The intelligence community’s venture capital arm, In-Q-Tel, published a helpful primer on explainable artificial intelligence.
I’m not an expert in AI and, if you’re not either, these excerpts may elevate your understanding:
As use of AI and Machine Learning (ML) becomes increasingly common across industries and functions, interdisciplinary stakeholders are searching for ways to understand the systems they are using so that they can trust the decisions such systems inform. This effort is sometimes referred to as “Explainable AI” or “XAI”.
The focus on trust and understanding that is driving the XAI movement relates to important questions of law and policy. An explanation for an AI or ML system can put the system’s reasoning into the open for debate about whether it is equitable or just, or may enable some sort of actionable understanding around why a decision was made.(5)
Some researchers, like Facebook’s Chief AI Scientist Yann Lecun and Google Brain’s Geoff Hinton, have argued that asking systems to “explain” themselves is a complex, infeasible task that may not lead to actionable insight.(6,7) Others disagree, arguing that explainability is necessary, as technologists need to consider the social implications of all parts of their AI systems.(8,9,10) Moreover, they argue, evolving research may make the task increasingly feasible.(11)
One helpful way to characterize efforts in XAI is by applicability — for example, whether a technique can be used to interpret or justify a single model or many, or whether it can be used to interpret or justify a single decision or larger trends.
While most new work and research on these techniques is coming from the academic sector, XAI tools are beginning to materialize in the market. Whether XAI companies will be able to stand on their own, or if these tools will primarily be absorbed as a feature by established AI/ML players, remains to be seen.
Because “responding to foreign interference requires a whole of society approach,” the U.S. Department of Homeland Security has published resources that help educate the public on ways hackers can impact U.S. elections.
Malign actions taken by foreign governments or foreign actors designed to sow discord, manipulate public discourse, discredit the electoral system, bias the development of policy, or disrupt markets for the purpose of undermining the interests of the United States and its allies.
The initiative is part of Homeland Security’s #Protect2020 campaign to “enhance the security and resilience of election infrastructure, and to ensure the confidentiality, integrity, and availability of the free and fair elections foundational to the American way of life.”
“These efforts to open up the results of Federally-funded research promise to increase the return of Federal investments in scientific research, bolster the reliability of that research, accelerate scientific discovery, stimulate innovation, promote entrepreneurship, and enhance economic growth and job creation.”
In 2019, however, we’re still a long ways away from realizing innovation potential by creating a culture that defaults to immediate, open access of federally-funded research.
Case in point, a recent federally-funded grant that researched the impact of government-funded research on innovation, such as private-sector intellectual property and inventions, is inaccessible to most of the people who funded it.
“This research is an effort to detect, in a more nuanced way, the myriad fingerprints that U.S. federal research leaves, directly and indirectly, on innovation by others. We hope that it provides insights for the government, corporations, and citizens about where this funding goes and the downstream impact it has on innovation. And let’s not forget, that does not include the social and economic impact of federally supported research – but that’s for another day.”
The irony here is that it’s a government-funded report on the innovation impact of government-funded research, and it’s not open or made available for immediate consumption and collaboration.
While current standards make a one-year embargo acceptable from a policy perspective, if federally-funded research is to have optimal impact, it must be made accessible to everyone at the same time.
This particular research most likely won’t impact global innovation, however, it’s the principle of not opening it immediately that continues to foster arcane thinking of and treatment around information access.
For the taxpayers who funded this research and wish to see it, that’s for another year.
The Director of National Intelligence announced the creation of a new DNI executive position focused specifically on election threats. The new position will be led by Shelby Pierson, who previously served as DNI election security crisis manager during the 2018 midterm elections.
From the announcement:
The ETE will serve as the DNI’s principal adviser on threats to elections and matters related to election security. Additionally, the ETE will coordinate and integrate all election security activities, initiatives, and programs across the IC and synchronize intelligence efforts in support of the broader U.S. government.
“Election security is an enduring challenge and a top priority for the IC,” said ODNI Director Daniel R. Coats. “In order to build on our successful approach to the 2018 elections, the IC must properly align its resources to bring the strongest level of support to this critical issue.
CivStart is a new government-focused start-up accelerator that wants to ensure civic technology products “don’t get made in a vacuum — that they serve the needs of our most vulnerable and underserved communities.”
Co-founders Nick Lyell, Anthony Jamison and Sarah Kerner share their mission, and why they started CivStart.
What problem does CivStart want to solve?
In short, local and state governments are responsible for serving their communities in ways that have significant impacts on people’s lives, but have not always tapped into the best resources to do so. CivStart wants to help governments innovate their processes and tools by connecting them to effective new solutions.
We look at what we are solving from two different viewpoints, based on our audience; state and local government leaders and startups.
There are a couple issues we are addressing here at CivStart:
Identifying startup technologies that are providing solutions which address the challenges and issues state and local governments face on a day to day basis as they plan for the future. Whether that is understanding where your vulnerable populations are during a disaster so that you can deploy assets strategically or providing affordable transportation options to your communities so that they don’t have to take multiple bus lines. Our goal is to find these technologies and offer them to government leaders so that they can ensure that their communities are healthy, secure and vibrant.
Helping startups scale and enter the market the right way. We understand that startups have a mandate to grow and to grow fast (as we are startup ourselves). However, the state and local market is incredibly unique and complex to navigate for many large companies, let alone startups. A lot of business in this sector is won through relationship. Government decision-makers want to know that they can trust you, so selling to state and local requires a different approach than what a lot of these companies are used to when cornering the market. Startups need to know what the pressing issues are, and position their solutions in a way to address those challenges.
What was the inspiration for starting CivStart?
In our experience we’ve noticed:
Governments are often unaware of new technologies available to help them better serve their communities.
Many new companies don’t know how to navigate the public sector market and build relationships with the governments they want to help.
This inspired us to create a nonprofit that works with multiple stakeholders to bring these groups together and solve both issues.
What is CivStart looking for in its participating startups?
Of course, we want the biggest and brightest startups to be apart of our portfolio. However, working in the space that we work in, we can’t just be focused on the next best idea, solution, or service; instead, we seek startups who are solving real state and local problems and that we believe can have a real impact on improving people’s lives.
We try to prioritize our focus on access and opportunity for underserved and unconnected communities through health, public safety & emergency services, transportation & infrastructure workforce development, economic and community development, gender equity, civic tech, digital and financial inclusion ventures. One of our main organizational goals is to have our cohort members promoting gender and racial/ethnic diversity within the tech community.
How is CivStart supporting your portfolio companies?
CivStart helps startups forge meaningful connections with leaders in the public and private sectors to turn compelling technology into viable, scalable, solutions for the state and local space.
Each startup is in our program for 24 months, during which we’ll offer educational programming, facilitated mentorships and advisory relationships, and help cohort members build their networks in strategic ways.
We empower technology entrepreneurs to work with governments towards positive localized social and environmental change.
What does success look like for CivStart?
Success for us in many forms.
The obvious measure of success is the growth rate of our startups. We fail if our startups do not win market share; however, being an honest broker of solutions for state and local governments is also a key indicator of success.
We want governments, and the people that work with them and for them, to know that we are thinking of how we can strategically serve their needs and challenges when we engage with selecting startups technologies. They can come to us knowing that we put these startups through a program that emphasizes treating governments as partners and not just customers.
The three-hour event will be held in tandem with Drupal GovCon on Tuesday, July 23, 1-4 p.m. ET, at the National Institutes of Health in Washington, D.C. The agenda includes an overview of key open data concepts, case study presentations and breakouts.
Particularly pointed are his comments on open data portals, which haven’t innovated much on user experience since the first iterations created by private and nonprofit organizations. Kin doesn’t talk about the poor experience of data portals, but I think a substantive part of this failure is open data movement’s inability to capture the imagination and interest of the design community.
“Today, there are plenty of open data portals,” writes Kin. “The growth in the number of portals hasn’t decreased, but I’d say the popularity, utility, and publicity around open data efforts have not lived up to the hype.”
Secondarily, a truly sustainable, open community of open data leaders never materialized. Harvard’s Civic Analytics Network and GovEx are available, but largely inaccessible to the broader community.
Kin’s opinions are a little more anti-entrepreneurial and punk rock than mine, but it’s hard to have experienced the energy of bright technologists at the early stages of the open data movement and — seeing where it stands today — not think it’s all now extremely incremental in realizing its true potential.
Hopefully, those who consider themselves open data leaders will take the time to meditate on Kin’s thoughts and use them to reinvigorate the next iteration of the movement.
The National Science Foundation issued a statement admonishing governments that “endeavor to benefit from the global research ecosystem” and fail to uphold the agency’s values of openness, transparency and collaboration.
In the statement, NSF said:
“The values that have driven NSF and its global research partners for decades are openness, transparency, and reciprocal collaboration; these are essential for advancing the frontiers of knowledge.”
Actions NSF took to encourage governments to better cooperate with these values:
policy guidance for researchers on requirements to disclose foreign and domestic support
a study that will provide recommendations for NSF to better protect its merit review system and for grantee institutions to maintain balance between openness and security of scientific research
new policy requirement that NSF personnel employed can’t participate in foreign government talent recruitment programs “that may jeopardize the integrity of NSF’s mission and operations”
The U.S. Government Accountability Office published a bullish report on the impact cloud services has had on federal government agency technology savings.
Thirteen of 16 agencies reviewed indicated they saved $291 million between 2014 and April 2019 from cloud services use. Because of a lack of reporting standards or guidance, “it is likely that agency-reported cloud spending and savings figures were underreported,” GAO wrote in its analysis.