AI

Perspectives On AI From The North Dakota Chief Data Officer

Read more at www.forbes.com

North Dakota is not often one of the states in the United States that is considered to be foremost for technology adoption, but the state’s thought leadership in a wide range of technology areas shouldn’t be underestimated. After all, the current Governor of North Dakota, Doug Burgum, was not only president of Billion-dollar tech company Great Plains, which was acquired by Microsoft, but also he was the head of Microsoft Business Solutions, chairman of the board for Atlassian, and on the board of SuccessFactors, and co-founder of Arthur Ventures.

It should come as little surprise, then, that North Dakota has been very forward thinking in their use of data, automation, and AI. In a recent AI in Government event North Dakota’s Chief Data Officer Dorman Bazzell shared how states are approaching technologies such as automation and advanced data analytics, as well as how AI plays an increasing role in local government. He provides more detail in a recent interview on  the AI Today podcast and he shares insights in this follow up interview here on Forbes.

What are some of the unique challenges involving data at the state government level?

Dorman Bazzell: Data at the state is verticalized, meaning it is aligned to state agency business architectures. The result is the rampant duplication of data, and hence increased costs of managing, securing and using the data. What is necessary is the shift to horizontal, state thinking about data. That is, data exists once throughout the state and agencies become process engines against that data. Agencies then become stewards upon the data with which their processes interact. Citizens become responsible for their data. Through that inverted thinking, the state can create personas that offer services to citizens. Today, if I need the services of all 47 agencies of the state, I will need to navigate 47 websites, fill in my name and address 47 times, etc. I still may not know, at the end of that journey, what services I can receive as there are multiple forms that need to be completed and mailed in. The citizen should not have to guess at services. A digital citizen platform focused on serving up services to citizens, using AI and citizen-persona while protecting privacy, becomes possible even desirable. Personally, there are services for which I may likely be eligible for but for which I would need to spend valuable time mining for. Imagine knowing that as a single parent with three kids and trying to find time after work, homework, baths, story time and a two year that has decided not to fall asleep until after 11pm.

The state has a moral obligation to protect the data of its citizens. Our legislators have invested heavily and successfully in our cyber capabilities. That protects us from the bad guys outside of the four walls of state government. 360 degree governance is what protects our citizens from the state. I will address that in the following questions.

What are some of the technical and practical challenges states such as North Dakota face around data collection, data usage, and data privacy?

Dorman Bazzell: The vision that our Governor has laid out is one of citizen self-management of data for the state. This would provide a platform for citizens to manage their own privacy as they would only need to provide the data required for a particular service needed from the state. One goal is for state employees to stop being data-entry personnel for citizens. There is a bit of tongue-in-cheek in that goal but also a great sense of urgency as the state looks to modernize its technology platforms and reinvent government.

What do governments need to prioritize, from a data management and governance perspective, as they seek to utilize AI applications?

Dorman Bazzell: I believe there are two imperatives for AI: Auditability and Explainability. From an audit perspective, being able to articulate a model’s veracity thru validated data lineage and quantified data quality lenses better positions the model for trust.  Explainability, other than the obvious nature of being to have predictable outcomes, really goes to the heart of AI technologies: did we do a good enough job of creating a use case and its outcomes such that the outcomes will always protect privacy, insure inclusiveness and diversity and there is a sense that the case for humanity is improved. 

AI, and specifically, ML have an imperative: Governance. If I could only have two table of content topics within the AI governance dimension they would be Auditability and Explainability. The more I read and understand about the ‘failures’ of ML, the more I understand that those are the macro areas where not only researchers and practitioners have concerns, but those who are responsible for the outcomes. The technology is ‘easy’. The people, process and culture of implementing ML outcomes are hard. Governance should not hinder innovation of ML. Governance should enable confidence and adoption of it.

Those may sound lofty, but they are achievable goals. These goals require a combination of actors at a state level: legislators, agencies, and IT. The 360 degree nature of governance is what gives it the ‘teeth’ to create value outcomes – services – for citizens.

What are some of the ways the state is using AI and especially robotic process automation?

Dorman Bazzell: IT has set a goal to automate 20% of the manual, redundant, mind-numbing processes of the state. The estimated saving is ~$600M. We need to execute rapidly on that goal for two reasons. The first, the state has 24% of its workforce who are eligible for retirement in the next four years. That is about 3,000 – 4,000 people. That is both a huge loss of knowledge and an opportunity. Beyond the savings opportunity, upskilling our workforce is an imperative. Upskilling is the second reason. We need a workforce that can take advantage of the digital workplace; driving data-inspired decisions, empowering execution and accountability for citizen outcomes to all state workers, and creating opportunities for the workforce of the future beyond the next set of retirees and integration of emerging technologies. 

RPA is certainly part of our technology arsenal. But, RPA alone will not solve the 20% problem and I do not see it as an endpoint technology but more of a transition technology. To achieve that 20%, and beyond, goal we will need to fundamentally examine our processes and either retire the obsolete, rewire processes such as form creation thru automation like Microsoft Dynamics or leverage RPA for such things as invoice consumption and HR on boarding until we have funds and imperatives from agencies to re-invent the process.

What are some interesting or surprising insights you can share about how North Dakota seeks to apply automation, ML and AI?

Dorman Bazzell: North Dakota has invested and continues to invest in ventures which look to further enhance our state as leaders and contributors to improving humanity. In agriculture, we are the number one producer of soybeans for the world. 80% of our soybean production finds it way to overseas markets. In the private sector, the Grand Farm initiative is looking to create and be the standard for the Farm-as-a-Platform thru autonomous agriculture. This includes the ability to track, monitor, predict and manage the entire farm-to-table value chain. Autonomous ag will deliver connected capabilities to not only farm equipment, but also IoT devices monitoring soil, air and water quality and connectivity to Grand Sky UAV to provide real-time land surveillance for sunlight and soil telemetry. Add to this the potential of the Department of Transportation to provide autonomous capabilities on roadways which deliver seed-to-silo as well as predictive weather and traffic pattern data in real-time.

Where do you see AI having the biggest impact on government services?

Dorman Bazzell: Governor (Doug) Burgum has given us six simple words for how state employees should think about state government: Empower People. Improve Lives. Inspire Success. If we focus that mindset and cultural aspirations on citizen services, AI, then, will function as one of the key enablers of those services. If you start with that end in mind, it will naturally lead you to governance, automation, AI, improved processes, and a lower cost to serve. In fact, citizens will demand it.

Perhaps, one of the biggest impacts will be that citizens do not see AI at all, but rather come to expect it as part of their state engagement. For example, when you Google something, or use your iPhone to find the fastest route to a restaurant, you do not realize that AI has an ambient presence. Where the state will excel is the ability to deliver a governance model that citizens will, again, not see, but come to rely upon as ubiquitous and a part of their experience.

What can states and local governments do to attract the skilled workforce needed to keep up with technological innovations? (Answering 7 & 8)

Dorman Bazzell: North Dakota has several ancient technology devices also known as mainframes. Not much of marketing lure for prospective state IT employees. For new agency employees, such knowledge might conjure up such things as green bar paper reports. To attract both IT and agency team members, we are actively deploying and marketing technologies such as Microsoft Azure, Dynamics, Cognitive Services, RPA, PowerBI that encourage innovation for both IT and agencies. Our data lake and adjoined data science model are encouraging analytics at the point of action, increased need for advanced degrees in mathematics, core competencies – not degrees – in Java Script, Python and .Net.

IT has invested heavily in infrastructure, security and technologies, governance and disciplined to push capabilities such as self-service analytics to the edge decision makers. We also are continuing to push capabilities such as Agency-Developer to the point of action to lower the cost of service to our agencies and citizens. That ‘push’ requires all teams to learn new technologies that re/up skills out IT teams into emerging technologies and encourages our agency teams to learn new capabilities as we shift away from manual processes.

COVID forced North Dakota, in fact, all states, to accept some form of virtual worker. Not everyone wants to live in North Dakota. However, as evidenced by resume and application submissions, many want to be a part of our highly progressive state. These individuals bring a variety of experiences and technologies into the state which help us think differently about the services we offer.

What AI technologies are you most looking forward to in the coming years?

Dorman Bazzell: A couple of innovative startups that are exploiting the ideas of AI and blockchain – Ocean Protocol and Circles UBI – offer some data exchange capabilities that could potentially transform how I think about consumption-based and charge-back models between agencies.

I am very encouraged about the connected nature of autonomous everything. For an obvious example, the idea of connected vehicles to connected, smart cities is intriguing because COVID has changed the paradigm of how we commute and interact with the cities in which we live. Also, the Grand Farm opportunity for farmers being being able to focus on the business of farming – versus having to physically sit in a tractor for hours – due to the connected autonomous nature of a farm. I grew up working for farmers. They spent significant hours in the field. With Grand Farm they could spend significant hours focused on optimizing the farm-to-table value chain to feed a growing world population. 

For the state, I have a vision of open data sources (anonymized data to preserve citizen  privacy) that has an open, model driven layer of AI tools available to citizens, business, the world, that is basically an open invitation to come to North Dakota to live and invest. This layer would be available with state supplied models, but would also provide tools allowing people and business to develop their own models to understand their communities and shape new forms of business and skills for our state. In fact, I have challenged my data science team to develop a model that predicts the next industry and skills for North Dakota!

There was a 2018 article in Smithsonian Magazine that asked ‘What will society look like when AI is everywhere?’ It was a bit of an anachronistic question as even as 2018 AI was everywhere. Ambient. As then, today we are walking around with AI in its embedded form of the smartphone. What encourages me about that reality is the opportunity, for example, if I am in medical or physical peril, my smart watch could sense that and place an emergency call to the nearest medical facility or police station and I could get assistance. Or, what if a sexual perpetrator has their smartphone on them and are near a school an alerts could be sent to school officials and children – and my grandchildren – protected. The alternate reality is the invasion and maybe false representation of me in some digital persona. That is where governance becomes imperative.

Read more at www.forbes.com

Show More

Related Articles

Back to top button