Dive Brief:

• The AI hype has businesses clamoring to integrate its capabilities into their technology stack, but applying it in a valuable way isn't clearly understood for most companies. "A misconception is somehow AI is fairy dust that you can just sprinkle over your company and sprouts will happen and you'll get growth," said Peggy Johnson, EVP of business development at Microsoft, at a New Work Summit panel Tuesday.
• About two years ago, AI began cropping up in the presentations of every startup looking for investment, and while it can improve the entire product life cycle, just having AI in a business model will fall flat unless the appropriate solution is being harnessed, said Frank Chen, partner at Andreessen Horowitz, at the panel.
• With every AI application comes the discussion of job disruption. But people have been predicting the end of labor pools with every new automation technology, even though that usually hasn't been the case, said Chen. With the introduction of spreadsheets and ATMs, many predicted the end of accountants and bank tellers, respectively, but in reality these positions grew with the new technologies. Companies need to get out in front of affected jobs with education and retraining and go farther back to what kids are taught from the beginning, said Chen. 

Dive Insight:

Focusing on practical, useful applications of AI over more glamorous ones is critical to effective adoption of the tech.

Johnson and Chen described AI applications across the world, from cow insemination in Japan to delivering blood in Rwanda to sorting cucumbers. As a business tries to figure out how it wants to integrate AI into its technology stack, technology leaders need to ask, "What's the cucumber sorter in your organization?" said Chen. 

Whether its automation applications or workplace analytics, AI ultimately boils down to the data a company has. The first order of business is stepping back and understanding the kind of data a company is using and what insight will come from that, said Johnson.

Building up the necessary talent to implement AI is the next challenge.

The skyrocketing salaries and competition for higher level AI talent is well documented, but most companies can cultivate AI talent by retraining existing software engineers. After just a few weeks in a Udacity class, existing employees can begin making a huge difference in the space, said Chen.