Does it Pay to be a Whistlebower? and ChatGPTs General Answer.

In HBS Working Knowledge and Podcast ... 

Business Research for Business Leaders

Topics, Sections

14 FEB 2023 COLD CALL PODCAST follows ... 

Does It Pay to Be a Whistleblower?

In 2013, soon after the US Securities and Exchange Commission (SEC) had started a massive whistleblowing program with the potential for large monetary rewards, two employees of a US bank’s asset management business debated whether to blow the whistle on their employer after completing an internal review that revealed undisclosed conflicts of interest. The bank’s asset management business disproportionately invested clients’ money in its own mutual funds over funds managed by other banks, letting it collect additional fees—and the bank had not disclosed this conflict of interest to clients. Both employees agreed that failing to disclose the conflict was a problem, but beyond that, they saw the situation very differently. 

One employee, Neel, perceived the internal review as a good-faith effort by senior management to identify and address the problem. The other, Akash, thought that the entire business model was problematic, even with a disclosure, and believed that the bank may have even broken the law. Should they escalate the issue internally or report their findings to the US Securities and Exchange Commission? Harvard Business School associate professor Jonas Heese discusses the potential risks and rewards of whistleblowing in his case, “Conflicts of Interest at Uptown Bank.”  ...    '

Part II:

ChatGPT  answers, much more generally, 

Does it Pay to be a Whistleblower?

Being a whistleblower can come with significant risks and consequences, both personal and professional. However, it is important to note that there are also potential rewards and benefits to blowing the whistle on wrongdoing.

In some cases, whistleblowers may be eligible for financial rewards, such as a percentage of the amount recovered in a government investigation or qui tam lawsuit. Whistleblowers may also receive protection from retaliation, such as being fired, demoted, or harassed, under various whistleblower protection laws. These laws can protect whistleblowers from retaliation by their employers or colleagues, including by providing legal remedies, such as reinstatement, back pay, and compensatory damages.

Additionally, whistleblowers may also experience a sense of personal satisfaction and fulfillment from doing the right thing, and from knowing that they have helped to expose and prevent wrongdoing that may be harmful to others.

However, it's important to note that being a whistleblower can also come with significant personal and professional risks. Whistleblowers may experience retaliation, such as being ostracized, bullied, or harassed by their colleagues and superiors, and may face difficulty finding new employment in the same industry or field. The process of blowing the whistle can also be stressful and time-consuming, involving legal proceedings, investigations, and media attention.

Ultimately, the decision to blow the whistle on wrongdoing is a personal one, and depends on a variety of factors, including the severity of the wrongdoing, the potential risks and consequences, and the individual's personal and professional values and goals. It's important to weigh the potential benefits and risks carefully and seek the advice of a qualified attorney or whistleblower advocate before making a decision.   END

Compare,  Usefulness?  

Simulate, Constrain, Repeat, Learn

Berkeley Bair posts an interesting look at Reinforcement Learning.   Made me think,  But once you get beyond the paras below this gets quite complicated and technical.   Anyone who has written significant simulation packages can be amazed at what they can accomplish. And embedded with reinforcement learning to provide direction, consider the possibilities.  And suggestions that anything can be a 'simulation'  gives us pause.   But how accurate can it be in real contexts?  Worth thinking it. 

Learning What To Do by Simulating the Past    By David Lindner, Rohin Shah    May 3, 2021,    Berkeley Bair

Reinforcement learning (RL) has been used successfully for solving tasks which have a well defined reward function – think AlphaZero for Go, OpenAI Five for Dota, or AlphaStar for StarCraft. However, in many practical situations you don’t have a well defined reward function. Even a task as seemingly straightforward as cleaning a room has many subtle cases: should a business card with a piece of gum be thrown away as trash, or might it have sentimental value? Should the clothes on the floor be washed, or returned to the closet? Where are notebooks supposed to be stored? Even when these aspects of a task have been clarified, translating it into a reward is non-trivial: if you provide rewards every time you sweep the trash, then the agent might dump the trash back out so that it can sweep it up again.1

Alternatively, we can try to learn a reward function from human feedback about the behavior of the agent. For example, Deep RL from Human Preferences learns a reward function from pairwise comparisons of video clips of the agent’s behavior. Unfortunately, however, this approach can be very costly: training a MuJoCo Cheetah to run forward requires a human to provide 750 comparisons.

Instead, we propose an algorithm that can learn a policy without any human supervision or reward function, by using information implicitly available in the state of the world. For example, we learn a policy that balances this Cheetah on its front leg from a single state in which it is balancing.  ...."

Advances for Reinforcement Learning

Very interesting,   The very first para below does a good job of  'why' this could change RL methods,  the rest of the article then carries on more technically.  Supporting images are at the link . Is this a big deal?  Humans determine they have reached a solution by comparing it to something they perceive is 'correct'.  Like an image of correctness.   Considering how this would be most useful.    Could we teach a system to learn to learn patterns of correctness? 

Recursive Classification: Replacing Rewards with Examples in RL

Wednesday, March 24, 2021    Posted by Benjamin Eysenbach, Student Researcher, Google Research

A general goal of robotics research is to design systems that can assist in a variety of tasks that can potentially improve daily life. Most reinforcement learning algorithms for teaching agents to perform new tasks require a reward function, which provides positive feedback to the agent for taking actions that lead to good outcomes. However, actually specifying these reward functions can be quite tedious and can be very difficult to define for situations without a clear objective, such as whether a room is clean or if a door is sufficiently shut. Even for tasks that are easy to describe, actually measuring whether the task has been solved can be difficult and may require adding many sensors to a robot's environment.

Alternatively, training a model using examples, called example-based control, has the potential to overcome the limitations of approaches that rely on traditional reward functions. This new problem statement is most similar to prior methods based on "success detectors", and efficient algorithms for example-based control could enable non-expert users to teach robots to perform new tasks, without the need for coding expertise, knowledge of reward function design, or the installation of environmental sensors.

In "Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification," we propose a machine learning algorithm for teaching agents how to solve new tasks by providing examples of success (e.g., if “success” examples show a nail embedded into a wall, the agent will learn to pick up a hammer and knock nails into the wall). This algorithm, recursive classification of examples (RCE), does not rely on hand-crafted reward functions, distance functions, or features, but rather learns to solve tasks directly from data, requiring the agent to learn how to solve the entire task by itself, without requiring examples of any intermediate states. Using a version of temporal difference learning — similar to Q-learning, but replacing the typical reward function term using only examples of success — RCE outperforms prior approaches based on imitation learning on simulated robotics tasks. Coupled with theoretical guarantees similar to those for reward-based learning, the proposed method offers a user-friendly alternative for teaching robots new tasks.  ... " 

Reinforcement Learning for Skill Discovery

Can skills be dsicovered.  That is, a means to find better behavior that leads to prescribed real-world goals?   Here in the Google Research blog,  addressing the use of unsupervised reinforcement learning (RL). Note the determination and inclusion of constraints.   Like in classic optimization problems. Largely technical, but thoughtful positioning.  Considerable links in the article below.

DADS: Unsupervised Reinforcement Learning for Skill Discovery
Friday, May 29, 2020
Posted by Archit Sharma, AI Resident, Google Research

Recent research has demonstrated that supervised reinforcement learning (RL) is capable of going beyond simulation scenarios to synthesize complex behaviors in the real world, such as grasping arbitrary objects or learning agile locomotion. However, the limitations of teaching an agent to perform complex behaviors using well-designed task-specific reward functions are also becoming apparent. Designing reward functions can require significant engineering effort, which becomes untenable for a large number of tasks. For many practical scenarios, designing a reward function can be complicated, for example, requiring additional instrumentation for the environment (e.g., sensors to detect the orientation of doors) or manual-labelling of “goal” states. Considering that the ability to generate complex behaviors is limited by this form of reward-engineering, unsupervised learning presents itself as an interesting direction for RL.

In supervised RL, the extrinsic reward function from the environment guides the agent towards the desired behaviors, reinforcing the actions which bring the desired changes in the environment. With unsupervised RL, the agent uses an intrinsic reward function (such as curiosity to try different things in the environment) to generate its own training signals to acquire a broad set of task-agnostic behaviors. The intrinsic reward functions can bypass the problems of the engineering extrinsic reward functions, while being generic and broadly applicable to several agents and problems without any additional design. While much research has recently focused on different approaches to unsupervised reinforcement learning, it is still a severely under-constrained problem — without the guidance of rewards from the environment, it can be hard to learn behaviors which will be useful. Are there meaningful properties of the agent-environment interaction that can help discover better behaviors (“skills”) for the agents?  ... "   ... ' 

A Peer to Peer Bonus Reward System

Intriguing bonus system architecture, or a crowd sourcing of rewards?

Implementing Peer-to-Peer Bonus System by Savita Pahuja on Aug 03, 2015  

Jurgen Appelo, CEO at Happy Melly and author of the book Management 3.0 explained peer-to- peer bonus system. He says that in a business that depends on collaboration, you should receive your bonus from your colleagues (not from your manager) with a peer-to-peer bonus system. Jurgen mentions that his team members are supporting, evaluating and crediting each other continuously.    ... " 

Random Freebies Creating Loyalty

In Retailwire: This is a classic psych finding, that random rewards can produce a kind of loyalty. Though mostly found in experiments with pigeons.  This was an experiment that was proposed in one of our grand experiments, but ultimately not implemented.  How well does it work with loyalty programs?   Still interested in a well constructed experiment.

Microsoft PowerBI Preview

Walter Riker points me to the Microsoft PowerBI site.   Preview free trial.  Which adds data visualization and dash boards to many MS and other data sources.    "Power BI works for your role ... Reduce risk and increase rewards - unlock insights to drive more informed and confident decisions. ... ".  Also now has an IOS App:  Microsoft Power BI   Which I have installed.  Will be looking more closely.

More on Power BI from article in the MSDN blog announcement:

" ... This morning at Microsoft’s Convergence Conference, we announced new products and services that will enable organizations to create a data culture in all parts of their organizations. Adopting a data culture starts with making information easily accessible to all - our Power BI offering is helping to accelerate that adoption by changing the way people access, use, manage and share data.

First unveiled in January, the new and improved Power BI is a cloud-based (software-as-a-service) business analytics service for non-technical business users. With just a browser or a Power BI mobile app, customers can visualize and analyze their data with easy-to-use, accessible tools and innovative technology  ...... 

With Power BI our customers can keep a pulse on their business by bringing all their data together and monitor it through a single pane of glass. Live Power BI dashboards and reports show visualizations and KPIs from data that reside both on-premises and in the cloud, providing a consolidated view across the business regardless of where the data lives."  ... '

Power BI Blog.

Dysfunctional Teams and Healthy Ecosystems

In FastCompany: An interesting view.   Somewhat over segmented.  Have seen some of these examples.   It then suggests healthier ecosystems.  Good idea, but also harder to implement in many cultures.   And where are the rewards to make these systems operate beyond just 'leadership'?

" ... In nature, one quality of healthy ecosystems is their vibrant and interdependent biodiversity, enabling adaptation to environmental shifts and threats. Healthy leadership teams, especially in today’s dynamic world, display divergent perspectives that they respect and value. They avoid getting stuck because they can evaluate options based on data and members’ individual knowledge, and ultimately find places of agreement. They see conflicting opinions as dilemmas to grapple with rather than fights to win. ... "

Risk and the Nimbleness of Brands

In K@W:   " ... Brands that take bigger risks are reaping greater rewards in the world of digital marketing, write Google’s Eric Solomon and Gopi Kallayil in this opinion piece.  ...  No matter what the task, taking risks is always scary. However, with increasing clutter in the digital landfill, these risks may soon represent the only real opportunity for brand advertising to stand out. For big and small brands alike, even small risks can lead to big rewards. It’s time to make that risky move on digital. And demonstrate mighty impact.  ... " 

Setting your CIO Free to Innovate

In Innovation Excellence:

“The CIO’s role is changing driven by cloud and an increasing number of ‘as a Service’ solutions that progressively reduce the organisations dependency on maintaining a high number of specially skilled ICT staff. Under these new conditions organisations have two options – they can make their technology managers, architects and administrators redundant and reap the short term rewards in an uptick in profits and share price or they can retain these highly skilled staff who sit at the intersection of the business and technology and create an impressive innovation team that adds considerably more value. ”

Games are for Learning

On Games and learning.  The idea has been around for a long time, but have seen relatively few really good examples of success.  Our own successes stemmed from competitive solution exploration.

" ... When I entered the games for learning business a little over two years ago, there was one word everyone wanted to talk about: “Gamification.” I was asked about gamification by top philanthropists, accosted at the Game Developers Conference about the subject, and even had to drive by a gamification billboard every evening on my Silicon Valley commute.

The “gamification” concept goes something like this: Take an existing set of activities – say banking, or exercise, or rote schoolwork (the more mundane the better, apparently) – apply a set of “game rewards” in the form of points (or leveling, or badges), and as if by magic the world will become more fun, workers more efficient, and learning more effective.

As a game designer of more than twenty years, this idea rubbed me wrong. Like, really wrong. ... " 

Web Style Analytics in the Retail Store

Good Forrester report, requires registration.

" ... Retail stores have been living in the analytical "dark ages" in comparison to digital channels, using metrics that rarely take into account customer behavior. But new advances in location-based analytics are transforming how physical store retailers win, serve and retain customers..... 

In this special report, senior Forrester analysts investigate the challenges and rewards of using in-store analytics to drive increased engagement with customers and greater efficiency in retail store operations. ... " 

On Rewards for Motivation

An overview of motivational rewards as used in gamification.  In particular a comparison of instrinsic and extrinsic methods.   What works in which conditions.  Which leads to the typical methods of Points, Badges and Leaderboards (or PBLs).  Good introduction to the basic methods behind gamification.

Kellogg Consumer Centric Data Case Study

An Ad from Aimia.

" ... Transforming CPG Loyalty With Consumer Centric Data    
It is an exciting time for the CPG industry. Building loyalty with consumers is no longer limited to retailers, airlines and hotels. CPG's also have the ability for direct dialogue and access to end users without the historical dependency on the retailer to provide this data. With more power than ever over the consumer-brand relationships, loyalty programs provide the means to capture direct insight into CPG end users in turn fostering deeper relationships. ... " 

" ... These are just a few of the benefits loyalty programs can provide. Download the Kellogg's Family Rewards case study to learn how Aimia helped create a program for Kellogg Company, designed to drive engagement and build real relationships with their consumers... " 

Sounds like P&G's 1 Consumer Place.

Complete Overview description from GMA.
 
For complete report.  Requires registration.

Kraft Product Breakthrough Innovation

In K@W:  Insightful thoughts about how innovation is important and essential in CPG.   What drives it when you have lots of well known current brands?   How should you invest? What is the meaning of a product category in this context?  Podcast and text:

" ... Like many food companies, Kraft Foods has had to deal with the rising costs for commodities, as well as the changing wants and needs of consumers. Several years ago, the company — which has annual revenues of more than $18 billion and a 27-brand portfolio that includes Velveeta, Jell-O and Kool-Aid — was launching new products at a rapid rate, but it wasn’t really investing in any of them. In recent interview on the Knowledge@Wharton show on Wharton Business Radio on SiriusXM channel 111, Barry Calpino, Kraft’s vice president for breakthrough innovation, discussed how Kraft rethought its strategy and reaped the rewards of a multi-year, multi-channel mindset to marketing. ...  "

Wharton Customer Analytics Initiative

Their latest newsletter.  Subscribe.  I follow them and write about their work from time to time, always something interesting to see.    Most recently results from symposia.  (not online here, but you can likely ask for the presentations)

 " ... Purchase evolution under a loyalty program.
Yuping Liu-Thompkins, Old Dominion University

Which is better for predicting customer retention: purchases or redemptions?
Sanjay Bapna and Greg Ramsay, Morgan State University

Which factors enable firms to benefit from sharing a rewards program? 
Valeria Stourm,The Wharton School

How managerial attitudes and customer attitudes and behavior affect employee engagement. 
Mindy Bergman, Texas A&M

Variability of customer experience as a predictor of customer value.
Daniel Korshun and Yanliu Huang, Drexel University  ... " 

 http://wcai.wharton.upenn.edu

Google Zavers

Zavers now about a year plus after launch:

About Zavers by Google
Zavers is a fast and easy way to offer the right coupons to the right shoppers, and track  redemption in real time. Shoppers can find  manufacturer discounts on their favorite retailer websites or across the web, and add the digital  coupons to their online accounts. Savings are automatically deducted at checkout when  shoppers swipe their rewards card or type in their phone numbers — no scanning or  sorting necessary.

With Zavers, retailers can reward loyal customers with relevant coupons, extend incentive programs, and speed up settlements. For manufacturers, Zavers provides the ability to manage and track the effectiveness of coupon distributions, giving marketers access to realtime redemption data so  they can tailor and optimize their campaigns  ...  "

Systems to Prevent Deception Fail

A discussion of systems that are meant to prevent deception and unethical behavior.  In Knowledge@Wharton.  A topic in any system that is meant to gather human data when there are rewards and consequences.   Will cognitive science ultimately solve this?   Neuromarketing is an attempt to deal with this.  Now from a forthcoming book:

" ... Wharton professor Maurice E. Schweitzer found quite the opposite in a recent research study. He says that unethical behavior not only can leave no negative emotional reaction but also can, in fact, trigger positive feelings.

Schweitzer, along with co-authors Nicole E. Ruedy at the University of Washington Foster School of Business, Celia Moore at the London Business School and Francesca Gino at Harvard Business School, discuss this result in “The Cheater’s High: The Unexpected Affective Benefits of Unethical Behavior,” to be published in the Journal of Personality and Social Psychology. ... " 

iBeacon Lands with InMarket

In Digital Trends: (Includes a video):

" ... Apple’s location-sensing iBeacon technology appears to be of growing interest to retailers and app developers, with mobile shopping startup inMarket the latest company to announce it’s making use of the tech.

Initially partnering with supermarkets Safeway and Giant Eagle, inMarket’s iBeacon Mobile to Mortar platform sends out a variety of information to iPhone-owning store visitors, so long as they’ve opted in to use the service via one of its compatible apps, such as CheckPoints.

By enabling the service, shoppers can expect to receive notifications to their Apple handset such as discount coupons, loyalty rewards, and reminders about what to pick up.

“Besides Apple themselves, we are the first to take this groundbreaking technology out of beta tests and into consumers’ lives throughout the heartland of the US,” Todd Dipaola, CEO and co-founder of inMarket, said in a release. ... " 

Kellogg Mobile Augmented Reality Marketing

A nice example of using augmented reality in marketing by Kellogg.   With music. " .... Kellogg is joining General Mills and Mondelez International in bringing augmented reality to packaging. In Kellogg's case, users with the brand's mobile application and rewards program can scan boxes to bring up an offering of four different musical artists and the chance to buy tickets to a Live Nation concert ... "