I’m wondering out loud about the weather when my preschooler says, “Mama, you should ask Siri. She knows everything.”
“Gahhh! What the what?” is what the privacy fanatic and data scientist inside me shouts. Out loud, I choose to respond with the more neutral route: “That’s interesting. What makes you say that?”
“Because at school when you have a question, you can ask Siri and she knows all of the answers,” she tells me.
This is how I realized that bots are becoming part of my family’s everyday life, whether I like it or not. None of this connectivity happened overnight, but it made me think more about how we talk to kids of all ages about emerging technology and the issues surrounding them, like data privacy.
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
Bot, chatbot, or intelligent virtual assistant?
What we call bots have been around more than 50 years and started with basic chatbots programmed to interact through mediums like messaging programs and audio. The first bots were limited by costs of processing power and algorithms. Fast forward to today, where simple interactions are now real-time automations powered by artificial intelligence and natural language processing (NLP), both of which have advanced significantly over time.
Today’s bots can help you complete tasks as complex as booking hotels, responding to support questions, and organizing your meetings. The list keeps growing while new developer kits, plug-ins, and APIs (e.g., BotKit) make it easier to build more bots. In the tech world, the term “bot” appears everywhere. You can fill out your next Technology Buzzword Bingo card in one talk on bots alone. There’s a wide range of interpretation here, so let’s explore what bots are and what they are not.
When you say bots, you might mean: Inner bots (chatbots)
There is a set of bots designed to work inside products or services.They’re essentially “digital users within a messaging product” because they use conversational interfaces inside services. For instance, while you’re in Slack, Facebook Messenger, or Skype, you can ask for alerts or query specific information. Even though you’re talking with a computer program, not a person, you’re finding information in a conversational way without leaving the app or product.
When you say bots, you might mean: Intelligent virtual assistants
Intelligent virtual assistants deliver botlike services but are more sophisticated than bots. Typically, when there’s a question-and-answer or task-based computer program driven by AI, this is considered an intelligent virtual assistant. Think Google Assistant, Amazon Alexa, Apple Siri, or Microsoft Cortana. Intelligent virtual assistants can order products, respond in a conversational way using voice, and offer predictive intelligence. They’re cropping up in health care, retail, and IT support scenarios at a rapid pace.
On the backend, Intelligent virtual assistants are a combination of voice recognition, speech recognition, and text-to-speech, with the ability to affect your applications and the world around you. What differentiates them from bots is they perform many functions, acting as the interface and the fabric that holds your bot world together. They form the foundation of digital interactions with intelligent conversational interfaces.
When you say bots, you don’t mean: Robots
Most robots are considered high-functioning machines capable of operating autonomously or semi-autonomously for complex tasks in challenging environments. This is unlike bots and intelligent assistants, which are designed to work within a specific domain of queries and responses or to perform a task, like telling you what time your flight departs.
There’s still debate about the exact definition of robots. Generally speaking, robots involve mechanical, humanoid, and programming aspects, with some ability to move, perceive, and exhibit intelligent behavior. They’re built to perform complicated functions like medical procedures, patient assistance, and manufacturing roles that are dangerous or repetitive.
How do bots know what they know?
Here is what you and the kids need to understand: Bots are only as good as the information given to them.
How do they learn more? We give them more data to help shape their view of the world. Like kids and other people, they learn through interactions with users, developers adding more data sources (like information from weather services and sensors outside your house), and algorithms that make predictions about things they don’t know.
Do they know EVERYTHING? No.
Could they know everything? Depends on your definition of “everything.” They still have a limited ability to learn over time using data and computer programs. As they collect more data, computing power increases, and machine learning algorithms get more powerful, bots will continue to gain sophistication and usefulness. They are learning about the world around them because we continue to provide avenues for them.
More bots, more data privacy issues
The issue of data privacy is central when it comes to bots and the data used to “educate” them. As bots become more ingrained in daily life, and as AI and ML become more sophisticated at capturing and storing data, it’s important to stay informed. Some very smart people are working on data privacy in your home. You can learn more about their work from the Data & Society Privacy Initiative and Federal Trade Commission Coppa FAQ.
AI and ML capabilities are advancing at a phenomenal rate. On one hand, this raises concerns about data privacy and the line between humans and machines. On the other, this introduces a new universe that could speed our ability to share information and expand our own knowledge. How we balance the opportunities and risks remains an ongoing debate among both researchers and teams working to create real-world applications with these developing technologies.
So what did I tell my daughter about her idea of an omniscient Siri? Because I’m an engineer and a data scientist, my daughter has a kidlike understanding of a few concepts and a complete fascination with robots. We talked about how Siri was like a robot but on a computer.
“Teams of people write code that uses data to help understand your question,” I explained. “Then Siri decides the best way to try to answer your question.”
We talked about how a bot isn’t always right and how sometimes that’s really funny. We talked about how a bot gets smarter through more users and data. She walked away with the idea that bots are cool tools, dependent on smart, curious people like her. #ParentingWin
Amanda Casari is a data scientist and senior product manager on the Concur Labs team, where she collaborates with a team of engineers and fellow data scientists to incubate new technology ideas, interfaces, and prototypes designed to change business travel and expense management as we know it.
