Bots 101: Understanding chatbots, AI, digital aides
If you follow technology trends, you could be forgiven for having “bots” on the brain.
It’s a word that over the past year or so has become ubiquitous, not unlike “The Cloud” or “The Internet of Things” in past years. Bots, it sounds like, will help you o rd e r a p i z z a wi t h o n l y your voice. They’ll remind you when you’re late to an appointment. They’ll help your grandparents settle into retirement as friendly, plastic helpers who roam around the house, automating tasks.
Wait, is that last one actually a “bot?” Or something else? What is a “bot,” anyway?
Good question! It turns out that this catch-all term may not mean what you think and that even people in the so-called bots industry think the term has become overhyped and under-defined.
So here’s a simple guide to “bots” and all the things they are and are not.
What is a bot, anyway?
One of the reasons the word “bots” has become so popular is the rising popularity of the group messaging service Slack and the rise of user-friendly voice-activated home gadgets such as the Amazon Echo.
Brad Knox, a computer science Ph.D who does robotics research at the Universit y of Texas, says “bots” for a long time has in tech terms meant an automated program that runs over the internet. On Slack, someone could write a bit of “bot” software that automatically posts new Google Calendar events to a chat channel or one that spits back a relevant animated GIF when someone types in a trigger word.
The rise of so-called “chatbots,” however, which can perform tasks on services such as Twitter and Facebook Messenger or speak to Amazon’s Alexa to order a pizza has raised the profile of “bots.”
Last year, bots became such a huge topic of conversation that an Austin company, Howdy.ai, organized a conference around the subject of chatbots.
So I’m not the only one who’s confounded?
B e n L a m m , C E O a n d c o - f o u n d e r o f t h e c o mpany Conversable, has done work in automated, conversational software for companies including Wingstop and Whole Foods. He says “bots” has become one of those terms that has created confusion in the market and says the industry is actually starting to move away from the term.
“To date, there has been too much emphasis put on bots instead of core conversational intelligence powering the bots,” Lamm said. “People are starting to understand that ‘bots’ doesn’t mean intelligence and certainly doesn’t mean AI (artificial intelligence).”
Knox says that “chatbots” and “bots” have become interchangeable terms in some circles, adding to the confusion. “My sense is that there’s an Alexa-triggered gold rush going on, and the usage of ‘bots’ is part of the hype around that.”
Confusing things even more is that “bots” is still a word that people associate with “robots.” You know, those metal things that walk around in old sci-fi movies going “Beep boop boop!” (Apologies to “droids,” which probably deserve a separate article.)
Knox is currently heading up a playful Kickstarter project called bots_alive that uses “HexBug” toy spiders, a smartphone-connecting infrared blaster and obstacles to teach kids programming principles.
He says robots that exist in real life and software “bots” that operate online are not all that different when you think about it.
“A robot has the physical world as its environment. A software bot also has an environment, which is everything in the software system that provides context for how it should act,” Knox said.
Got that? Let’s move on.
How does ‘AI’ fit into all this?
For “bots” to have any value at all, they have to have some context. And in his business, Lamm says, there has to be a big picture where bots are just a part of what he calls “the conversation intelligence puzzle.”
“To use a metaphor, bots are the ears and the mouth — they listen and respond. They are an endpoint. You still need a centralized brain to make the decision,” Lamm said.
A larger part of the picture, for inst ance, might be a database of recipes or menu items offered by Whole Foods or TGI Friday’s and the mechanisms that allow a user to communicate with an SMS- or Facebook Messenger-based chatbot with a high degree of accuracy.
AI and machine learning have become huge areas of academic research and a growing part of the software industry. If AI is machines being able to perform tasks in a way that can simulate human intelligence, machine learning is the deeper concepts of computers learning from data and interpreting it for themselves. (Make implications of that as you will.)
It’s a major area of study at the University of Texas Computer Science department and has fueled startups including Diligent Droids, Cogitai and Apptronik. Every year, there’s a growing Data Day Texas event that largely focuses on these topics. And there’s even several AI and chatbot-focused meetups groups in Austin.
So what can bots do for me?
Even if the term falls out of favor, the tasks that bots can do for us only seem to be getting more useful.
“Chatbots” will continue to offer new ways to perform tasks over messaging services we already use.
Bots can be made to do novel things such as purchase stock and donate the profits any time president Donald Trump tweets negatively about a publicly traded company.
A bot could do a language quiz with you or allow you to play “Zork” via Messenger.
A s wi t h t h e hy p e t h a t greeted apps at the early years of smartphones, it’s important not to get too caught up on the idea that bots, however useful they might be, are going to make everyone’s lives easier.
“When the bot wave hit, there was a lot of noise and belief that a bot will solve any problem, as if bots were going to be the Iron Man’s JARVIS come to life ... bots by themselves are actually quite simple and can’t do much,” Lamm said.
B u t b o t s a r e p a r t o f a b i g - p i c t u r e s e a c h a n g e that, in a lot of ways, will change the way some of us get through the day, how we use the devices around us, and how we think about the machines that (for now, at least) serve us.