Alexa co-creator: Digital world forgot importance of real dialogue

"We got obsessed with the interface as opposed to how we use technology to really connect and understand the consumer," said Alex Spinelli, who now serves as LivePerson chief technology officer.

JULY 12, 2019 04:34

Alex Spinelli, LivePerson CTO and former global head of Alexa for Amazon (photo credit: PR) — Alex Spinelli, LivePerson CTO and former global head of Alexa for Amazon

(photo credit: PR)

Obsessions with swiping, tapping and shiny glass surfaces have led enterprises to forget the importance of truly understanding and building a real dialogue with their customers.

That is the strong verdict of Alex Spinelli, LivePerson’s chief technology officer and former global head of virtual assistant Alexa for Amazon.

“We argue that the experiences that we’ve built don’t work, and is one of the key causes for big businesses dying faster than ever before,” Spinelli told The Jerusalem Post at LivePerson’s Tel Aviv office.

“We got obsessed with the interface as opposed to how we use technology to really connect and understand the consumer.”

Approximately 60 big businesses that were listed in the Fortune 500 in 1955 remain there today, with Spinelli blaming the fall of many enterprises on excessive emphasis on interfaces and building digital experiences that don’t work.

Having held senior technology leadership positions at several major global companies prior to joining Amazon and LivePerson last year, Spinelli doesn’t absolve himself of all blame either.

“We have put proxies in the way of having real dialogues and conversations with the consumer, such as interactive voice response (IVR) technology, phone tree software systems and search and browse,” he said.

“There is a much better way to exchange information. Dialogue and conversations are how we build confidence and relationships as humans, and personalize the discussion.”

Significant advances in mathematics, artificial intelligence and natural-language understanding in the framework of the so-called Fourth Industrial Revolution, or Industry 4.0, are now breaking down technological barriers that previously prevented human-to-human dialogue at such a scale.

“We will be doing more and more with artificial intelligence, but we don’t feel it’s the goal to replace people with technology,” Spinelli explained.

Stay updated with the latest news!

Subscribe to The Jerusalem Post Newsletter

Subscribe Now

“The simple, repetitive tasks will be given to machines, helping humans really dig into the challenging problems and relationship-building.”

Spinelli is now responsible for leading the technology organization and overseeing all global R&D for LivePerson, a New York City-based company developing conversational commerce solutions based on artificial intelligence for major brands.

LivePerson employs more than 325 employees at its Tel Aviv and Ra’anana R&D centers, leading much of the company’s technology and intelligence data developments.

The company’s AI-powered chat-bots foster engagement with consumers via the most popular messaging channels, including SMS, Facebook Messenger, Whatsapp and Apple Business Chat.

“When you hear about chat-bots, the name is almost minimizing what the industry is talking about nowadays,” said Spinelli.

“We’re really talking about artificial intelligence-powered business, hyper-perzonalization, deeply integrating with enterprises and engaging with consumers to build relationships and dialogue.”

While Spinelli is yet to see any technological limitations to potential artificial intelligence applications, he does cite existing issues regarding consumer confidence.

The first step to increasing consumer confidence in artificial intelligence, he says, is transparency. Bots need to identify themselves and also offer an “escape hatch” to a human where complex problems are involved.

“Humans will be good at understanding complex problems for quite a while, but machines are already much better at things like matching patterns,” he explained.

Standards, guidelines and best practices are also necessary, he added, to allocate liability and traceability when errors are made by artificial intelligence, and to ensure the creation of unbiased systems.

Reflecting biases in society, human training of artificial intelligence systems with bad data can lead to discriminatory algorithms based on race, gender or other attributes.

“We need to hit these issues head on, talk about them and bring people together to put the right regulations in place,” Spinelli said.