The idea of imbuing our machines with smarts has been around since the 1950s. After some fits and starts, a breakthrough in 2012 kickstarted the current flood of new research and interest in artificial intelligence (AI). Much of the work has been done at a handful of world-class universities and big tech firms.
And more companies are catching on—all with the goal of improving customer experiences and business outcomes. Meet five.
Greening With AI: Microsoft Our planet is facing a host of environmental challenges, from decreased water supplies and deforestation to air pollution. To help combat these issues, Microsoft launched AI for Earth in 2017, a five-year, $50 million program that puts software tools and cloud access in the hands of researchers working in the AI and environmental space. The initiative has doled out 242 grants to support projects such as PAWS, a program that uses machine learning to predict where poachers will strike, and SilviaTerra, which uses AI to monitor forests for wildfires.
The program is more than just being a good steward of the environment, according to Lucas Joppa, chief environmental officer at Microsoft. “It’s the right thing to do for the planet, and it’s also the right thing to do for the business,” Joppa told CMO.com.
When AI can work in the field and handle messy, real-world testing, it bodes well for more controlled office uses of the technology. In addition, putting AI in the hands of a wide range of people around the globe is also good for user testing. “Anytime we can stress-test the technology, the better,” he said.
It’s not just Microsoft that could benefit from AI either. Research commissioned by the company looked at the economic impact of AI on four environmental sectors—agriculture, water, energy, and transport—and found that employing AI could contribute $5.2 trillion to the global economy by 2030.
“Companies and countries that embrace the advent of the AI era, and sustainable economies, are set to gain most from the upcoming changes and opportunities,” the authors wrote in the report.
Report Cards For Cities: Citibeats Modern cities have a tough balancing act: trying to rely on more data-driven solutions while keeping citizens in the decision-making loop. Citibeats takes the best of both worlds by collecting citizen-generated content and analyzing it using machine learning. Citibeats’ dashboard gives city managers and other officials a way to monitor what’s being said in real time, and sends automatic reports and alerts on issues they are keeping an eye on.
The Barcelona-based company does this by taking content from social media, news sites, blogs, review sites, and other open data sources to keep up with what people are saying about their cities. Using natural language processing (NLP) technology, Citibeats can discover the issues that people care about the most or identify new problems. It works for every language, even learning slang, leaving no voice out of the discussion.
So far, Citibeats’ tool has been used in places like Buenos Aires and Nairobi to monitor how well a city is doing in reducing poverty or how confident people are with local financial institutions. The tool also uncovered an interesting insight for Cork, Ireland. Citibeats found that in Cork, locals outnumbered visitors at “tourism” sites, like the harbor and English Market, by 3-to-1 in January, making the case to offer more activities for residents in these places, as least in winter time.
More Art, Faster: Vasjen Katro While crunching numbers and analyzing data points may not sound very right-brain, AI has a lot to offer the creative world. Vasjen Katro, aka Baugasm, is an Albanian artist who turned to AI as a new technical paintbrush for his palette.
“So many tools, so much stuff I want to do, but there’s not enough time,” he said in a video.
Yet Katro made time to test out what AI could do: Specifically, he tried out style transfer, a way to have one piece of art look totally different, such as make a modern painting look animated or a line drawing look like a watercolor. Style transfer is usually a painstaking process, but with AI, artists can experiment with distinct appearances almost instantly. In testing it out, Katro began with some of the ways he would typically approach a project, taking photos of things like pavement and leaves, as well as creating acrylic paintings that he filmed through a prism. Then, he turned to style transfer technology to try out different looks, including changing his acrylic artwork to look as if someone painting with thick van Gogh-like brushstrokes created the artwork.
From speeding up complicated processes to finding new inspiration to build from, plenty of artists are looking to use AI in creative ways. And the new approach could be lucrative. The first piece of AI-generated art to sell at Christie’s went for $432,500—over 40 times the initial estimate. AI offers promise for collaboration and co-creation, but that raises questions of how to determine ownership.
<Making Shopping Tech-Forward: Vue.ai The fashion industry has always been one to adopt technology, so it’s no surprise that style mavens are starting to embrace AI. That includes Vue.ai, which uses a cutting-edge AI technique calledGANs that can create media-like images and videos to design digital models. Its tool, called VueModel, uses AI to dream up models in realistic poses with a variety of skin tones and body shapes in a fraction of the time it would take to do a traditional photo shoot.
“It's essentially having designer humans made by AI for the right reasons, like representation across different body types, race, styles, and more,” said Ashwini Asokan, co-founder and CEO of Vue.ai.
The startup already has convinced brands including Macy’s, Levi’s, Diesel, Thredup, and Tata to become customers. And more developments are in the works—in April, Vue.ai raised $17 million to continue refashioning the industry.
Talking To Our Phones: Aiqudo Instead of using hands to control our technology, companies are banking on voice commands to be the future of interfacing with our expanding quiver of devices.
By 2023, 8 billion digital voice assistants are projected to be in use, up from 2.5 billion assistants at the end of 2018, according to the U.K.-based analysts at Juniper Research. So far, though, most voice assistants, such as Amazon’s Alexa or Apple’s Siri, only work within their own ecosystems. Aiqudo wants to make every app on a smartphone controllable with voice commands, regardless of what device a user might have.
To achieve this, the company has gone through thousands of mobile apps and defined a number of commands, called “actions,” that can be done in an app, such as telling Spotify to play a specific song or getting WhatsApp to message a friend. For users, all they have to do is download the Aiqudo app—which launched last year in seven languages (English, Spanish, Portuguese, French, Italian, German, and Russian)—and then speak naturally to any of their apps. The NLP technology in the background is built to understand the various ways a command might be said.
Aiqudo plans to sell its voice tech to other companies developing apps and wanting to add voice commands without having to create different tools to integrate with Alexa, Google Assistant, and other voice assistants. It’s also relevant for manufacturers, from cars to phones, that want to build voice controls into their products.
The app is available for any Android phone, with an iOS app coming later this year. The company also plans to roll out languages spoken across Asia and India later this year, to sync up with dozens more apps that are ready and waiting to be told what to do.
“Voice should be completely intuitive,” said John Foster, the CEO of Aiqudo.