How to generate the needed expert data?
To set up effective artificial intelligence software, a company needs a lot of data. But what happens if you don’t have that kind of specific information about your area? A Seattle-based startup, Mighty (formerly known as Spare5) will get it for you by paying subject-matter experts to spend a few minutes answering questions or performing tasks. Some examples include finding golf aficionados for IBM, people who can describe a photo for Getty Images, and radiologists or technicians to read tumor scans.
Race for data
The company, which renamed itself to reflect a focus on training tasks, is adding three new investors as part of a US$14mil (RM62.46mil) funding round: Intel Capital, Google Ventures and Accenture Ventures. It’s unveiling partnerships with Intel and Accenture, too. Spare5 was spun out of Seattle-based VC Madrona Venture’s labs in 2014, a mobile rival to Amazon’s Mechanical Turk program, which finds workers for tasks online. Around the same time, Getty, reached out to get help categorising images in its collection. Like Getty, customers increasingly wanted Fives – the people that take on the “microtasks” – to perform brief activities that train algorithms. So Spare5 refocused and renamed itself around that idea.
Getting data about Golf
“There’s an arms race in training data” for , said chief executive officer Matt Bencke. IBM, with its focus on Watson products, wanted to create a chat bot for spectators at the 2016 Masters golf tournament it sponsors. Using tablets on site or their own mobile phones, golf fans would be able to ask the bot questions or banter with it. The only problem? IBM couldn’t find enough annotated golf-related training data. So IBM sent Mighty a large body of information culled from the web that it thought was related to golf. Mighty found workers familiar with golf, had them tag information specific to the sport and compose questions and answers based on the material. That data became the basis of IBM’s Watson golf conversational agent […]