Krulua ©123RF.com

By Robert Gordon

Currently the Chief Strategy Officer at SBG Technology Solutions, Inc., Robert Gordon, is helping to introduce and implement Artificial Intelligence (AI) in truly transformative ways across the IT industry. In this second article about AI for G2X, he shares further thoughts regarding the Democratization of AI and its impact on the Health IT Community.

AI Exploding – Time to Get Informed!  

The use of AI is greatly expanding. Its volcanic effects, however, have benefited just a few sectors to date, such as those related to Big Tech. While the small businesses generate almost half of the economic activity in the United States; they have yet to fully embrace, leverage and benefit from the significant progress that AI has made in the areas of analytics, classification and improvement of business workflow. The same lag applies to many sectors in the Federal Government and non-profit community that can more fully adopt and operationalize AI. Democratization of AI can help to spread the positive effects that AI has had on Big Tech across other sectors, but it is essential to come to grips with a few strategic imperatives if democratization is to truly take hold significantly in the longer term. In this article I will discuss what democratization of AI means and furthermore, why it is important for the long term success and adoption of AI tools and platforms.

Let me first define “democratization” as a process through which substantive democratic changes occur to make something – in this case, AI – more accessible to everyone within a regime. In this case the “regime” I refer to is the multitude of sectors that make up the US economy, to include the intergovernmental relations, i.e., federal, state, and local governments; big business; small business; academia and the non-profit sector. These sectors play a critical role in shaping our health care sector in significant ways for delivery and administering care.

Typically, articles considering the democratization of AI focus on tactical aspects of AI, such as access to data, simplification of developing AI tools, and the dramatic reduction of costs in devising an end-to-end AI platform. In my first article on the rise of AI 2.0, I discuss the importance of these elements necessary for democratization. These factors are clearly important to democratization in the technology sector, and Artificial Intelligence (AI) in particular; but they are not sufficient.

Strategic Approaches for Adoption

There are three broader strategic approaches that are needed: establishing laboratories of democracy; charting an ethical pathway and strengthening the alignment between policymaking and AI. All are vital for democratization and its widescale adoption.

Laboratories of Democracy

There are bulwarks of the American system of Government that serve as guideposts for establishing a robust system of democratization for technologies such as AI. For instance, the US Constitution establishes a robust system of American Federalism, which divides power and authority between the US Federal Government and the 50 states. This engenders a concept called Laboratories of Democracy, which means that innovation can freely occur at the state level and can be diffused, or shared, across state lines and at times these innovations are adopted by the Federal Government. Our Constitution’s 10th Amendment holds that “all powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.” Author David Osbourne takes this concept a step further in his seminal book, Laboratories of Democracy. He argues that innovation and pioneering can be spurred by subunits, in this case states, and greatly benefit a larger governmental ecosystem to achieve breakthroughs and substantive change.

Innovation with respect to AI, and hence democratization of ideas, can follow the same path with a cross sector approach to harnessing AI education from one sector to another, making AI tools available in many sectors, establishing an environment in which these innovations can be shared, debated, and potentially adopted. Establishing conditions for “laboratories” among different sectors can lead to more frictionless interaction, collaboration, and the eventual evolution of a marketplace of ideas.

Charting an Ethical Pathway

Ethics that drive behaviors of individuals and organizations, based on the normative aspects of moral choice, are especially germane to the health care sector, and crucial to democratization since AI is a human-centric undertaking with potentially significant consequences on the human condition. Indeed, guidelines and rules of democratization, while necessary, are alone insufficient in forging a system of values that establishes a healthy and vibrant moral grounding to minimize the development and use of AI for nefarious purposes, while at the same time unleashing AI’s promise to positively disrupt health care to improve overall health care delivery and services. Organizations such as the Department of Defense and Department of Veterans Affairs are exploring clear and definitive AI ethical principles to serve as an ethical rudder to steer the AI community to understand and adhere to behaviors that are accountable, responsible and legal in the creation and operationalization of AI. These AI principles include benchmarks that focus on responsible, equitable, traceable, reliable and governable uses of AI. The point here is that to improve pathways to AI empowerment and broaden access to AI, key in harnessing AI’s benefits is to establish an ethical foundation to control for deleterious effects on privacy, security and safety of a society.

Good Public Policy and AI

AI is experiencing breathtaking breakthroughs, such as the creation of GPT -3. GPT-3 is an AI program in the form of an unsupervised autoregressive language model created by Open.AI. This incredibly impressive piece of AI has been pre-trained on 175 billion parameters, which means it has read and ingested just about everything on the public internet. GPT-3 is orders of magnitude more accurate and 100 times faster than its predecessor – GPT2. GPT-3’s capabilities are endless, and already fields such as the legal community see this technology having an incredibly democratizing effect, with its ability to auto produce documents that were once the charge of junior lawyers, and even put those documents in the “voice” of judges or justices, or anyone else for that matter.

Within the medical community, imagine auto created documents, in terms of white papers, research and even proposals to solutions for diseases, that GPT-3 could produce. Such an AI breakthrough can expand entirely new vistas around answers to complex questions and provide new frameworks for transformative solutions in health and medicine. Yet, it has already been determined that GPT-3 is not always accurate or precise … remember, it has digested nearly all the data from the public internet, including nonfactual and false information. This can create bias and begs the question of what policies we should put into place to control for the potential deleterious effects of this type of technology.

Good public policy intertwined with effective democratization must be considered if we are to ensure a balance between rapid breakthroughs in technologies like AI and potential harmful second and third order effects on basic rights and freedoms. The challenge is that policymaking establishing legitimate authority in a representative democracy takes time and does not evolve at anywhere near the speed and rate of technological change. If AI continues its rapid pace of evolution without a commensurate change in the pace of policymaking, the gaps between the overall effects of technology change and the ability of society to control those effects through policymaking will continue to widen, causing a social disequilibrium that could be very difficult to rectify in the longer term.

Democratization of AI Needed Now

In summary, democratization of AI is more than opening the spigot to the access of new AI tools that would foster and empower a greater number of cross sector participants to innovate and bring to market newfound solutions. Larger strategic considerations – establishing a viable marketplace of ideas through laboratories of democracy, charting an ethical pathway and realigning technology advances and public policymaking – are vital to establishing a viable and democratized AI environment that will have long lasting cross sector and societal benefits.

About Rob Gordon

Rob Gordon brings more than 30 years of experience conceptualizing and ensuring application of creative solutions for the Federal Government. With a background in high level innovation, large scale complex systems management, comprehensive strategic planning and execution, Rob leverages his broad networks and deep leadership and management experience to strategically grow SBG’s short, mid, and long-term market position within the federal technology landscape.

About SBG Technology Solutions

Founded in 2004, SBG Technology Solutions, Inc (SBG) is a Service-Disabled Veteran-Owned Small Business (SDVOSB) that leverages extensive governance and engineering expertise to offer customers comprehensive, enterprise-wide approaches and innovative technical solutions for critical mission execution. With a primary focus on clients’ strategic business needs, SBG continues to deliver best-in-class solutions that govern modernization, optimize operations, and secure vital IT infrastructure.

SBG Technology Solutions is a T4NG Prime Contractor, Service Disabled Veteran Owned Small Business and industry leader in Health IT services.  For further information on Artificial Intelligence  please contact the author, Robert Gordon https://www.linkedin.com/in/robert-gordon-2a19771/ and visit our website, www.sbgts.com.

Advertisements

LEAVE A REPLY

Please enter your comment!
Please enter your name here