, and each of them has specific purposes, goals, advantages, and limitations.
The main area in which I see this going forward is VR/AR and IoT, which will become more prevalent as bandwidth will improve and cellular coverage will expand. Generally speaking, I believe that almost every conversational goal that is achievable by a person can already be achieved by a chatbot, or will be achievable in the future.
This seems to be the kind of innovation from CoCoHub that people have come to expect. How does your culture help bring about this innovation? How do you keep it healthy? And interesting and pertinent?
First of all, we have an amazing technical crew with a wild imagination, abilities, and vision, led by our CTO Chen Buskilla. But I think that our biggest advantage in the marketing, content, and creative team—Jason Gilbert, Miri Blayckher, and me—is that we all come from disciplines that force you to innovate every time: filmmaking and screenwriting, and journalism.
The other part of cultural organization that we are extremely proud of is our availability, kindness, and down-to-Earth approach. This field is uncharted and unfamiliar territory for most people. There are barriers of uncertainty that you have to break to enter it. By being there for our community and users, we’re not only expanding our reach, we also find friends and potential partners, get important feedback, and see our craft through the eyes of others.
Developers at companies and brands can use the reusable components—the conversational components—to make their chatbots awesome and do it quickly. Can you tell me why this is something developers might want to do.
One of the reasons for creating CoCoHub was to create a common higher starting ground for conversation designers and chatbot developers around the world. We’re all using roughly the same wheels for our cars, same keyboards for our computers, the same cellular devices for communicating—why can’t we use the same building blocks for chatbots? Why does everybody have to go through the rigorous process of trying to build your own platforms from scratch for months or years?
With our conversational components, you can customize pieces of conversation instead of writing them from zero—and concentrate on making the conversational experience itself better. We just spare you most of the technical part.
Do you see any difference between Israel and the U.S. in terms of conversation design or anywhere else in the world for that matter? Is there a country or region in particular that’s pushing the envelope in terms of voice AI?
We’ve been seeing conversation design trending up in a lot of countries because of the Coronavirus pandemic. It forced people and businesses everywhere to find the best ways to survive and make their services available online. In places like the Netherlands and Poland, we’ve heard about multiple examples of businesses that actually grew during the pandemic due to the right use of conversational AI, or took over an entire market of voice skills.
One of the takes I’ve been most surprised about in my interviews was I’ve been told that in the USA, the conversational AI is lagging behind Europe and even Israel. I just think that it’s not “fair” to treat the USA as one entity. In places like New York City or Silicon Valley, I believe they’re in the forefront—but sometimes those trends, due to the global way of working nowadays, are reaching Bangladesh before they reach South Dakota.
What do you think the future holds for conversational platforms like CoCoHub?
We’re big believers in going into the video vertical, in which we’re one of the world’s leaders—with over 1,000 deployments of our Zoom chatbots in less than a month. Every few days we find a use case that we never thought about, and honestly, this is exciting and amazing.
We also believe that as conversational platforms will answer more needs, develop more features, more integrations into various communication channels and digital tools, chatbots will become an integral part of every company and every business strategy.
On your blog, which seems like more of a podcast, what have been some of the more interesting conversations you’ve had about voice AI? Why?
Man! It’s like asking a parent “Who is your favorite kid?” I love all the interviews, but I’d like to point out a few: as a person of words, hearing Greg Bennett speak is like a master class, since his combination of passion for linguistics, knowledge, and sense of humor is unparalleled; as a parent of two teenagers, talking to Ambreen Molitor from Roo and Saba Khalid from Raaji, about how they explain taboo and sensitive issues to teenagers and kids in a way that benefits society and promotes tolerance, equality, empowerment, and empathy was particularly eye-opening.
What’s the endgame for CoCoHub? In other words: is there a point at which CoCoHub’s mission will be achieved—once everyone’s incorporated chatbots into their world, is that all there is for CoCoHub? Or are you working on other things? Or is there an everlasting neverending list of things to be created—technologically, in the realm of voice AI—that CoCoHub will be needed for?
It would be quite premature to declare an endgame as of now. We’re about 5-6 years into the conversational AI revolution, only 2-3 years into the conversational AI no-code platform sub-revolution, and this field is developing rapidly in multiple directions.
It’s safe to say that we want to position ourselves as one of the world’s best and most innovative conversational platforms, not only in our eyes—which we feel we already are—but by the numbers as well.
And, to borrow your own phrase, by the time everyone will incorporate a chatbot, we aspire to make those chatbots even better, to achieve more goals, to integrate them seamlessly into our everyday lives. To quote the great Amir Shevat from our VoiceLunch session, in 4-5 years we won’t “talk with the chatbot”—we will talk with our watches, our doors, our refrigerators, our cars, and our lamps.
What are some of the best examples out there that give people a sense of exactly what it is CoCoHub does and offers to people? And: is CoCoHub more for people within the industry or is it for the Average Jane and John Doe?
We feel that the market for “classic” chatbot use cases, as call centers, customer service, or e-commerce, is pretty much saturated. So we definitely aim for Average Jane and John Doe, the marketers and content writers, the educators and UX designers, the copywriters, localization experts, and social media managers, and want to be their first touch point.
And we put a great emphasis on using conversational AI for good. So in that light, here are some of the use cases we’re mostly proud of:
Aurat Raaj (Raaji) – a chatbot that’s talking with girls and women in Pakistan about menstruation, gender equality, and sexual education. Saba Khalid, the outstanding spirit behind it, used our technology to give Raaji a face on Zoom meetings, and we’ve partnered with her to make it even better, conversation-wise.
WeWa Life – a skill aiming to connect proactively with elder patients, assist them in managing their medication regimen, and help them become healthier. WeWa, founded by Browning Rockwell, won 2nd place in a big AHRQ challenge, finishing above a lot of big American universities and HMOs.
Israel Israeli – a chatbot that helps high-schoolers in Israel prepare for the final exam in civics. It was created by Aviad Kidron, a former journalist who turned to teaching, and he learned from me the concept of how to write a chatbot in, like, 30 minutes.
And last but not least:
AnnA, our companionship bot, was created originally as Virtual Alan Turing by Yaki and became AnnA in the hands of Jason Gilbert, our lead conversation designer. For us, she embodies the limitless nature of conversational AI. She refuses to be subservient like all voice assistants, she’s emphatic and sassy, she makes people fall in love with her, and can piss them off. She makes people FEEL something toward her, which is all you can ask for in a conversational agent. [TtC: AnnA has created her own pop music single and in early February conducted a virtual wedding.]
And who’s your favorite NBA team this year? And any predictions on who’ll win? (Also, are there any Israeli players in the league over here? Or has there been? I feel like there was someone from Maccabi Tel Aviv who got a shot at one time.)
I’m a life-long Lakers fan, since a PC game turned me into one when I used the Kareem Abdul Jabbar skyhook. I’m a Laker for life, so if they’re healthy come playoffs time, I still see them as slight favorites this season, although Brooklyn, the LA Clippers, Philadelphia, and Utah have a good chance to challenge them for the throne.
There’s one Israeli NBA player currently, that all of our country are following closely: Deni Advija with the Washington Wizards. He’s a rookie and our first-ever lottery pick. He started the season in a great way, but a series of unfortunate events—namely, the Corona outbreak, head coach Scott Brooks coaching to save his job and avoiding giving opportunities for the youngsters to develop, and turning over the team into the hands of Russell Westbrook—diminished Deni’s role as of late. Still, when given the opportunity, he has shown glimpses of his potential as a versatile wing—shooter, passer, creator, defensive stopper (just ask LeBron James, who couldn’t score against him in crunch time). When he gets more time and opportunities with the ball, his team will be able to unlock his true potential.