AI Literate Citizen

We talk about challenges in moving AI solutions from testing into production or commercialising AI solution for all users internall / externally. We haven't really focused a lot of attention on the other extreme: the data collection and developing the proof of concept.  

Where do you start ?  Ask yourself, what is the outcome you want to achieve? What is the business problem you are trying to solve? Not understanding this is the #1 blocker to success.  

Prerequisities: There are a lot of moving parts just to get started, some examples, but not limited to:  1. what data is needed & where are you going to get it 2. availabilty of data; do you have access to this data in a repeatable way 3. what about data privacy, white room (data can not leave the site) 3. how will the data be organized 4. who is going to own the data goverance 5. do the data scienists have the right technologies to do their job 6. is the organization ready for this to start  Additional topics discussed:  - How to priortize where to start - 80% of the work by data scientist is data rangling - Example use cases in the insurance sector - Example use cases in asset management companies - Proof of concept validates if the outcomes we want to or expect to see, can really happen - Establishing the architecture in a way we can scale if the POC is sucessful - Technologies data scientist should have access to  - Monitoring data drift effectively  - Different languages across the roles. Takes a lot of change management to get this right. 

Ultlimately is the organization 'data ready'?

My podcast recordings are usually under 30 minutes, but sometimes there is topic that is so fundemental and foundational to how we live our lives every day, the clock is just thrown out. This is one of those topics.  

While there were many 'Ah ha' moments, one moment of enlightenment was just how much carbon is produced for the inefficiency in training an AI model. We looked it up during this podcast. It was shocking. AI is used to support research in reducing carbon emissions, but we may not be fully aware how much AI technologies act as a carbon emitter.   

Our goal with this discussion is by then end you are asking yourself questions you may not have thought about and/or starting asking new questions at your job tomorrow.   

Topics covered today:  What is Circular Supply Chain? Impact of on primary, secondary, and other supply chain...markets, materials, and waste How do you wrap your arms around all that data? Considerations when building new products? 

Why should we care about Circular Supply Chain? Impact of carbon footprint when training AI technologies

Redefine the definition of your carbon footprint to include systems, AI technologies, etc. 

Are we really closer to a zero carbon footprint?  What is a true circular economy? What is Zero-waste really?  The defintion from 5 years doesn't exist today. In 3 years from now, we will want to redefine this again. 

Visible impact & roles for a citizen of the earth - what can I do?  Cultivating the right mindset - the negative perception of refurbished vs OEM Roles and responsibilities - what you thought your role was today is completely changing 

Do we take a serious look at scrap & how it can be resused in new creative ways? We illustrate a real world example.  ----------------------- 

Sneha's passion is leading supply chain process improvements while focusing on the reduction of  carbon footprint and improving sustainability.

AI technologies are not new. They only became popular after some time. They have been around for decades.  As a result of supporting components such as cloud architecture and semiconductor chips, using AI technologies has become more accessible. This accessibility brings this topic into the mainstream for discussing positive and negative impacts.  

Aruna Pattam,  Head of AI & Analytics for the Asia Pacific region, is my guest on this episode.  In 2021, she was Linked in Top Voice 2021 for Technology and Innovation and recently named of the 2023 Top 100 Brilliant Women in AI Ethics. Additionally, she continuously contributes her thought leadership as a group member of the Responsible AI Think Tank. 

 In Jan 2023, Aruna published her perspectives on the Top AI Trends for 2023. Today's discussion discusses those trends and why we should be watching these more closely in 2023.  1. Impact of AI and Cybersecurity 2. Democrationalization of AI 3. Edge AI 4. Generative AI 5. Responsible AI

This interview, our discussion, and our views are now a part of the global data set.   

According to Laura, data is impartial. It doesn't have human-like qualities. The unclean data, the unstructured data is the best mirror to ourselves. Ultimately, we are responsible for the data-sets that are pulled into other organizations. i.e. we buy, we post, we upload, we like, we repost. 

We can clean data to redefine a better 'prettier' answer,  but if we want to fix the data, we may want to shift that focus to ourselves.   "... you play an active role whether you have known it or not. You have been in it all along", Laura.  "We focus on all the things that can go wrong (when discussing AI technologies), but there are benefits that can be used for good with the right gaurd rails. We have to sift through the not good in order to acheive the good", Laura.   

Additional Topics Discussed:  - Definition of ethics - Applying ethic concepts to AI technologies  - What does it mean to be a human in today's world - Moral decisions and consequences - Bias exists comes from us, the human interaction - Broken mirror theory, Laura's theory based on years or observation and reflection - AI algorithms are not entirely responsible for all the bias themes - Ethics look at the harm when the result provide the wrong answer that is harmful - Ethics-washing - we always want the pretty version without the work - Books: How to Lie with Statistics & Merchants of Doubt - Transparency Fact Sheet - Are we missing anything? - Regulations are written after innovation, i.e. Copyright Act, written in 1976 - Policies.The future requires dedicated resources to revise policies on demand or every quarter. - Right, good, fair, and just - Laura' answers the question, 'what can someone do to start thinking differently about ethics, that impacts us personally' - Protecting your images, what your have contributed yourself you may not have thought much about until now  

https://www.linkedin.com/in/lmiller-ethicist/

AI-related projects are not typical IT projects with a traditional waterfall approach.  My conversation with Lisa Palmer, Chief AI Strategist & Ethicist, AI Leaders focused around her recent doctorial thesis and research on 'Artifical Intelligence: Are For-Profit Entities using the "Do the right thing' goverance to drive business results"  

When her customers claimed they were 'doing the right thing', Lisa wanted to take that further i.e. 'What does that really mean and more importantly, what does the customer believe that means?  This led to her doctoratal statement and how companies are: 

 1. Making decisions 

2. Taking action 

3. The results of those actions  

AI Technologies are still reletatively new in how they are being used in for-profit entities. She turned to podcasts to collect the foundational data related to her thesis statements for recent and relevant information. Listening to 172 podcast episodes and narrowing this down to 46 specific scenarios included in her study, covering a date range from the last 3 years up to April 2022.   

There are 3 ways AI technologies are being used in enterprise companies: 

1. Efficiency of processes 

2. Revenue opportunities 

3. Risk Avoidance  

Other questions discussed during this episode:  -Self-adopted policies - What are companies actually doing from a policy perspective? -What are the types of companies which are doing this well? -Is the AI champion a benefit to the enterprise organization, or are there unintended consequences -What about companies that are focused too broadly? -What is a reoccurring theme for successful companies? -How important is the quality of diversity in viewpoints or successful companies? -How influential is company culture when executing AI-related technologies? -Can AI policy and regulation restrict our ability to innovate?  -Is there a disadvantage to the US which is not writing AI-related policies and regulations but instead -This remains a fluid topic. What do we need to address in 2022-2023?

A recent graduate from Nichols College in Boston, MA and now working as an Automation Engineer at Paragus IT, Skyla started her educational journey with an Accounting degree,  but later changed that concentration to technology and automation.   

In a conversation with her advisor, it was suggested perhaps Accounting wasn't the right fit. These discussions led to an internship through Center of Intelligent Automation (CIPA) program at Nichols College.  Skyla defines automation as the improvement of manual processes that are rule based and repetitive. Mostly data entry that was occuring daily or weekly. She using Power Automate with Microsoft to support her automation  Two (2) examples where she has made a visible impact to her internal stakeholders are:     

- The same email is sent every week, from the same supervisor with the same body of content to remind folks about their time sheet.     

- Employees submit reimbursement requests via paper. Through the use of mobile app, employees can send those request very quickly and can managers can quickly approve or deny.  

Other topics discussed wtih Skyla:  

- What prompted the change - the appeal and challenge - Types of activities as an automation engineer - Not every process can be solved by automation.  - Do stakeholders reach out to the department or is it a struggle for internal stakeholders to embrace automation? - Listening is the #1 skill set. The process is explained differently by different people or doesn't exist on paper. - The value of current state processes - Skyla's recommendations for other women just starting their career journey.  https://www.linkedin.com/in/skyla-w/ https://www.linkedin.com/company/paragus-it/

AI4Good. We talk about this often but haven't focused a lot of time on understanding the 'good' part of this phrase. It is women like Shiran Somech who are using AI technologies that truly have a global social impact.   She has found the perfect blend of her two (2) primary passions of AI technologies and wants to make a visible social impact on women's rights. Our discussion talks about her initial idea and carrying that forward from concept to reality, leveraging AI Technologies to create something that may not have been possible five (5) years ago.   

Social Media Campaign 'Listen to my Voice' An AI-based campaign speaking up against domestic violence uses AI technologies to bring the voices of women murdered by their initimate partners who are no longer here to tell us their stories.   Our conversation covers the partnerships, the impact, and some technology challenges she had overcome in bringing this campaign to life.  This could not have been possible without the support of the families. To realize this concept, collaborate with non-profit government partners, tech creators, and sponsors. As a result of its extraordinary success, the campaign is now being duplicated around the world: https://www.listentoourvoices.co/en/. 

Volunteer Special Projects Tochnit Saleet, is a program sponsored by the Ministry of Welfare & Tel Aviv Municipality to help women break out of the sex-work cycle and integrate into the hi-tech workforce. In this role, Shiran helps to raise funds, provide mentorships, and upskill women, enabling them to enter the Israeli technology market.

"Best Friends is transforming the animal welfare industry and data is transforming Best Friends", Michelle Dunivan, PhD Dunivan, Analytics Director of Best Friends Animal Society.  As we approach the 2022 holidays in the United States, the adoption of kittens & puppies increases as gifts to family members.  

What we don't see is everything behind the scenes to ensure your local animal shelter or rescue center has the right balance of animals to service local demand. We never see the amount of data which needs to be collected,  to resolve key business challenges for the betterment of animal welfare. 

In talking with Michelle, data provides her organization with the ablilty to ask questions no one thought to consider five (5) years ago because the data wasn't as accessible as it is now.  In the future, data can help Animal Welfare Industry to be proactive instead of reactive to really understand the pyschology of adoptions.  

What if data could help us understand (or predict) the outcome when a type of dog breed is accepted by the local shelter in a geographical area, at a certain time of year, and possibly the time of month.  

We packed a lot information in just under 30 minutes. Other topics during our discussion include:   - Roles & reponsibilities supporting data and analytics in animal welfare  - Research in the real world compared to the obstacles of academia research  - Studies: Animal Behavior (kpeeing animals living longer)  - Research: Community cat programs  - Supply and Demand: what animals are needed in which states  - Myths: Pandemic Puppies are bring returned. The data refutes this perception  - Data collection: Collecting from animal shelters across the United States  - Establishing the right amount of data to achieve our business goals   - Reporting to bring actionable decisions not just to have lots of data  - Constituents, volunteers, donnations, and local state and county policies

"There is no such thing as a bad project, only poor project planning." Personally, this has been my quote for 15+ years. Executing a good project plan is a balance between decisions and data. The decisions we make before and during the project and data availability facilitate those decisions.  

We have all been there. Using XLS instead of Microsoft Project to create a project plan with just enough detail to allow you to track the progress, and yet somehow, we cannot make accurate decisions with a high level of confidence about the success or failure of the project. Yes, we can say the project is ahead or behind, and we know when a project goes longer than expected, we can use the existing run rate of our resources to estimate how much more money will be needed to finish the project.  

What about the majority of projects that continue to go over budget or run long multiple times before the project is completed?  How do we know when and where to shift resources to other workstreams, so the project is back on track?  More importantly, can we predict a project will be successful or fail before the kick-off and explain why?  

With the accessability of AI technologies, Greyfly.ai combines the foundational concepts of project management to generate true project intelligence. This software solution enables project managers to predict proactively the expected success or failure of a project and, during the project, outline what project parameters need to be changed to bring the project back on track.   

In talking with Marcia Williams, Chief Product Officer of Grefly, leveraging AI technologies on top of the data enables predictive analytics to address project questions with higher confidence. The fact-based decisions reduce the risk that can drive potential savings (or cost-avoidance. The historical data can be used to highlight if a project has the potential to be a failure and address the 'why.'  

There is a US cartoon in the 1980s I used to watch on Saturday mornings called  GIJoe. Their slogan was 'knowledge is power and 'knowing is half the battle. With improved project intelligence, using AI technologies is helping to harness the power of project data and to answer the one question we could not figure out before 'Why?'.  

Other Topics Discussed:   

- a professional journey from IT auditor to Product Development 

- data quality -

- drive a fluid business culture to allow the customer to take you where you need to go 

- the product officer's role 

 - resource capacity through an organic growth structure