I built my first chatbot in the 90s, and I haven’t really thought about it since…
Technology has evolved dramatically in the last few decades alone. Think back to your first computer — the one that probably ran Windows 95 and that you likely spent a few thousand dollars or more on. The one that you had to buy a new (and sturdy) desk to fit in your home. Now, consider that an off-the-shelf laptop purchased from any regular big box electronics retailer is not only many times more powerful than that was, but many times less expensive, too.
Software in particular has gone through several “golden ages” — first you had the “Swiss Army Knife” programs that “did it all” like Microsoft Office, which gave way to the “app-centric” revolution that we’re currently living in.
But throughout all of that, one thing has remained largely unchanged — the ways in which we communicate with that technology. Even though graphical user interfaces and the way we control our apps has evolved from point-and-click adventures to touch-screen interfaces and back again, we’re still not interacting with software the way our brains would like to. We’re not communicating with a piece of technology like we would with another human being.
This, in essence, is what chatbots are all about. Chatbots are simply defined as computer programs that are designed to “simulate conversation with human users, especially over the Internet.” Imagine that you send your friend a text message telling him to buy you a plane ticket to Madagascar next month. You wait a few minutes, your friend emails you confirmation of your purchase. Now, imagine that your friend isn’t a human at all — he’s a piece of software designed to act and INTERact like a human.
This is the raw potential that chatbots bring to the table and, in truth, they’ve been around for a lot longer than you probably realized.
The Rise of the Bots
Anyone who was an avid user of IRC during the 1990s to early 2000s can tell you that chatbots are hardly a new idea. They were a common thing even then, and could be used to complete a wide range of different tasks like:
- Managing channels; you could give a chatbot user privileges and allow them to boot spammers, provide information, etc.
- Tell jokes and instructional information.
- Search for applications, both above-the-board and illicit (ie: warez).
- Provide stats and reports to users and more.
In those days, creating different bots to automate different tasks wasn’t just common — it was fun, too. It was hardly user friendly and it required at a minium basic programming and Linux skills, but it became something of a facinating hobby for many.
Eventually, people set to work creating tools that helped them make better chatbots that could even run in your IRC client. Like ircN add-ons. Essentially making you automatically respond to things when you were “AFK”.
Along with this came ALICE, or “Artificial Linguistic Internet Computer Entity.” It was the earliest form of what we now refer to as AI, or “artificial intelligence.” If you ever typed a question into “Ask Jeeves” and phrased it as an actual question (and were lucky enough to get a page of results that were actually relevant to what you were looking for), it was all made possible due to some very similar concepts and ideas.
Since as early as the 1990s, this methodology was very common in the computer game space in particular. Though this may seem like a “wasted” use of groundbreaking tech, it is actually one of the major reasons why AI has come as far as it has. Developers used AI to make entities like fully autonomous characters that could interact with the player (think: “The Sims”), non-player characters that could aid you on your adventures (think: “Half Life”) and even entire cities that could function largely independently (think: “Sim City.”)
Throughout this period there were a huge range of different memorable games that were all built on solid, conversational-based AI. Just a few of them include but are certainly not limited to ones like:
- Zak McKracken and the Alien Mindbenders
- Leisure Suit Larry
- Escape from Monkey Island
- Dragon Quest IV
The New Trend
Flash forward to today and it should come as a surprise to absolutely nobody that the gaming industry is still at the forefront of conversational AI. However, the smartphone revolution that is still taking shape has led to a number of essential advancements.
This is particularly true when it comes to virtual assistants like Siri, Google Now and Cortana — all of which are widely used around the world. You may not type your conversation and instead speak it out loud, but the basic premise is still the same. You use them for things like navigation, sending texts or emails, calling contacts, web searches and more — causing the software to gain a lot of momentum in the process.
We’re even seeing a wide range of different bot commands turning into real, natural languag processing thanks to advancements in machine learning at the same time.
Software as a Conversation: The Way of the Future
We believe that the natural evolution — and natural next step — of all of these concepts is what we’ve been calling Software as a Conversation. Think about it: with services like IBM Watson, Microsoft Cognitive Services and Wolfram Alpha, we have access to powerful tools that we can use to understand speech, context, emotions, readability, images and much more. The new user experience emerging as a result is more about structuring and scripting conversations, guiding users by asking questions and offering choices in natural language.
It’s context aware and analytical. It’s the best of all possible worlds, converged into one simple and sophisticated solution.
We’re not alone in this thinking, either. New platforms that make it easy to create these bots are popping up seemingly every month:
Even platforms where these bots can live have become increasingly popular:
Chat apps already use text, audio, video, links, buttons, emoticons, gifs and more to allow people to communicate efficiently together. When you can connect software to this platform, why SHOULDN’T you be able to do so through a flexible chat app like Slack?
So far, we’ve been using Software as a Conversation for a host of unique tasks like:
- Connecting to visitors and users of website/apps to support and sales via chat widgets
- Running daily standup meetings with our employees and partners spread across the world
- Collecting and making knowledge available in Slack through https://niles.ai
- Give reminders of daily focus goals
- A wide range of commands, like sending SMS and email by comments in Podio
- Informing developers system errors
- Collecting micro feedback for HR purposes
- Notifying clients and partners of new features, when bugs are fixed or when their feedback is needed
- Marketing and lead generation
We’ve even developed a prototype that asks you about your mood and replies with a relevant motivational quote to make you feel better!
As far as the future is concerned, there really is no telling what it may hold.
Of course, we have some ideas.
Imagine a personal assistant that creates, keeps track of and informs you about important events and tasks at precisely the right moment.
You get important reminders and questions related to keeping a quick, five minute journal in Messenger or Slack.
Think about being able to connect FAQ and KnowledgeBase resources to a website chat widget, allowing vistors to ask questions, get answers and create new support tickets in real-time.
How great would it be to be able to search and deliver files as easy as sending a text message, or being able to create and manage projects entirely by guided software conversations with an AI heart?
This is not only the future we want to live in — we believe it’s the one we’re working towards as we speak. This is one of the many, many reasons why Software as a Conversation is such an important idea.
It’s also one that we need to keep working towards, together.
We imagine a future where the contact list is the new app drawer on your smartphone.
What is your idea of Software as a Conversation?