The technology world is abuzz about “bots.” Or, at least, they have increasingly become an area of focus and funding, according to Mizuho analysts Neil Doshi and San Phan after attending the MobileBeat “Bot” Conference in San Francisco this July.
While hype may seem outsized to reality, Mizuho finds promise in emerging bot technologies after discussions with VCs and entrepreneurs at the conference. Particularly with advances in machine learning and artificial intelligence, bots are prompting a spike in attention from techies and investors alike.
So what are bots exactly, and what’s to come?
The Bots We Know Now
A bot is software that automates certain basic tasks that can be accomplished by a device, such as sending a message, making dinner reservations, adding an event to a calendar, or, in the future, accomplishing tasks like retrieving your parked car. Bots can be enabled using both voice and text.
In fact, you’ve likely already interacted with a bot or two, the most common being Apple’s Siri, which first became available with the release of the iPhone 4S in 2011. Capable of accomplishing relatively basic tasks, such as conducting an Internet search or displaying the current temperature, Siri was among the earliest examples of a vocal bot.
After half a decade, it is clear that Siri is not the end game for bots. While it may have been the first to market, the leader right now is Amazon’s Alexa, considered by Mizuho’s Doshi to be “the bot of bots.” Alexa is currently able to master more than 1,500 “skills,” including household duties such as controlling lights and adjusting the thermostat. Developers continue to create additional skills for Alexa at a rate of several hundred per week.
A surprising counterpoint to Amazon’s success, Doshi notes, is that Google seems to be late to the table in developing consumer bot products, despite owning some of the best artificial intelligence and machine learning capabilities out there.
Another prominent application for bot technology can be seen as companies seek to meet consumers where they are by providing a frictionless experience, particularly leveraging text-enabled “chat bots” through social media and other digital touchpoints. Chat bots require textual input and respond in the most appropriate manner, which can include automated actionable responses.
According to Mizuho, Facebook now has around 20,000 developers building chat bots for Facebook’s Messenger, rendering it the leading platform for chat bots at the moment.
Burgeoning Bot Technology
The software behind bots has not been without issue, specifically the limitations of current artificial intelligence in understanding human speech and developing responses beyond a limited set of specific phrases.
As VentureBeat explains, bots understand little of what humans are actually saying, instead scanning for a preset list of keywords to decide what response is most appropriate. The bot then chooses from a list of agents, or programs within the bot, that are designed to accomplish different tasks.
For example, asking for the front door to be locked would call for a specific agent, while scheduling a dentist appointment calls for another. With multiple available agents, the bot has trouble choosing between them. VentureBeat calls this the “multi-agent problem.”
Going deeper than simply responding to basic, pre-determined scripts, bots are developing the ability to provide more intelligent responses, such as including information that may be relevant to the question at hand, according to International Business Times. For example, when asked about a musician, a bot may intelligently recognize that the musician will be playing in a nearby city and ask the user if he or she wishes to see ticket information or transportation options to the concert.
A Bright Bot Future
While bots sport a bright future, investing heavily in a “bot-strategy” may not make sense for many brands at the moment, assert Mizuho’s Doshi and Phan. While in certain cases bots augment opportunities to scale customer service, facilitate transactions and improve in-app search experiences, initial development costs are still quite high.
Bots transcend pop-culture prominence as AI, machine learning and chat capabilities continue to advance. Today, while some of the dominant industry players have yet to throw their hat in the ring, the constant pace of technological development and ravenous consumer appetite for automated services points to an impending bot invasion.
Fortunately, these bots are much more likely to order you a pizza than appear in the next Hollywood science fiction reboot.
Simon Hylson-Smith is a former financial industry editor and currently CEO of Paragon.