The arguments for greater urgency to attain AI supremacy have raged for years among technology experts and national policymakers.
Some have cited serious concerns due to U.S. lethargy in tech development.
Others posit the race for AI supremacy is actually one between U.S. and Chinese private-sector companies like Amazon and Alibaba.
While important for national security and global business competition reasons, the discussion obscures the foundation upon which AI even becomes possible: data supremacy.
Developing Data Supremacy
Data supremacy is the precursor to AI supremacy. Data is the food for AI. Without unassailable data inputs, genuine AI and AI supremacy remain unreachable.
Many AI systems pull from limited sets of black-box data. As AI grows, however, so does the number of disparate sources from which algorithms will retrieve data to generate output. AI systems must be able to authenticate verified data sources and assess the provenance of individual data.
That approach would guarantee a transparent data management system – free from bias and manipulation.
A foolproof way for enterprise, midsize, and startup companies to establish verified data provenance would be to use distributed ledger systems. Input data within such systems would enable AI professionals to build audit-algorithm tools. Those tools, in turn, would yield explainable information – an immutable and cryptographically secured data management approach. One with a blockchain structure that delivers a historical lens for data scientists in the private and public sectors to review.
Distributed ledger technology automatically defends against attacks on AI algorithms, warding off “data poisoning” – or, the manipulation of input data by attackers to modify the outcome of an algorithm.
This is why regulators and company executives responsible for proprietary or open-source data-management systems must be cognizant of input data in the same way they are already focused on certain AI outcomes. Systems that capture and validate algorithm input data are a key plank to instituting data supremacy.
U.S. Steps Toward Data Supremacy
The U.S. Department of Defense formally began developing a strategy in May 2021 to transform the DOD into a data-centric organization when Deputy Defense Secretary Kathleen H. Hicks signed a memorandum to transform the Pentagon into a data-centric organization.
That memo represented DOD’s inaugural “data decrees” that would enhance the Pentagon’s data proficiency specifically in “joint all-domain operations, senior leader decision support, and business analytics.”
The memo “empowers” the acting DOD acting chief data officer to augment the Pentagon’s entire data ecosystem – which encompasses “people, technology and culture” – as well as its data sharing, data architecture, data lifecycle management, and data-ready workforce abilities.
Individuals and Data Supremacy
Regular, everyday American citizens are also integral to achieving data supremacy. They leave traces of data every second throughout the internet.
Much of this data is bought and sold – and ultimately could become part of an AI that affects citizens. Proactive steps by companies are crucial for strengthening self-sovereign identity and other technologies that allow individual data ownership. Self-sovereign identity technology fights against bias, asserts more control over privacy, and builds public trust in the emerging commercial AI landscape. Each person would then have control over the collection, sharing, and use of individual data in web applications and AI systems.
Businesses that make that core to their data-related operations will ultimately be those that gain a leg up in this intensely competitive arena – particularly as data privacy and individual ownership over one’s data continue to become paramount among citizens throughout the world. National and transnational governmental entities are already implementing such policies.
The emerging European Union’s Data Governance Act is a model for incorporating citizen rights, such as personal privacy and trade secrecy. It emphasizes investments that promote data availability, transparency, the common good, and trusted and secure data sharing. A more-informed society is a better society. We need to build structures for truthful, unbiased data reporting.
The EU has also created a European Digital Identity that will be available to all EU citizens, residents, and businesses who would like to identify themselves or confirm specific personal information. That identity is usable online and offline for EU public and private services.
Ultimately, once people own data unique to them, they should have the option of selling it if they want – particularly when data is becoming the digital equivalent of human fingerprints.
The Private Sector
Robust agency afforded to individuals is more important than ever when one considers how some businesses already have achieved a certain level of data supremacy by both collecting data on their customers, and the investments they have made to construct modern data-stack infrastructure that gives them the edge among global competitors.
Amazon, Apple, Coca-Cola, Google, Netflix, Nike, Starbucks, and Tesla are just some of the corporate behemoths that have harnessed research and development dollars into creating state-of-the-art data management practices. Each of these companies uses data technologies to better understand who their customers are and what their customers desire. Those technologies and practices have enabled those companies to rise to the top within their respective industries.
Data Supremacy Tenets
Realizing data supremacy for any organization requires the creation of data infrastructures, systems, solutions, and policies that lead to optimal data management practices for people, as well as the organization itself.
It is the dual mastery of data-centric policies like the DOD’s 2020 VAULTIS – visible, accessible, understandable, linked, trustworthy, interoperable, and secure – strategy and the provision for maximum ownership and control by individuals of their respective personal information.
That mastery not only yields more equitable stakeholder data access, but it also manifests better-informed decisions due to better-quality data and enhanced data collaboration capabilities. Organic results from this approach also include elevated data security, privacy, compliance, and integrity.
Whether a company executive or government official, leaders must reorient themselves toward envisioning data supremacy as the bedrock for accomplishing AI supremacy. Forward-thinking entities and individuals will benefit from developing a list of actionable items to tackle – one that makes the VAULTIS principles a matter of course from the data-management side, while simultaneously moving toward data privacy, protection, and ownership on the side of individuals.