How Siri, Alexa and Amy enable our conversational future
Written by: Lisanne van Nieuwkerk
6:30, the alarm sounds and with some reluctance I get up. Entering the living room I ask: “Hey Google, what’s the weather today?”. The coloured Google dots light up and a friendly female voice says: “Good morning, today in Amsterdam it will be cloudy, with a few showers and a forecast high of 18 and a low of 14 degrees celsius”. “Do I need an umbrella?”, I ask. “Yes, there will be showers between 8 and 10 AM.” Looking at my coat rack makes me realise I don’t have one. To make sure I won’t be having this situation again I ask Google to order an umbrella for me. “It will be delivered tomorrow evening.” “Thanks Google, you rock.”
Interacting while using natural language
Siri, Alexa, Google home and Amy are all examples of conversational bots: digital agents that interact with users using natural language. These computer programs are designed to mimic human interaction and are powered by artificial intelligence.
And it works. Although I’m clearly talking to a small round device, I’m using natural language while responding, just like I would have to a human. And the more I do that, the more the bot learns to answer humanly. It’s not a coincidence that most bots, computer programs that converses in natural language, get a human name like Alexa. Maybe it is not even a coincidence that these are most of time female names. Perhaps this is due to the kind character of the bots, but a relation to the female talent of talk cannot be denied right?
How the conversational ball got rolling
Have you seen the movie ‘The imitation game’? The film is about the English computer scientist Alan Turing and the machine he built that conquered ‘Enigma’: the encoding machine the Germans used during world war II. Around 1950, 4 years before he died, Turing invented the so called Turing-test: a test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.
In 1966 Eliza, a bot which mimicked human conversation by matching user responses with scripted responses, first passed this test. This bot was used in the psychiatry. This was the start of building and using human-like bots.
In 2004 A.L.I.C.E was released and got rewarded as best bot. They created a well working bot by using heuristic patterns in digital conversations which made the program ‘human’.
And then, in the 2010s the last so called AI-winter is over. Machine learning clearly is progressing rapidly and the error-rates of image- and speech recognition are dropping significantly: enabling the advancement of intelligent and self-learning robots.
In 2010 apple launched Siri, a cheeky and intelligent personal assistant based on natural language processing. Rapidly others followed: Google Now (2012), Alexa by Amazon and Cortana by Microsoft (2014).
The anchoring of these digital assistants
Now, in 2017, according to Vertuo Analytics Siri loses some growth but is still the most popular digital assistant with more than 41 million users (US) a month. Meanwhile, Amazon’s Echo (based on Alexa) and Cortana show huge growth numbers.
These numbers are only US, but also the worldwide statistics show the anchoring of the conversational agents. Tractica calculated the use in 2017 and stated that 710 million people are already using digital assistants like Google Home, Alexa, Cortana & Siri.
The estimate for 2021 is that almost 2 billion people will service themselves with the use of the agents. A study by Google and Northstar Research show that 55% of teens use voice search on their smartphones daily. Acceptance of the digital assistant will grow as etiquette evolves and capabilities grow, so they say.
Messaging will be the next platform
The rise of conversational agents by voice is accompanied by the rise of the use of chat-messaging apps. With a growth of almost 400% (source: Flurry) the messaging and social apps definitely show that people are getting used to service themselves via a messaging app: get restaurant options via your whatsapp, pay directly by credit card and share the location directly with your friends. Why not do it all in one app instead of using four?
It gets more and more likely that we will start using messaging platforms as the baseline for running our daily life. And so we believe: messaging will be the new platform.
Got convinced? Up next: The potential of the conversational future.
Getting excited or getting scared? We always like a good conversation, connect with us via www.sound.team or just reply on this article.
Get your eyes opened: Blockchain
Every month we help companies embracing the future by setting up a eye-opener breakfast event. Within 2 hours you are not only ready for the day, but also for the future.
Apply here with your name, email-address and phone number for our next session on November 8, 7:30–9:30 @ our office.
