Blog

Share
Industry News

The failure of chatbots shouldn’t kill your faith in AI

by Lewis Leong on Sep 5, 2017

The chatbot era might be burning out as quickly as it started. With a failure rate of 70%, it’s clear that chatbots have a long way to go if they’re going to survive. This, as well as examples like Tay’s overt racism, and even further failures with sisterbot Zo have caused the tech press to declare that artificial intelligence as a whole has been largely a failure.

A PR problem for chatbots has turned into a PR problem for AI, but you shouldn’t let that kill your faith in AI’s possibilities. Chatbots are but one very small part of what makes up the entire field of AI, and it is far from the peak of its potential.

AI won’t be just another buzzword—it will be the foundation of future technologies. Let’s first agree on a better definition of AI, and then we can take a better look at where it is now and where it’s going.

A problem of definition

Chatbots are only one small part of the larger “Artificial Intelligence” umbrella.

Artificial Intelligence is an umbrella that includes technologies such as machine learning, deep learning, natural language processing, and others. A product falls under the AI umbrella when it applies software-driven learning based on a collection of data that produces better and better results without any help from humans.

Over the last five years, progress in neural networks and the expansion of deep learning into wide-ranging practical applications such as language translation, machine vision, virtual assistants, and music production have created a renaissance for AI. That renaissance has attracted the attention of technologists and investors while producing multiple hype cycles for AI’s different applications.

Buzzy hype cycles tend to be associated with empty marketing buzzwords and vaporware thanks to over-hyped flops like mobile wallets and big data. With AI, that couldn’t be further from the truth, as the industry already has some proven successes.

Today, AI encapsulates popular voice-powered assistants like Amazon’s Alexa, predictive algorithms like Netflix’s show recommendations, and smart services like Cogito, which detects human signals to guide customer service representatives in real time. These are just a few examples of what AI is capable of.

Chatbot issues aside, AI has the potential to be the next mobile

The way I see it, the innovation cycle is about 15 years. Every 10 to 15 years we get radical changes like the PC, Internet, and mobile, and the smartest people leverage new technologies right away. Zuckerberg made the right move at the right time with Facebook when he saw mobile’s rapid rise.

Right now we are in the mature phase of the mobile innovation cycle. AI will lead us to the next big wave of innovation. That said, AI might not be the revolution itself. It is the most interesting thing for investors, tech companies, and the public to explore, but we have to manage expectations because groundbreaking developments take time. Even though it was clear mobile was going to be a game-changer in 2009, it wasn’t that great at the time.

AI is absolutely over-hyped right now, but much of the hype will eventually come true. Self-driving cars, digital assistants that are actually smart, and cancer diagnoses predicted by AI are not that far away.

It’s said that hindsight is 20/20, but perfect hindsight doesn’t make a business successful. Although AI has been poorly defined, you shouldn’t let chatbots ruin your perception of it.

AI is an opportunity every business should be investing in. It has the potential to be as big of a breakthrough as mobile, and if you don’t ramp up your AI efforts, you might find yourself disrupted like many companies were during the mobile revolution. Start investing in AI now so that you can reap the benefits now and later.

Lewis Leong is AppLovin's Content Marketing Manager.

We’re hiring! Apply here.