Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

5 trends in AI that shaped 2018

2018 was a big year in the world of AI. Here we look at five of the leading trends in the use of AI to see how they set the stage for what should be further gains next year.

With 2018 quickly fading in the rear view mirror as we drive ahead (autonomously?) into 2019, this makes a great...

time to take a look at some of the big trends in AI, from use cases and implementations of AI to machine learning developments that propelled the industry forward. 2018 was a banner year for AI in many ways, and it seems that 2019 will continue the pace without slowing.

Autonomous vehicles got real

2018 will stand out as a big year for autonomous vehicles. Going into 2018, Uber and Google's Waymo self-driving vehicle unit were facing each other with lawsuits over intellectual property. However, just a few months into 2018, the two companies settled their differences, and proceeded to make autonomous vehicle industry news perhaps in ways not intended. Uber's self-driving vehicle incurred the industry's first fatal autonomous vehicle accident, causing the company to put on hold its self-driving initiatives. Other autonomous vehicle units soon followed suit putting their autonomous vehicle programs on hold. By the end of the year, Uber was back on the roads with self-driving vehicles, but in manual mode -- it's unclear how that makes the vehicles autonomous, but baby steps are warranted here. Uber just this week admitted some blame for the fatal crash through lack of oversight.

On the other side, Google's Waymo unit announced the availability of commercial self-driving rides under the Waymo One brand. Perhaps the future is finally here with self-driving vehicles, but early experience is already showing that full autonomy for vehicles might still be a distance away. Regardless of the industry's fits and starts to making autonomous vehicles happen, it's clear that 2018 was a watershed year in pushing the industry forward. There's no doubt that 2019 will also be a year to watch for self-driving developments.

Voice assistants everywhere

Voice assistants played a leading role in another of this year's leading trends in AI. Those devices got a big visual boost this year with Amazon, Google, Microsoft, Apple, Samsung and others releasing devices in a wide range of colors, sizes, features and capabilities. Amazon is pushing their video capabilities while Google and Apple are battling for sound quality and other consumer-focused features. But what we really care about is their intelligence. Research firm Cognilytica released a benchmark showing that all of these voice assistants are lacking in intelligence. Amazon soon after announced a feature that will let Alexa get back to the user with questions it was not able to answer.

Of course, Google proceeded to make things even more interesting and creepier with their demo of Google Duplex. The company intended to demonstrate the evolved sophistication of intelligent assistant technology by having it interact in a prerecorded, simulated live interaction with a real person. The problem is that people don't like being fooled about whether or not they are talking to a bot. Despite the concerns, Google announced plans to roll out Duplex to a limited number of users by the end of 2018.

AI goes money crazy

As much as investors were interested in AI in years past, venture capital went absolutely bonkers for AI companies in 2018, funding many companies at eye-watering levels and making this one of the most important trends in AI. In the U.S., AI companies raised billions of dollars with companies such as Dataminr, Cylance, Pony.ai, Automation Anywhere and others instantly becoming so-called venture unicorns, with valuations over $1 billion each. However, while the U.S. VC market always likes a good frothy bubble, what makes the AI space interesting is that U.S. VC investment in AI was actually overtaken by Chinese investment. Chinese AI companies raked in ludicrous amounts of investment money with companies like SenseTime, Yitu, ByteDance and others raising billions of dollars each, mostly from government sources. It's clear that investors and governments alike are seeing big strategic opportunities putting their money into AI-related startups.

On the flip side, 2018 also saw some notable business failures in the robotics realm of AI. Rethink Robotics, led by notable AI leader Rodney Brooks, was forced to shutter its doors after being unable to raise necessary capital. Similarly, Mayfield Robotics and Jibo both shut down operations in 2018 after failing to find customer demand for their offerings. On the one hand, money seems to be flowing to AI startups, and on the other, money is not readily available for robotics companies. We'll have to see where the wind blows in 2019 for venture funding of AI-related companies.

Governments turn strategic with their AI plans

A little over a year ago, China released details of a three-step program outlining its goal to become a world leader in AI by 2030. This year, their initiatives made it clear that the nation sees AI as both a competitive and strategic imperative. The U.S. has doubled down on its strategic AI planning with the Pentagon announcing this year its plans to invest $2 billion in AI-related initiatives. Meanwhile, European countries are also betting big on their AI initiatives. France's President Emmanuel Macron released a national strategy for AI this year with plans to invest more than €1.5 billion (about $1.7 billion) in AI-related research over the next five years. Germany announced the creation of Cyber Valley, a new tech hub region in southern Germany hoping to create new opportunities for collaboration between academics and AI-focused businesses. Even the United Arab Emirates and South Korea have announced major AI initiatives. 2018 was the year of the country-level AI strategy.

Continued widespread adoption of AI in enterprises

But of course, the biggest news of all in 2018 is that AI, machine learning and cognitive technologies continues to gain widespread traction in enterprises across a range of industries and use cases. 2018 was the year that chatbots became pervasive in customer service, and AI-enabled IT self-service management matured. 2018 was also big for predictive analytics companies embracing machine learning and becoming AI-powered analytics companies. We saw companies develop advancements in natural language processing that gave enterprises more visibility into their reams of unstructured data.

AI-enabled marketing was also one of the top trends in AI in 2018, with the concept of hyper-personalization gaining steam and companies realizing the benefits of augmented intelligence in a marketing context. We also saw enterprises using AI and cognitive technologies across a wide range of use cases, from internet of things and edge devices powered with AI- to machine learning-based content intelligence systems. In 2018, retail also jumped on the AI bandwagon with major AI-enabled commerce systems going live, including Amazon's plans to launch over 3,000 Amazon Go stores in the next few years.

AI continued to show its strength in 2018, and certainly shows no signs of slowing as we make our way to 2019. We're looking forward to an AI-powered 2019 with even more big news and trends on the horizon.

This was last published in December 2018

Dig Deeper on AI technologies

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

What do you think were the most important trends in AI for 2018?
The emergence of chatbot was the most important trend for me as small businesses would opt for a bot to be included on their website to hiring a new customer representative.
In my context many companies are now experimenting with AI. Cognitive chat bots are a relatively easy first application, but we see a growth in the use of AI associated with logistics, marketing, client relations and even financial decisions. Very little yet related to Human Resources or accounting.
The biggest trend we saw at Indico was the transition of AI from the "what" to the "how".  At the start of the year, many prospects were reaching out with an AI agenda, rather than a particular pain point.  By the end of the year we saw this mostly inverted and lots of great discussions around AI as enabler, specifically in our case of back office, document-intensive processes that AI can automate.