«  Return to all blog entries

Is Artificial Intelligence the Future of Business Intelligence?

There is a lot of buzz these days on artificial intelligence (AI) in business intelligence products. At a recent tweetchat of my #BIWisdom group of users, vendors and consultants, a participant asked: Is AI becoming a part of everything we do, the way analytics have been added to apps of all shapes and sizes across the business and personal space?

Although companies tend to overreact to emerging technologies, AI is not a new technology.  One of the tweetchat participants commented that AI is crossing a chasm, either in awareness or approach and that perhaps its evolution to this point is due to the prevalence of data scientists. Another member of the group agreed, tweeting that the data scientists craze has helped more organizations embrace machine learning and AI to get deeper insights. He added that analytics doesnt give enough deep insights.

Learn more about this topic at our upcoming Real Business IntelligenceTM conference, July 11-12, 2017 on the campus of MIT. Keynote speakers include Dr. Cathy O'Neil author of the recent book "Weapons of Math Destruction".

As our discussion progressed, I tweeted that I question whether AI is mature enough for generalized approaches in the business intelligence space.

Someone in the group asked, If all organizations see views of the same information, where will a competitive advantage come from? My AI is better than yours? His observation reminds me of what happened with ERP  if everyone adopts it, then everyone has the same benefits and limitations.

As with any technology, there are pros and cons. One the people attending the session that day tweeted a concern: I would hope organizations using AI for a long time, such as NASA, have developed some standards and guidelines.

Others discussed concerns about whether vendors should embed AI into their BI products. It could result in users thinking of AI as BI and taking it for granted, someone tweeted. Another in the #BIWisdom group countered that view with an opinion that vendors making AI invisible to users would help increase adoption. An observation from someone else was that some vendors probably will differentiate their products by not labeling AI. And another concluded that maybe it doesnt matter if users know. That was followed with another tweet: AI is a way to give users something they need but didnt know they wanted; its super useful for novice BI users.

Learn more about this topic at our upcoming Real Business IntelligenceTM conference, July 11-12, 2017 on the campus of MIT. Keynote speakers include Dr. Cathy O'Neil author of the recent book "Weapons of Math Destruction".

I brought up another concern: AI has the capability to transform industries. So how can organizations audit AI to make sure its working properly? Does this make data scientists even more important with responsibility for monitoring and auditing algorithms? And there are legal complications; so who gets sued when AI fails, especially if vendors embed AI in their products?

The group responded with some interesting tweets:

  •     Future accountants will study auditing and need to get AI auditing attestation.
  •     The regulations around AI would be fascinating. We're still struggling with net neutrality, so I cant imagine having government regulations around AI.
  •     Would AI algorithms be able to think beyond a regulation?
  •     Would the regulations be black and white, so to speak? Would they be in the form of guardrails or more like safety nets?
  •     This is why you need data scientists rather than accountants. Accountants think in black and white.
  •     Which business unit would develop and maintain the AI?

Bottom line: In my mind, all the concerns about AI come down to determining the goal of artificial intelligence in BI products. Will it be used to guide users, augmenting human capabilities? Or will it be used to replace them? 

The answer depends on whether artificial intelligence is better suited than humans for doing data analysis. Its important to keep in mind that artificial intelligence and cognitive machine learning change the game. These technologies can go beyond just enabling business intelligence; they can advise people. Arguably, AI wouldnt have bias or private agendas like humans, so it could be totally objective. But in reality AI could be trained to be biased. It could inherit the biases of a programmer or the entity funding the BI initiative. This could be an issue where AI augments human analysis by recommending new data sources and helping with metadata. The old maxim, he who has the gold makes the rules, is a key consideration. Already we have examples of differences is recommending data sources: Amazon uses Alexa, Apple uses Siri, Google uses Google Now and Microsoft uses Cortana. 

As organizations embrace artificial intelligence more and more to gain deeper insights, will those insights be objective or have built-in bias? As AI is crossing a chasm in awareness, organizations must keep in mind that it is not a mature technology and they should consider the impact of bias, just as they would be careful with bias in advice from humans.

Learn more about this topic at our upcoming Real Business IntelligenceTM conference, July 11-12, 2017 on the campus of MIT. Keynote speakers include Dr. Cathy O'Neil author of the recent book "Weapons of Math Destruction".

Howard Dresner is president, founder and chief research officer at Dresner Advisory Services, LLC, an independent advisory firm. He is one of the foremost thought leaders in Business Intelligence and Performance Management, having coined the term “Business Intelligence” in 1989. He has published two books on the subject, The Performance Management Revolution — Business Results through Insight and Action, and Profiles in Performance — Business Intelligence Journeys and the Roadmap for Change. He hosts a weekly tweet chat (#BIWisdom) on Twitter each Friday. Prior to Dresner Advisory Services, Howard served as chief strategy officer at Hyperion Solutions and was a research fellow at Gartner, where he led its Business Intelligence research practice for 13 years.

Signoff ac76b40517b7cc2ba93d8063981d746be69ffc8e8ae97f4cd71e2938514c56aa