July 24, 2019

Artificial Intelligence: What is it and how does it work?

Artificial intelligence is much more than just a trend. Within the current business environment, it’s turning into an amazing tool for automating processes and optimizing resources.

Thanks to this technology, companies can develop innovative strategies in different areas. They can make their employees’ tasks easier and improve their customers’ experience.

In fact, a study conducted by Adobe found that, today, just 15% of companies are using AI, and 31% said it is on the agenda for the next 12 months.

Despite its popularity, sometimes the artificial intelligence field can be intimidating, but that ends today. In this article, Juan Biondi, Head of AI at Aivo, clarifies concepts and answers questions so you may learn about or keep updated on key AI knowledge.

Juan is a computer engineer and leads research about natural language processing, machine learning, and deep learning applied to Aivo’s solutions.

His catchphrase?

“Artificial intelligence is wonderful, but there’s no magic in it.”

Can you briefly tell us what artificial intelligence is?

When we talk about artificial intelligence, the first thing that comes to mind is a machine that imitates cognitive functions of “intelligent beings,” such as reasoning, learning, or problem solving.

A more intellectual definition for AI is “the ability of a system to correctly interpret external data, learn from such data, and use that knowledge to achieve concrete tasks and goals through flexible adaptation.

”As opposed to what many believe (maybe sci-fi inspired), AI isn’t “general intelligence” yet; it’s still years away from that. It isn’t a system capable of solving any given task as a human being would (like the Terminator or the Bicentennial Man.)

AI as we know it today represents “narrow intelligence.” That is to say, it’s a system that is intelligent enough to complete a specific task. It doesn’t know how to do anything else besides the activity it was trained for.

What do we call “intelligence” when we talk about artificial intelligence? In other words, what makes technology “intelligent”?

In general, this technology is defined as “intelligent” because it can imitate what is understood as intelligent behavior. For example, it learns from experience and figures out how to solve a certain problem.

Additionally, in order to consider how “intelligent” the current systems are, we can take into account what Andrew Ng, a renowned artificial intelligence researcher, suggested as an “imperfect rule of thumb.” This rule states that “almost anything a typical human being can do with less than one second of mental processing can be automated with artificial intelligence.”

In the future, these machines’ “intelligence” will keep on evolving. In fact, 42% of AI researchers predict that, by 2030, artificial intelligence will be similar to human intelligence.

What’s the difference between Machine Learning and Deep Learning?

There are many ways of developing artificial intelligence. It’s a very broad field with different areas.

One of them is Machine Learning or automatic learning. It could be defined as a way of developing systems without specifically programming them to perform a particular task. In other words, a system is trained so that it learns on its own.

To do so, an algorithm is implemented, defining how the system must learn to perform a task. Then, it’s trained through examples. Telling the system which features it has to keep in mind for task completion is essential. That’s how it can make decisions based on the established information.

Furthermore, Deep Learning is a branch of Machine Learning that uses algorithms capable of “realizing on their own” what the important features in a group of raw data are, based on the context.

Do you think it’s essential for companies to add AI technologies? Why?

I wouldn’t say it’s essential, but it is rather convenient. Artificial intelligence can make many processes easier and faster and can considerably reduce costs, errors, and delivery times.

Mainly, it lets us focus on more complex activities and leave repetitive tasks to autonomous systems. For example, in the Customer Service field, we can let an intelligent bot answer most of the inquiries autonomously. This way, human beings will only address situations that call for human service.

The advantages are many: 24/7 availability, fast troubleshooting, less customer effort, and greater security and privacy. Also, thanks to AI, interactions can be customized for every customer type based on where the conversation is going.

What’s your advice for a company that has never used artificial intelligence and wants to add it to their Customer Service strategy?

Some of my suggestions for someone that’s considering adding this technology are the following:

1. Do not use AI just for the sake of innovation. Its implementation has to be based on a strategy that is aligned with the objectives and mission of the business as a whole. Moreover, it has to solve problems.

2. Not every company has the ability or the resources to develop their own AI. In this case, it’s advisable to seek support from AI solutions partners and vendors that already have this technology developed and ready to use in different cases and industries.

3. It’s advisable to understand from the beginning that AI won’t solve every problem or replace employees. This technology has to work together with the teams. There are certain human features and abilities, such as creativity and sensitivity, that allow for solving more complex problems. None of them can be replaced by a machine. What AI can do is widen human beings’ possibilities by suggesting automatic responses, offering shortcuts, collaborating in troubleshooting, and assisting with daily task performance.

4. For effective problem solving, AI needs a fair amount of high-quality data. These must include enough information in order to be able to answer questions and solve problems. Moreover, they need to be diverse so that they can represent as many situations as possible and thus train the system.

How do you imagine technology in the next 10 years? What do you think will be incorporated into daily life that doesn’t exist yet?

I’d say artificial intelligence is already part of our lives. If we think about it, today we have virtual assistants like Alexa or specialized chatbots that answer the majority of our questions.

I think that in less than 10 years AI will be so internalized within our lives that it’ll be natural, everyday stuff.

The current uses are already in diverse areas and are almost infinite. Many of these uses are implemented for user entertainment. This is the case of Affine Layer, which creates a photorealistic image from a doodle.

Others contribute to more data generation in areas of higher impact, such as healthcare and the entertainment industry.

In fact, the technology used to develop “entertainment” apps such as FaceApp is the same used to detect and identify different types of breast cancer, for example.

In that sense, the future will be inhabited by artificial intelligence. Autonomous cars, household chore robots, and virtual assistants are already part of our lives and, in the future, they will be even more present in our routines.

How does Aivo implement Artificial Intelligence in its solutions?

Artificial Intelligence is one of the main aspects of our value proposition. Our product Suite, that includes chatbot, voice, live chat and knowledge solutions, assists companies in customer service and helps them provide a great customer experience using AI.

Overall, we use AI for conversational purposes, language understanding and empathetic interactions. With it, we make sure that our solutions understand the intention behind the questions of the users.

For this, among other things, we use traditional NLP, Deep Learning based on “embeddings” with LSTM networks and different types of models of Machine Learning. This technology is useful not just to understand customers, but also to assist human agents when they are interacting live.

Thanks to AI, our solutions allow companies to interact with their customers automatically, on a deeper level, with natural language, and at all times. If you’re curious about this, here you’ll find more about how our AI works.

And what do you think about artificial intelligence?

With so much information, I imagine you have tons of questions.

Don’t be shy; give us your feedback. We love to talk about these topics (about all topics, actually).

See you soon!

Are you looking for new ways to improve your CX?

Our customer service solutions powered by conversational AI can help you deliver an efficient, 24/7 experience  to your customers. Get in touch with one of our specialists to further discuss how they can help your business.