The world's biggest tech companies are investing heavily in artificial intelligence (AI): software that can learn to think and solve problems in a human-like way.
According to Deloitte: "By end-2016 more than 80 of the world’s 100 largest enterprise software companies by revenues will have integrated cognitive technologies into their products."
Each company takes a slightly different approach to making business processes smarter through the development and deployment of machine learning, or cognitive computing. Here we have rounded up some of the biggest names in the industry and what their approaches are to AI.
"The evolution of enterprise software"
David Schatsky, head of the trend-sensing program for the US innovation team at Deloitte said at the AI Summit in London earlier this year that the big victories in artificial intelligence over the last couple of decades have all been in games, from IBM's Deep Blue mastering chess in 1997, to Deep Mind's AlphaGo beating Lee Sedol at Go this year.
Now he asks: “How do we translate these undeniably important achievements to create value in business? What we have realised so far is that it’s a journey and we are going to build it together.”
Dave Elkington, CEO at Insidesales.com, a tech unicorn that uses predictive algorithms to help sales teams close more deals, sees data as the key to opening up the possibilities of AI in the enterprise.
“The data is what is most interesting," he says, "and most people incorrectly think it is the algorithm. It’s about the data. It’s the evolution of enterprise software.”
One company that is a gatekeeper for all of this data in the enterprise is German software giant SAP. Markus Noga is responsible for introducing disruptive technology to existing networks at SAP, where he has been focusing on making its business applications more intelligent.
“Machine learning starts and ends with the data and we have a fantastic advantage with some of the most precious data sources in the enterprise world, and the way that data flows through our network,” says Noga.
SAP delivers its machine learning capabilities as APIs on its enterprise cloud and embeds them into existing applications, such as its travel booking platform Concur and its HR offering Success Factors. Concur processes $50 billion (£34 billion) of travel transactions every year and Success Factors is installed on 245,000 systems. That’s a lot of data to crunch.
SAP started out by creating simple invoice-matching and CV-matching applications, where computers learn to read and match documents, freeing up human workers from these mundane tasks.
Speaking to ComputerworldUK, Noga couldn’t resist taking a shot at the competition: “We look after real-world business problems. The boring things that matter to businesses. We’re not going to showcase it on things like winning a game of Go or Jeopardy. [Our focus is on] transactional problems that drive significant value.”
It is early days for SAP on its machine learning journey, though. According to Noga: “We have screened over 180 machine learning use cases. The first wave prioritised horizontal apps near our core and we are currently prioritising a second layer and all spaces that are of value to our customers, such as the Internet of Things (IoT) and analytics, specifically predictive capabilities.”
He adds: “We recognise that not every customer has machine-learning expertise, and they shouldn’t need to. This is why we have an application-led strategy and ready-made solutions for the most pressing business problems and they deploy like any application. So you don’t have to be an expert in how the app is designed.”
Paul Chong, director of Watson Group EMEA at IBM told the AI Summit in London: “We want to get to this stage where you simplify the use of the technology to the point where you actually put it in the hands of the business owners.”
IBM Watson was commercialised in 2015 and IBM has been working on bringing cognitive computing to business projects, with the likes of CitiGroup, Imperial College London, Under Armour and various healthcare organisations.
“We’re creating the cognitive platform to be the API economy of choice for you to build cognitive systems. We’re making this open, we believe there are a huge number of opportunities and I think about the number of clients we are working with and it is varied.
"Our hopes and aspirations are that the platform Watson, or others in terms of AI, will start to be infused through the API economy with all business processes. It won’t be just the client or for the decision maker, it will be open to all,” said Chong.
Adam Evans, CTO at cloud computing pioneer Salesforce.com, tells ComputerworldUK: “AI is technology that Salesforce is incredibly interested in. At Salesforce we have moved from systems of record to systems of engagement and now we are moving towards systems of intelligence.”
Salesforce already has machine learning and predictive capabilities baked into products like its customer relationship management (CRM) platform SalesforceIQ, as well as Marketing Cloud Predictive Intelligence and the Service Cloud Intelligence Engine.
Evans explains: “SalesforceIQ’s technology unifies customer data by leveraging data science to automatically capture, analyse and surface information and predict patterns, and then proactively recommends actions.”
Some of the biggest enterprise software vendors are suspiciously quiet on the AI front. Obviously there are predictive capabilities with Oracle’s database product if you have the right people in-house to leverage it, but the software giant has yet to launch any products or services with an obvious machine-learning layer.
Microsoft laid out its enterprise AI vision at Build 2016, and it seems like it is banking on bots, as well as investing in predictive capabilities to make its Cortana digital assistant even smarter through the Cortana Intelligence Suite for developers.
Richard Peers, director of financial services industry at Microsoft wrote: “It’s like Siri on steroids and, as a bonus, nobody gets to hear you clumsily and repeatedly speaking out loud into your phone like you’ve gone clinically insane. But we need quality bots.”
Speaking at Build 2016, Microsoft CEO Satya Nadella said: “As an industry, we are on the cusp of a new frontier that pairs the power of natural human language with advanced machine intelligence. At Microsoft, we call this Conversations as a Platform, and it builds on and extends the power of the Microsoft Azure, Office 365 and Windows platforms to empower developers everywhere.”
Speaking with ComputerworldUK, Dave Elkington at InsideSales says: “Qi Lu, the chief product guy at Microsoft is a friend of mine. He and I will once in a while make wagers over who can predict things more accurately. Running Cortana and his predictive stuff versus ours and once in a while, despite being a smaller company, we still give them a run.”
When Google spent £400 million on the little-known artificial intelligence startup Deep Mind the question of how it would enrich Google’s business was at the forefront of many people’s minds.
Co-founder Demis Hassabis told The Verge that Deep Mind's role is to “turbocharge” Google: “Of course, we actually work on a lot of internal Google product things, but they’re all quite early stage, so they’re not ready to be talked about. Certainly a smartphone assistant is something I think is very core — I think Sundar [Pichai] has talked a lot about that as very core to Google’s future.
Speaking to Wired last year, head of applied AI at Deep Mind, Mustafa Suleyman, said: "I've got five teams, working on YouTube, search, health, natural-language understanding and some Google X projects. We're working at applying the core engine that sits behind the Atari games player across the company.
“One is in YouTube-recommendation personalisation. We try to learn from the types of videos that lots of users are watching in aggregate to better recommend at the right time, in the right place, what they'd like. Then search: think of search as a process of querying an engine, browsing through links generated, then refining your query in this iterative feedback cycle. Over time we can improve results."
Away from Deep Mind, Jeff Dean, senior fellow at Google, said at its Cloud Next conference earlier this year: “Our goal is to create applications that can see, hear and understand.” At the same conference the search giant announced Cloud ML to help machine-learning engineers to build “sophisticated, large” models based on its recently open-sourced TensorFlow deep-learning library.
Not every company is Google, though, and buying the brightest minds in machine learning and giving them fairly free reign to turbocharge your already multi-billion search and advertising business is not a luxury most of the enterprise can afford. Time will tell if Google’s strategy of bringing AI into its enterprise will pay dividends, or prove to be an expensive, yet impressive, experiment.