The U.S. government is increasingly using open source software as a way to save money and rollout the best technology. The government is also pushing development of artificial intelligence to improve everything from combat readiness to Naval maintenance programs.
The Office of Management Budget released its open source policy in 2016. The goal is for the government to have more sharable, reusable code and to stop licensing the same proprietary code over and over again.
But open source is not always free, especially not the kind that’s needed for mission-critical initiatives in government, military or business. And artificial intelligence relies on good data in to get good intelligence out.
Before deciding on open source and AI, any government agency—or even private sector company—should, as outlined in the policy, ask several questions to ascertain if open source is the right selection:
The policy requires agencies to examine the total life cycle cost of IT purchases. Open source software has a huge pricing spectrum from free to very expensive. There is a reason that IBM recently paid $34 billion for Red Hat, the largest provider of open source-based solutions. More often than not, there is an enterprise edition of open source software that packs a commercial license, and only a stripped-down “community edition” with a free license.
Also, product support is critical. When enterprises license commercial software, they typically buy a level of support along with it. In an open source community, there is no single source obligated to assist users. Open source software support comes through forums, chat and other community-led resources and is voluntary. Vendors do provide support options, but those come with a price tag.
While minimizing capital expenses by acquiring “free” open source software is appealing, the up-front cost of any software endeavor represents only a fraction of the total outlay over the lifecycle of ownership and use. And while cost effectiveness is important, it must be carefully weighed against mission effectiveness.
What’s the technical debt being delayed?
Open source software usually requires assembly to meet the unique and complex demands of government mission-critical projects. So you might be shifting cost from the software to the labor and integration category. It always sounds easier than it is to make software work optimally — especially in today’s rapidly changing technology landscape. Open source software has spurred a revolution, and the thought of “doing it yourself” is very attractive — until you realize that you are, in fact, doing it yourself. Open source software enables users to see how the product works, and users also typically must understand how to install, integrate, configure, customize, and optimize the software. If an IT staff lacks expertise, there may be long or difficult deployments and the need for additional IT staff.
Also, if staff is deploying new software, it means they’re not doing something else. Is putting together, testing, managing, upgrading, securing, and governing a set of components the best way for your staff to spend their efforts? There may be cases where it is, but the choice to be an integrator is not a small one.
Is the data feeding your AI trustworthy?
Data is the lifeblood of AI. Without data, AI cannot learn and make predictions. If AI is based on untrustworthy data, AI will make untrustworthy predictions.
Open source-based databases may require additional management to ensure data security, good data governance and rapid access to mission-critical data.
Any military force, for instance, must integrate complex data across organizational boundaries to uncover actionable insights that promote sustainable forward presence. This requires a more complete, secure and governed view of critical data across the command in order to deliver combat ready forces.
Good data governance means that there’s assurance that the data has not been changed or, if it has been changed, that there’s a record of when, how and by whom. Data governance policies—including privacy policies—need to stay with the data so that, as data moves, the policies remain in force.
When assessing data governance capabilities, a full review of everything involved—including integration, data trustworthiness and AI-readiness—is required when deciding open source or not.
A place for open
Clearly, open source software has a place in government, as does the emphasis on open standards and making custom code open to other federal agencies. But the absence of a price tag doesn’t mean there is no cost to open source, and the need for high-quality data and data management to feed AI is a critical issue. If agencies don’t have the integrated, trusted, accessible data that they need to meet mission objectives, projects and success will be at risk.
Brigham Bechtel is Chief Strategist for Intelligence and Defense at MarkLogic.
Open Source and AI: Ready for primetime in government?
...
The U.S. government is increasingly using open source software as a way to save money and rollout the best technology. The government is also pushing development of artificial intelligence to improve everything from combat readiness to Naval maintenance programs.
The Office of Management Budget released its open source policy in 2016. The goal is for the government to have more sharable, reusable code and to stop licensing the same proprietary code over and over again.
But open source is not always free, especially not the kind that’s needed for mission-critical initiatives in government, military or business. And artificial intelligence relies on good data in to get good intelligence out.
Before deciding on open source and AI, any government agency—or even private sector company—should, as outlined in the policy, ask several questions to ascertain if open source is the right selection:
Find out how to best drive desired outcomes using artificial intelligence and automation in our new ebook, sponsored by Maximus. Download today!
Is it really free?
The policy requires agencies to examine the total life cycle cost of IT purchases. Open source software has a huge pricing spectrum from free to very expensive. There is a reason that IBM recently paid $34 billion for Red Hat, the largest provider of open source-based solutions. More often than not, there is an enterprise edition of open source software that packs a commercial license, and only a stripped-down “community edition” with a free license.
Also, product support is critical. When enterprises license commercial software, they typically buy a level of support along with it. In an open source community, there is no single source obligated to assist users. Open source software support comes through forums, chat and other community-led resources and is voluntary. Vendors do provide support options, but those come with a price tag.
While minimizing capital expenses by acquiring “free” open source software is appealing, the up-front cost of any software endeavor represents only a fraction of the total outlay over the lifecycle of ownership and use. And while cost effectiveness is important, it must be carefully weighed against mission effectiveness.
What’s the technical debt being delayed?
Open source software usually requires assembly to meet the unique and complex demands of government mission-critical projects. So you might be shifting cost from the software to the labor and integration category. It always sounds easier than it is to make software work optimally — especially in today’s rapidly changing technology landscape. Open source software has spurred a revolution, and the thought of “doing it yourself” is very attractive — until you realize that you are, in fact, doing it yourself. Open source software enables users to see how the product works, and users also typically must understand how to install, integrate, configure, customize, and optimize the software. If an IT staff lacks expertise, there may be long or difficult deployments and the need for additional IT staff.
Also, if staff is deploying new software, it means they’re not doing something else. Is putting together, testing, managing, upgrading, securing, and governing a set of components the best way for your staff to spend their efforts? There may be cases where it is, but the choice to be an integrator is not a small one.
Is the data feeding your AI trustworthy?
Data is the lifeblood of AI. Without data, AI cannot learn and make predictions. If AI is based on untrustworthy data, AI will make untrustworthy predictions.
Open source-based databases may require additional management to ensure data security, good data governance and rapid access to mission-critical data.
Any military force, for instance, must integrate complex data across organizational boundaries to uncover actionable insights that promote sustainable forward presence. This requires a more complete, secure and governed view of critical data across the command in order to deliver combat ready forces.
Read more: Commentary
Good data governance means that there’s assurance that the data has not been changed or, if it has been changed, that there’s a record of when, how and by whom. Data governance policies—including privacy policies—need to stay with the data so that, as data moves, the policies remain in force.
When assessing data governance capabilities, a full review of everything involved—including integration, data trustworthiness and AI-readiness—is required when deciding open source or not.
A place for open
Clearly, open source software has a place in government, as does the emphasis on open standards and making custom code open to other federal agencies. But the absence of a price tag doesn’t mean there is no cost to open source, and the need for high-quality data and data management to feed AI is a critical issue. If agencies don’t have the integrated, trusted, accessible data that they need to meet mission objectives, projects and success will be at risk.
Brigham Bechtel is Chief Strategist for Intelligence and Defense at MarkLogic.
Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.
Related Stories
Bad data costs Americans trillions. Let’s fix it with a renewed data strategy
Securing the AI data pipeline with confidential AI
Taming the breach: Is U.S. incident disclosure working?