Rolling out some new Blogs
Deconstructing Deepfakes. How do they work and what are the risks?
Last month, Microsoft introduced a new deepfake detection tool. Weeks ago, Intel launched another. As more and more companies follow suit and more concerns arise about the use of this technology, we take a look in today’s WatchBlog at how this technology works and the policy questions it raises.
What is a deepfake?
A deepfake is a video, photo, or audio recording that seems real but has been manipulated using artificial intelligence (AI). The underlying technology can replace faces, manipulate facial expressions, synthesize faces, and synthesize speech. These tools are used most often to depict people saying or doing something they never said or did.
How do deepfakes work?
Deepfake videos commonly swap faces or manipulate facial expressions. The image below illustrates how this is done. In face swapping, the face on the left is placed on another person’s body. In facial manipulation, the expressions of the face on the left are imitated by the face on the right.
Deepfakes rely on artificial neural networks, which are computer systems that recognize patterns in data. Developing a deepfake photo or video typically involves feeding hundreds or thousands of images into the artificial neural network, “training” it to identify and reconstruct patterns—usually faces.
How can you spot a deepfake?
The figure below illustrates some of the ways you can identify a deepfake from the real thing. To learn more about how to identify a deepfake, and to learn about the underlying technology used, check out our recent Spotlight on this technology.
What are the benefits of these tools?
Voices and likenesses developed using deepfake technology can be used in movies to achieve a creative effect or maintain a cohesive story when the entertainers themselves are not available. For example, in the latest Star Wars movies, this technology was used to replace characters who had died or to show characters as they appeared in their youth. Retailers have also used this technology to allow customers to try on clothing virtually.
What risks do they pose?
In spite of such benign and legitimate applications like films and commerce, deepfakes are more commonly used for exploitation. Some studies have shown that much of deepfake content online is pornographic, and deepfake pornography disproportionately victimizes women.
There is also concern about potential growth in the use of deepfakes for disinformation. Deepfakes could be used to influence elections or incite civil unrest, or as a weapon of psychological warfare. They could also lead to disregard of legitimate evidence of wrongdoing and, more generally, undermine public trust.
What can be done to protect people?
As discussed above, researchers and internet companies, such as Microsoft and Intel, have experimented with several methods to detect deepfakes. These methods typically use AI to analyze videos for digital artifacts or details that deepfakes fail to imitate realistically, such as blinking or facial tics. But even with these interventions by tech companies, there are a number of policy questions about deepfakes that still need to be answered. For example:
- What can be done to educate the public about deepfakes to protect them and help them identify real from fake?
- What rights do individuals have to privacy when it comes to the use of deepfake technology?
- What First Amendment protections do creators of deepfake videos, photos, and more have?
- Deepfakes are powerful tools that can be used for exploitation and disinformation. With advances making them more difficult to detect, these technologies require a deeper look.
January 2021, CICO writerStaff Reporter Karen Howard
How To Be A Data-Driven Company: 5 Ways To Embrace Data
Gambling your company’s future value on intuition isn’t a sustainable strategy—but getting what you need out of data can seem overwhelming. Here’s how you can tackle data technologies to meet business goals.
In recent years, as terms like “big data” and “real-time analytics” have invaded the enterprise lexicon, numerous analyses have reinforced that data-driven companies tend to outperform competitors. A recent Deloitte survey, for example, states that “Organizations that reported having the strongest cultural orientation to data-driven insights and decision-making are twice as likely to have reported exceeding business goals in the past 12 months.”
Despite this seemingly obvious fact, many companies still rely on intuition and questionable judgement—rather than data-driven decisioning—for much of their problem solving.
This can happen for a number of reasons: cultures that prioritize praising “winners” rather than those who make the best decisions; IT operations that don’t make data accessible or easy to interpret; the constant pressure to ship; the tendency to privilege personal experience over data; and so on.
Despite the challenges, building a data-driven company is not out of reach. Asking the right questions—and getting the right data to the right people—is the hard part. In this article, we’ll examine five practical steps a company can take to create a data-driven culture.
1. Get the data flowing:
Newton’s first law of motion states that an object at rest will stay at rest unless acted on by an outside force. We should pattern a new first law of data after Newton’s observation, since corporate data in the enterprise works much the same way: Data at rest tends to stay at rest. Research has shown that data flow at the edge of a corporation’s network can predict financial performance—so it’s important to get data moving, and to create interactions around your data by getting it to the right people in the right contexts.
After all, if the value of shopping centers corresponds to the number of shoppers they attract, then digital interactions—the shoppers of the digital world—should drive increased value to the corporation in much the same way. The more digital footfall (i.e., the greater the usage of the application programming interfaces, or APIs, that connect data to applications and user experiences), the bigger the financial outcome should be. This is even further magnified by internal APIs, which predict sales more strongly than external APIs.
2. Make product decisions based on data:
This one seems obvious, right? One of the biggest challenges faced by large enterprises is that there is often no single source of data that can be leveraged. The data associated with customer interactions, for instance, must be aggregated from multiple elements—CRM systems, ads data, various application databases, log files, and more. Creating a single pane of glass—a focal point to aggregate and collect this data so that it is available to decision-makers—is a key step toward driving an objective decision-making process.
In the era of digital ecosystem competition, in which many services and customer experiences rely on technologies from multiple companies or platforms, single-pane access to actionable data is frequently married to not only data analytics products such as BigQuery, Google Cloud’s enterprise data warehouse, but also the API management platform. Since virtually all digital interactions pass through APIs, API management offers a single point at which all transactions and behaviors can be aggregated, analyzed, and leveraged, both through visualization tools and machine learning (ML) models.
3. Produce new data based on data:
The process—and promise—of delivering ML value to the enterprise is frequently challenged by the amount of data that needs to be used for training and then re-training. This is magnified by the emerging discipline of continuous delivery (CD) for ML modeling, in which models are constantly re-trained and released, producing even more data. It can be easy to focus simply on the volume and storage costs of data, but keep in mind the larger goals. Cost rationalization is important, but not at the expense of delivering new value to customers.
4. Put data in everyone’s hands:
Getting access to the right data isn’t just about APIs. Almost always, even if the appropriate APIs already exist, they are only available to developers who write software that consumes and leverages those APIs. A new set of tools, known as low-code or no-code platforms, democratizes access to data by targeting business users (sometimes referred to as “citizen developers”), empowering them to design, build, and deploy apps and business automations that harness the enterprise’s data.
This can help all parties that need to make data-driven decisions to do so, whether they are software developers, business analysts, operations managers, strategic negotiators, or virtually any interested party—and whether they need solutions for the office, for working from home, or for working in the field. Application design and development are no longer the exclusive purview of the developer, allowing anyone in the workforce to do more with data.
5. Lean in to strategic openness:
Many enterprises have historically shown reticence toward openness, keeping data within the corporate firewall and locking off access to digital assets. This is the equivalent of building corporate castles—replete with moats, drawbridges, and heavy castle walls—around the data: It might feel safe, but it makes it awfully hard to get the data moving. But what if the best security posture is actually to lean in to openness? Instead of locking down data, enterprises can build a security culture that identifies sensitive vs. non-sensitive information, provisions who is accessing what, and provides self-service access to data while still giving administrators the ability to restrict access to the right people.
This can help build a culture around developer (and citizen developer) enablement while still keeping data protected. In some organizations, data is siloed even between business units, impeding development and management of products that are sensitive to a 360-degree view of customer interactions. These impediments can lead to negative customer experiences, such as having multiple apps to log in to, multiple different usernames and passwords for a company’s apps, and even multiple different payment methods. For many enterprises, leveraging and contributing to an API catalogue is a great step toward changing the working model of a company from guarded, possessive, and nervous to collaborative, dynamic, and modern.
Becoming a data-driven company requires assessing an organization from multiple angles: infrastructure rationalization and developer engagement to strategic alignment, new KPIs oriented to encourage employees to do more with data, and potential network effects that may arise from expanded data efforts. The transformational business results that come from being a data-driven company won’t materialize from a single gesture, but the results can drive the growth strategies that sustain and propel a business forward.
December 2020, CICO writerStaff Reporter Dave Feuer
Employers Must Act Now To Mitigate The Impacts Of The Pandemic On Women’s Careers
It may be years before we comprehend the full ramifications of COVID-19 on our society and places of work. But while we are still learning to navigate the pandemic, we each have had to adapt our daily lives to respond to it. Working women, in particular, are being impacted in profound ways, facing tremendous challenges and commonly taking on expanded duties at home while continuing to juggle their careers.
In order to understand how and to what degree women’s day-to-day lives have changed – and how they feel these changes could impact their careers – we recently conducted a survey of nearly 400 working women around the globe at a variety of career levels and spanning various industries.
The pandemic is taking a heavy toll on the daily lives of working women
What these women shared sheds light on the extent to which the pandemic is affecting their work/life balance, mental and physical health, and confidence in their long-term career prospects.
Over 80% of the women we surveyed said their lives have been negatively disrupted since the onset of COVID-19. Additional caregiving responsibilities, extra household responsibilities, and heavier workloads were cited as common impacts, causing many women to experience negative tolls on their mental or physical well-being or feel unable to balance their work/life commitments.
Alarmingly, nearly 70% of women who have experienced these disruptions are concerned about their ability to progress in their career. And 60% questioned whether they actually want to progress when considering what they perceive is currently required to move up in their organization.
We should be concerned about these results in terms of the immediate impacts on women’s daily lives, the potential long-term effects on their future careers, and the broader threat to the progress made in recent years in achieving gender equality in the workplace. But our research also reveals how leaders can take action to mitigate these impacts.
Actions taken by employers will be critical in ensuring women continue to thrive
Our survey asked women what employers could do to support them in progressing during and beyond the pandemic. Using their answers and other insights from our research around key barriers and enablers, we believe there are six important steps organizations can take to ensure women continue to progress:
1) Make flexible working the norm. Going beyond “working from home” to offer a range of options that enable everyone (not just working parents) to have a manageable work/life balance is critical for making progress on gender equality. Of the 60% of women surveyed who said they questioned whether they want to progress in their organizations, more than 40% cited lack of work/life balance as a reason. Moreover, just under half of those surveyed cited having more flexible working options as something their employer can do to help them stay longer term. But this is not just about policies – these options must also be underpinned by a workplace culture that supports employees in taking advantage of them without any fear of career penalty.
2) Lead with empathy and trust. The need for leaders and managers to have open and supportive conversations with their teams has never been stronger, and 44% of women surveyed said that having more regular team check-ins to understand how individuals are doing is a key action leaders can take. Open dialogue can help leaders understand any short-term constraints their employees face and make sure their long-term prospects within the organization are secured.
3) Promote networking, mentorship and sponsorship as ways to learn and grow. 46% of women surveyed told us that the provision of such opportunities would entice them stay with their employer longer-term. These resources can be meaningful platforms for career growth, provided they are offered in ways and at times that accommodate different schedules and needs.
4) Create learning opportunities that fit within employees’ daily lives. With 40% of women saying they want more learning and development opportunities, introducing approaches to learning and development that provide access to expertise and skills in flexible and practical ways can be key to supporting women, many of whom remain keen to take on more responsibilities despite the constraints imposed on them by the pandemic.
5) Ensure that reward, succession, and promotion processes address unconscious bias. With over half of those surveyed citing getting a promotion and/or a pay raise as actions employers can take to make them stay longer-term, it remains critical that organizations address unconscious bias in their reward and succession processes. This includes looking at these processes in the context of remote working and addressing any negative perceptions of unavoidable commitments outside work, such as caregiving responsibilities.
6) Above all, make diversity, respect, and inclusion non-negotiables. Of those women who said they were questioning whether they wanted to progress in their organizations, around a quarter cited lack of diversity, poor or no role models, and poor culture, and 30% cited non-inclusive behaviors experienced (e.g., microaggressions, exclusion from meetings/projects) as reasons. Beyond having the right policies and processes in place to advance gender diversity, leaders must address these non-inclusive “every day” behaviors, such as microaggressions and exclusion, through clear and visible action since this is clearly still a significant factor to ensure women remain engaged.
We are at an inflection point. With no end to the pandemic currently in sight, organizations must meet the call to support the women in their workforce and ensure they can thrive both personally and professionally—or our economy and society could face long-standing repercussions.
November 2020, CICO writerStaff Reporter Emma Codd
Growing A Culture Of Innovation: 5 Lessons From Google
Organizations are facing unprecedented change and challenges stemming from a confluence of natural and artificial conditions. These forces are driving many to rethink the tools and technologies they use, and the places they need to be, to grow and to innovate. Below, Vinton G. Cerf, Vice President and Chief Internet Evangelist at Google, shares five lessons on growing a culture of innovation.
1. Sustained competitive advantage cannot be achieved with technology alone.
To create a more innovative business, you must rethink how people, structures, and processes interact every day—we refer to this as organizational culture. The teams you rely on to build must have systems and processes that keep them engaged, amplify their ability to produce, and keep them consistently forward-looking.
At Google, we’ve spent years thinking about how to maintain and improve a culture that fosters transformation and innovation. This has led to alignment around certain core principles that have informed our approach and supported Google’s culture for two decades.
2. Measure, make decisions, and be transparent in that process.
Measurement is at the heart of everything we do at Google. We measure everything—from how our systems are running, to how productive we are, to how people are feeling. All the data we gather is extremely valuable, because it exposes problems faster than if we simply scratched our heads and wondered. Once we gather that data, we still need to spend some time interpreting it, but at least we have a basis for judging how well our organizational structure is working.
It’s important to recognize that a feedback system only works when people believe changes will be made as a result of their feedback.
A culture of measurement results in a collection of anecdotal information as well as quantitative data. Both are necessary to inform change. We perform a number of different measurements—for example, encouraging everyone to participate in an anonymous employee satisfaction survey every year. That data and that feedback loop facilitate our decisions to change how we’re doing things, as needed.
Once we’ve gathered the data and made a decision, it’s time to actually put those changes into motion. It’s important to recognize that a feedback system only works when people believe changes will be made as a result of their feedback. So the trick is to ask the questions and then actually do something with the result.
Transparency is another important part of Google culture. It’s important that we be transparent about the feedback we heard, and how we went about addressing it. Being transparent as a company increases customer trust on one hand, and employee trust on the other. It’s important that people understand why we prioritized the changes we made. That’s core to the company’s DNA.
3. Don’t be afraid of failure.
Sometimes science learns more from failure than it does from success. If you ask why something didn’t work, you often learn more than you would have if it actually did work. And so, even at Google, we try a lot of things out that don’t work—and we learn from them and refine our practices. And eventually, we hope, we get to the point where the things that we want to work actually do work. Science is a lot like that. Google is a lot like that as well.
You have to have the willingness to allow failure. I’m not suggesting we should fail all of the time—that would be a problem! I’m talking about the freedom to try things out without absolute certainty of success. This is the fundamental difference between engineering and research.
With research, you don’t start off knowing the answer. With engineering, you think you know the answer, and you just have to build it. But what can happen with engineering is that you build it and then it doesn’t work. These two disciplines interact in the most wonderful ways. The engineer says, “I built it and it didn’t work.” The researcher says, “Why not?” And the engineer says, “I don’t know, can you help?” Together, they discover there’s a fundamental reason why this particular path for implementation didn’t work—and they learn from that. And then you get to develop a new design that takes this into account.
At Google, we’ll go down a number of different paths as we explore new capabilities in the system, and we often encourage people to go down these paths, even if they might end up at a dead end. And we share, blamelessly, with others the fact that there was a dead end, so everyone learns. That’s how we advance everybody’s ability to carry out their work.
4. Don’t forget that culture is always a work in progress.
Over time, as the mix of people joining the company changes and as the scale of the company gets bigger, we have to remind people about the cultural norms that we would like to maintain.
You have to periodically refresh the cultural elements that matter.
For example, one of the things that Google tries to accomplish is to give people the freedom to try things out, which resulted in a policy of allowing engineers to spend 20% of their time doing things that they weren’t originally assigned to do. People use 20% time to learn outside of their assigned duties and it actually acts as a stabilizing component of employee satisfaction.
The idea of 20% diminished for a while as we grew, until we reminded everybody that that 20% was fundamental to Google and was a cultural element that we wanted to maintain. It’s important to remember that you have to periodically refresh the cultural elements that matter.
5. Stay open.
If I were trying to give advice to an enterprise CIO, one of the things I would say is this: Don’t think that you have all the answers. In fact, the probability is very high that you don’t have very many of them at all. Take advantage of opportunities to share knowledge with your colleagues, your friends, even your competitors to better understand what others have learned in order to solve the same problems you have. Openness is your friend. The same thing is true when it comes to not taking all the credit. It’s important to acknowledge other people’s contributions because it gives them the incentive to continue contributing. And so this kind of openness of spirit is just as important as openness of ideas.
Technology alone does not guarantee success. You need a culture that supports change and acceleration—which paves the way for innovation. People have always powered technology, and today that’s especially true as teammates must collaborate and solve big problems together, even if they’re not in the same room. Fostering a culture of innovation helps lead to identification of new opportunities, and quick action to create new ideas and get ahead of the competition.