Speech Technology Magazine

 

Digital Assistants to Bring in $623 Billion by 2020

Digital assistants are expected to generate $7.9 billion in revenue this year, but that amount is expected to balloon to $623 billion by 2020, says an industry expert
By Leonard Klie - Posted Nov 10, 2016
Page1 of 1
Bookmark and Share

Digital assistants are expected to generate $7.9 billion in revenue this year, but that amount is expected to balloon rapidly to $623 billion by 2020, predicts William Meisel, president of TMA Associates, in a new industry report.

Meisel says he hesitated to predict such large numbers, “but the logic to me is inescapable.”

Digital assistants, he explains, “are increasingly becoming the new way companies and applications will interact with users.”

A main factor behind this growth is the fact that “personal assistants are fighting to be your portal to the web, including outside companies and services,” he adds.

The report defines digital assistants broadly as any cloud-based application that responds to natural language inputs.

Though the total market landscape includes about 170 vendors, including the providers of full digital assistants, messaging apps like Facebook Messenger, specialized solutions, core technologies, and niche players, Meisel points out that only 39 vendors “are capable of supporting the building of a full specialized text or voice digital assistant.”

In terms of impressive technology, “Amazon’s Alexa uses voice only, is quite accurate, can even understand speech while it is playing music, has low latency, and is probably the best example of what can be done with natural language voice interaction,” Meisel says.

Among some of the other industry leaders mentioned in the report are Apple’s Siri, Microsoft’s Cortana, Google Assistant, Viv (recently acquired by Samsung), and solutions from Nuance Communications, [24]7, Interactions, and IBM.

Viv, Meisel notes, “is likely to be a significant competitor when launched, using a collaborative approach to building its knowledge base.”

Although IBM is a relative newcomer to the market, Meisel expects it to become “a major entrant with its conversational and speech APIs.”

And while many of the current applications, including the wide array of chatbots and messaging bots, emphasize text right now, Meisel suggests even they will add a speech option very soon. “Adding a speech recognition front end is a relatively simple next step” once natural language processing has been incorporated into the application, he says.

Another motivation for using speech, as opposed to text, is that an individual is more likely to speak in fuller sentences and provide more information. This can allow more accurate natural language understanding.

Meisel also expects the nature of the speech interface in most digital assistants and bots to change.

When the interaction is by voice, front-end speech recognition converts the speech to text to allow natural language processing. Today, that is done largely by cloud-based large-vocabulary speech recognition that is often independent of the specific application.

“In principal, since the context is limited, the speech recognition could be tuned to the application, and thus be more accurate, but that would involve specialized training of the speech recognition software,” he says.

Meisel also expects to see more integration between natural language processing and speech recognition, “so that the language that the NLP can understand will impact the probabilities in the speech recognition algorithms’ language model.”

Also look for natural language processing to change, particularly as companies experiment with neural networking, machine learning, and artificial intelligence. “It can be tolerant of speech recognition errors if so designed,” Meisel points out.

If, for example, an application’s speech recognition makes typical errors, such as “a count” instead of “account,” the natural language engine can be designed to still carry out the requested task.

“I’ve seen Siri do something like this when it displays an incorrect transcription but takes the right action. This is similar to what humans do in a noisy environment, when we don’t hear every word in a conversation but get the gist enough to carry on the conversation.”

Still, for all the business applications available, Meisel and others expect the enterprise segment to make up only about 17 percent of the total market in 2016, dropping to 13 percent of the market in 2020 as consumer assistants expand rapidly.

Though market research firm Tractica predicts a fourfold increase in the use of intelligent virtual digital assistants, virtual agents, and chatbots in the next five years, it predicted recently that the number of people globally who actively use the technology will grow from 390 million today to 1.8 billion by 2021. The number of business users will grow from 155 million today to 843 million by by 2021, it predicts.

Page1 of 1