artificial intelligence vs. intelligence amplification


AI vs IA: The Intelligence of Amplification

March 15, 2017

Analytic Partners’ Senior Vice President Mike Menkes looks at a quiet phenomenon in analytics computing that is reaching fruition before it’s more famous cousin.

Artificial Intelligence is a headline-grabbing phenomenon that may evoke either images of evil robots making humans obsolete, or utopian dreams of AI solving all our toughest challenges.  And in recent months, AI has garnered growing attention, having been incorporated as an element of Google Translate and pushing boundaries in IBM’s voice recognition technology.  But despite an enormous amount of noise, the reality of this type of technology continues to trail well behind the hype.  As a recent article in The Atlantic put it “In most cases, the systems making claims to artificial intelligence aren’t sentient, self-aware, volitional, or even surprising.  They’re just software.”

The fact is, AI faces limitations in many areas of application.  Full Artificial Intelligence relies on millions of data points: the “ingredients” it requires before it can begin to think for itself.  This makes it more effective for static systems with nearly unlimited data or insights to leverage.  So Artificial Intelligence can play chess, and now poker too.  But highly dynamic systems like marketing?  While there has been a lot of buzz and plenty of investments with an eye on the future, at present, it’s just not there yet.

What’s actually more interesting right now is AI’s less famous cousin -- what’s referred to as Intelligence Amplification, or IA.

No Artificial Ingredients

The germinating evil robots may not like to hear this, but there are researchers today who feel that IA offers a promising model for partitioning artificial intelligence as a powerful aid to human decision-making, rather than an independent decision maker.  Intelligence Amplification is programmatically less onerous and more attainable in the here-and-now than Artificial Intelligence.  In many cases its smaller footprint alone makes IA a better candidate for adding value.  And, when it’s in the form of an embedded expert database that delivers context-sensitive tips, it can also be organic to the workflow.

Here’s a quick example:  In recent years, we’ve seen a profusion of data visualization tools flood the market.  An inexperienced user can now easily deploy a hundred sophisticated charts from a drop down menu.  But is the most colorful choice necessarily the best one to communicate the story told in your data?  What if that visualization software could display a few tips during the design process?  “Given your data, we suggest you use one of the following four charts…”  That would be like having an information or UI designer as a free consultant looking over your shoulder.

That “free consultant” is an example of Intelligence Amplification (IA).  That is, the integration of human experience within the architecture of a software system in a way that can be selectively leveraged by the user.

Marketing Analytics and IA

Few things have changed more dramatically in the last decade than marketing.  Alongside all these changes, quantum advances in analytics have raised the demand for pinpoint measurement in areas where gut feel was formerly the rule.  But there is no analytic tool on earth (or on the drawing board) that, by itself, can account for the profound and continuing changes we are seeing in marketing.

Today, the most successful marketers have internalized the old and tested marketing dynamics and use them as a foundation to explore and internalize the new world.  Marketers, of course need to know how marketing works, but also need to assess data quality and filter out the noise.

Which brings us to a principal requirement for IA systems: you need to possess intelligence before you amplify it.

The ROI Genome

At Analytic Partners, we’ve brought IA theory into practice with the ROI Genome.  We believe analysis performed and informed by a more comprehensive understanding of marketing and business performance is crucial to measurement success.  The ROI Genome is a growing and accessible trove of marketing wisdom and knowledge based upon our collective experience as a firm.  Analytic Partners has built an incredible database of marketing response information spanning a wide cross-section of industries and marketing tactics across almost 50 countries to go beyond benchmarking – and we’ve gone the extra step to embed this intelligence within our GPS Enterprise analytics software suite.  Now our clients and modelers have the help of customized benchmarking, tips and tricks to help make intelligent and informed decisions as we analyze and consult for our clients.

So as our human experiences accrue through the changing markets ahead, the ROI Genome becomes an increasingly rich resource… with no artificial ingredients.  Artificial Intelligence may be the future of analytics, but in the meantime, we are accumulating and utilizing the intelligence today to feed tomorrow’s AI algorithms.

To learn more about…