- Artificial Intelligence
- Internet of Things
- Big Data
- RT Analytics
- Machine Learning
- Use Cases
- Black Box
- Business Semantics
- Business Analysis
- Cognitive Computing
- Data Strategy
Data Science: A Fancy Black Box Is Still A Black Box
Why would anyone risk their career on a mystery, especially if the mystery is avoidable?
And yet this is what too many managers and executives are asked to do where machine learning, or even analytics, is concerned.
Managers rebel against mystery however. The manager's mantra is "no black box". One doesn't build companies or products by not knowing what's inside.
It's no surprise then when confronted with a machine learning or analytics black box solution that many executives just say "no". (See Data Decisioning's recent podcast with Decision Management Solutions for a real-world example of "no".)
Our Market Scan is capturing more and more signs that business is taking the black box problem seriously. This week's lead quote is an example: an almost heretical suggestion to focus less on outcomes -- a kind of black box -- than on the "millions of preceding smaller decisions".
That chain of decisions is what at Data Decisioning we call the "data-to-decision value chain". We didn’t' call it a "data-to-decision black box"!
Interestingly, what's inside any data-to-decision value chain is not generic. It's always specific to a given sector.
Collecting data from cows and collecting data from loan transactions could not be more different, even if they both involve 1's and 0's. So business knowledge and experience-based judgment counts.
Make a bet on great technology, combined with what you know. That's a risk worth taking ♦
Image Credit: Pixabay
"Becoming decision-driven is better than focusing on 'business outcomes,' because outcomes depend on many preceding smaller decisions, perhaps hundreds, thousands, or in the case of algorithmic trading or automated manufacturing, millions of small decisions."Matt Podboy via Mark Palmer - TIBCO - Be Decision-Driven Not Data-Driven: Q&A - #DecisionsFirst #OutcomesDriven #DecisionDriven
"Machine learning is eating up the world around us, and it works so well that it is easy to forget that we don’t know why it works."
Ankur Moitra of MIT Mathematics - MIT News - On a quest through uncharted territory - #BlackBox
"Stakeholders are often reluctant to trust ML projects because they don’t understand what they do. It’s hard for decision-makers to relinquish control to a mysterious machine learning model, especially when it’s responsible for critical decisions. AI systems are making predictions that have a profound impact, and in some industries like healthcare or driverless cars, it can mean the difference between life and death."
Lak Lakshmanan - Google - Why you need to explain machine learning models - #BlackBox
"Rational and scientific methods necessitate data without which neither can we have information, nor knowledge or wisdom."
Anu Raghunathan - The Indian Express - In Data, Let Us Trust - #DataDriven #Rationalism
"Some respondents noted that, even if workable ethics requirements might be established, they could not be applied or governed because most AI design is proprietary, hidden and complex. How can harmful AI 'outcomes' be diagnosed and addressed if the basis for AI 'decisions' cannot be discerned?"
Lee Rainie, Janna Anderson and Emily A. Vogels - Pew Research Center - Worries about developments in AI - #BlackBox #Ethics #Accountability
"While the potential of big data is irrefutable, is it the panacea for all decision-making situations? Put differently, could a strong emphasis on data and analysis backfire under some circumstances? We explored this in our recent research. Our intuition was that data-driven decision making could be counterproductive under extreme uncertainty."
Oguz A. Acar and Douglas West - HBR - When an Educated Guess Beats Data Analysis - #DataDriven #Uncertainty #Judgment
"In the past, such crucial [operational] decisions were made based on general rules of thumb or traditional business intelligence. Today’s managers and leaders, however, have the support of technology and advanced data analytics to make well-informed decisions that optimize value on all levels."
Kristel Lawrence - Global Trade - How Data Analytics Can Help in Making Better Operational Decisions - #Judgment #Rationalism
"Leaders and managers in legacy enterprises often hear that they should approach their business as if it were Amazon, Netflix or Spotify. These great companies have built enormously successful and admirable businesses; however, their models and solutions characterize twenty-year-old e-commerce companies that were designed to be data-driven before they even launched. That model often doesn’t generalize to a hundred-year-old global manufacturing company. "
Tom O'Toole - CMO Network | Forbes - Why Legacy Companies Struggle With Data Cultures—And What Leaders Can Do - #Corporates #Production #BusinessModels
"IRCC [Immigration, Refugees and Citizenship Canada] is using advanced data analytics to sort and help process temporary resident visa applications from some countries where there is a high volume of applications … Advanced data analytics systems recognize patterns to help accelerate our work and better inform decision makers … Officers always make the final decision on applications. The [analytics] systems never refuse or recommend refusing applications. When applications are refused, an officer made the decision."
Immigration, Refugees and Citizenship Canada - Government of Canada - Digital transparency: Automation and advanced data analytics - #Fairness #Accountability
"A sleek robotic trimaran retracing the 1620 journey of the famous English vessel had to turn back Friday to fix a mechanical problem. With no humans on board the [AI-driven] ship, there's no one to make repairs while it's at sea."
AP - NBC Boston - AI-Powered Mayflower, Beset With Glitch on Voyage to Mass., Returns to England - #Augmentaion #Robots #Vision
"If a picture paints a thousand words, then a map paints a million! Given the unique ability for geospatial information to provide a common canvas on which to make complex business decisions, ‘geography’ is now considered a new business platform for change and transformation at local, national and global levels of business."
Paul Synnott of Esri Ireland https://twitter.com/EsriIreland - Geospatial World - If Location Intelligence is the answer; does it matter what the question is? - #GeoSpatial #Decisioning
"There's clear evidence that in today's fast-changing environment, being data powered is fundamental to the success of consumer product and retail organizations. As competitive intensity increases both from within and outside [CPG and retail], companies need to foster a culture that enables them to gain insight and act fast."
Tim Bridges via Esther Shein - TechRepublic - Consumer products and retail companies must accelerate their data maturity to become more resilient - #CompetitiveEdge
"The platform revolution has enabled plug and play of all kinds of different applications and tools and services. You just put them together the right way to serve a business community and you’re off to the races, basically.
Kirk Borne - Datanami - In Search of Data Science Talent with Dr. Kirk Borne - #PlatformRevolution
"One in three organizations currently uses AI to make some decisions around pay, according to Gartner research. Seventy-nine percent of those report better pay standardization, and more than half report it has improved their efforts to match pay for performance."
Laura Starita - Gartner - Would You Let Artificial Intelligence Make Your Pay Decisions? - #HRAI #BlackBox #Accountability #Fairness
"Increasingly, data generated at the edge are used to feed applications powered by machine learning models. There's just one problem: machine learning models were never designed to be deployed at the edge. Not until now, at least. Enter TinyML."
George Anadiotis - ZDNet - Machine learning at the edge: TinyML is getting big - #EdgeML #UbiquitousComputing
"Infrastructure as code: that’s a great way to build systems. It reflects a lot of hard lessons from the 1980s and 90s about how to build, deploy, and operate mission-critical software. But it’s also a warning: know where your infrastructure is, and ensure that you have the talent to maintain it."
Mike Loukides - Radar | O’Reilly - Code as Infrastructure - The Next Critical Talent Shortage Won’t Be Fortran - #Maintenance #Infrastructure