This post first appeared on IBM Business of Government. Read the original article.
To provide comments on any of the below proposals, please email businessofgovernment@us.ibm.com.
Proposal One: Improving Knowledge Delivery Through The Next Generation of Intelligent Chatbots
The “new” normal is one fraught with many risks. However, new challenges bring opportunities for innovation and creativity. Such is the case for the Federal Government in managing risk and coping with change under these pandemic conditions.
Artificial intelligence (AI) for developing user-centered Chatbots can be of great assistance in the federal government during these stressful times. Some work has already been done, including a 2019 AI-based Chatbot at GSA to provide online advice and improve customer service for users of the USA.gov website. However, many of these existing chatbots do not apply the advances in machine learning and natural language processing. Earlier this year in January, Google released Meena, a 2.6 billion parameter end-to-end trained neural conversational model and scored 79% versus humans (86%) in the sensibleness and specificity average. Gartner predicts that by 2024, the workload of managers will be reduced to 69% of current levels, due to AI advancements.
As COVID-19 has created a “more work from home” online environment, there will be a greater need for Federal Government agencies to apply AI Chatbots using the available and emerging AI/NLP technologies. This should improve user engagement and knowledge delivery to reduce the risks of not providing the right information at the right time to the right individual. Federal agencies should embrace the next generation of AI chatbots which should improve internal/employee and external/taxpayer delivery of services, both knowledge-driven and actual products. According to the March 12, 2020 Forbes article, the next generation of intelligent chatbots will transform the way services are delivered. The Federal Government should encourage the development and implementation of these intelligent chatbots in the near future.
Proposal Two: Re-thinking the Strategic National Stockpile
In order to see the full spectrum of PPE supplies available to us and discern those that are not truly available to us (e.g. ghost stock) the country needs to re-vamp its perspectives on its Strategic National Stockpile, and embrace a PPE sourcing strategy that seeks to enhance supply chain immunity over resilience.
Flexible. A key component of a future state supply chain response is the ability to withstand different requirements that need to be pulled together. This requires advanced planning, effective category intelligence, and strategic sourcing plans for every key need that might arise in an emergency.
Traceable / Transparent. Contractual requirements must be supplemented by inventory visibility systems throughout all healthcare networks. This can be best achieved through blockchain transaction channels, along with a QR or barcoding system that is attached to every item of inventory in the system.
Persistent / Responsive. A national response system must be decisive and efficient in making decisions, based on real-time data provided by the visibility system. Because events in a crisis such as COVID move rapidly, the materials system must also be able to deploy material based on actual values tracked in real-time (not in a batch or audit system).
Globally Independent. Outsourcing of manufacturing capabilities in North America has been on-going for more than 20 years. A national policy is needed that truly develops an understanding of the risks of localization vs. globalization, and the critical categories that must be co-located to promote national security.
Equitable. During a pandemic we have seen large integrated delivery systems, individual hospitals (in and outside of these systems), government delivery systems including military and VA, prisons, nursing and senior residential facilities and rural hospitals and clinics all seeking products.
Proposal Three: State Tax Administration Will Never Be the Same
The COVID-19 pandemic has changed so many aspects of American society. For those of us who work in the public finance field, the health crisis has irrevocably changed state tax administration. if you dealt with state revenue one year ago, you would not recognize the field today or next year. Everything from audits to appeals have gone virtual. And they will remain so for two reasons. First, state governments were surprisingly prepared to deal with remote tax administration. And that will improve. Second, taxpayers, or more importantly taxpayer accountants are and will be working from home. That home may be in a different state. The convergence of new technology and a permanent remote workforce will mean the government mechanism for collecting $1 trillion a year will change for good.
Proposal Four: COVID-19, AI, and Allocating Governance Tasks
As AI tools have increased in their capabilities, governing organizations have also increased their use of these tools. This increase in use has opened up both new opportunities and threats for government management, with consequences both for decision making processes and the organizational structures of those governing organizations. These opportunities and threats present new and complex challenges for risk management in particular.
COVID-19 has been a shock not only to our general governing system, but also to the work that is done by governing organizations. Work, and in particular much administration work has moved from face-to-face to distance digital work. This has upended flows of both information and work for completing governance tasks. While this is a systematic challenge to governance, it also presents a unique opportunity to more intelligently design and manage the risks of both these flows for the organizational decision-making process.
At the intersect of AI tools, COVID-19, and digital distance there is an opportunity to more intelligently allocate governance tasks across and throughout organizations. If this allocation is done well, governing organizations could improve along the important dimensions of effectiveness, efficiency, and equity. For this to be done well, work tasks need to be allocated across three general task conditions, given the rise of AI tools: (1) Tasks to be completed by humans, (2) Tasks that are co-produced across humans and AI tools, (3) Tasks to be completed by AI tools. This task allocation needs to be made with careful consideration to the type of task, the context of the task, and the consequence of the task. Finally, some consideration is also given to how this task assignment process may influence the shape of the organization and the more general flows of information and decision making throughout the organization.