The Merriam-Webster dictionary defines artificial intelligence (AI) as "An area of computer science that deals with giving machines the ability to seem like they have human intelligence," and offers an alternate definition of "the power of a machine to copy intelligent human behavior."
The first AI programs were developed in the 1950s and weren't commonly used in business. The big data analytics revolution has finally taken AI out of the halls of academia and research institutions, and has plugged it into commercial applications for business. Commercialized AI is still fundamentally new for many companies, though. The keys to business success with AI are to know how to use it and know what results to expect.
Here are four examples of AI systems that are working for businesses.
1: IBM Watson
There are few AI engines more finely tuned for commercial success than IBM Watson. Since its television debut on Jeopardy in 2011, IBM Watson has wowed users and is currently at work in a variety of applications and industry sectors, including finance, law, and transportation (PDF). In healthcare, Johnson & Johnson and Sanofi are using IBM Watson to identify new applications for drugs and to research through scientific papers that detail clinical trial outcomes. IBM Watson is also assisting the Cleveland Clinic with medical diagnoses.
The business value proposition: IBM Watson is an adaptable application of AI with an established track record. If you have a clear vision of where you want to implement AI in your business, IBM Watson can probably deliver it.
2: Robotics in the warehouse
The job market gained momentum after the economic downturn of 2008, and with it, warehouse managers were challenged to compete for new, young workers. Consequently, many workers in distribution centers and warehouses are older employees. Walking on miles of warehouse concrete floors each day to pick items for orders is hard work for older workers, so warehouse and distribution center managers have found that robots can do some of this legwork. Worker fatigue gets addressed, and order picks and fills are being expedited, which means companies can earn more revenue faster.
Amazon is a classic example of warehouse robots at work. Amazon uses Kiva robots to do the walking and the order picking, while the humans that work in the distribution center can stay in one place and process more orders faster.
Hitachi has just come out with a warehouse robot that grabs goods with two arms and places the items into shipping containers. These warehouse robots are ideally suited to repetitive packing of the same item. For online warehouses with high volumes, the robot can speed pick, pack, and ship times and also save human workers from highly repetitive and fatiguing work.
The business value proposition: If you are a fast-paced ecommerce retailer and you want to improve the speed of order picks and fills and/or you are a warehousing operation with high demands and an older workforce, robots can save footsteps and get orders out the door faster.
3: Predictive selling
Recent research from Salesforce (PDF) reveals that there will be a 58 percent increase in the use of sales analytics between 2015 and 2016, and that within the next 12 to 18 months, 74 percent of sales leaders plan to use sales analytics. As online consumers, we see predictive sales analytics all the time with etailers offering us sales and promos on books, recordings, clothes, and other items similar to what we have purchased in the past.
There are many choices in the commercial market for predictive sales analytics, but Salesforce stands out, given the company's established sales software base in business. Since Salesforce is cloud-based, entry price points do not require purchasing on premise hardware, and therefore enable small to midsized companies to afford the application.
The business value proposition: By getting to the consumer first, you can capture a greater portion of each consumer's potential spend, and grow overall revenues for your company.
4: Fraud detection
In 2014, fraud cost US retailers $32 billion, so AI that can provide fraud detection by analyzing each card user's typical usage patterns and spend ranges and flag any unusual aberrations is indispensable not only to retailers, but to banks, credit unions, and any other businesses that accept credit cards.
Major card processing firms such as MasterCard and Visa include fraud detection AI with their services. There are a number of independent fraud detection software solutions that are available to businesses; leaders in the market include FICO, ACI, NICE, and SAS.
The business value proposition: If your company is losing money through credit card fraud, you should do something about it. Fraud detection AI can give your company predictive and real-time analytics, putting your company in a better position to spot fraud.
Start with a business case before buying into AI
The businesses enjoying the greatest success with AI are those that have identified ways that AI can help them strategically and operationally before they invest in a solution; this is why purchasing any AI solution should always be driven by a business case. You should have a set of predefined key performance indicators (KPIs) or metrics to be discussed with vendors in advance of purchase to confirm that the AI solution envisioned can help your company meet its goals.
- AI, Automation, and Tech Jobs (ZDNet/TechRepublic Special Feature)
- Executive's guide to AI and the future of IT jobs (free ebook)
- After medical school, IBM's Watson gets ready for Apple health apps
- Coming to a factory near you: Man-machine hybrids
- Mark Zuckerberg on how Facebook's AI will be "better than humans"