Summary of Part 3
Chapter 14: AI Won't Save Us (Unless We Save It First)
As our society becomes increasingly dependent on AI, a pernicious assumption is spreading that software will allow us to escape the influence of bias. This is often not the case. If we code fairness in, though, AI could help us build more equity-centered workplaces. Failure to do so can embed bias into these systems for years, or decades.
To ensure that automated intelligence is inclusive and equations are equitable, developers should answer questions that would determine whether their code meets standards of fairness:
- Who, or what, is missing? The categories that an algorithm allows for can also allow certain demographics to go statistically missing.
- Who is creating the model? Diverse teams are more likely to provide diverse insights, which decreases the odds of exclusive designs. In addition, to avoid building a model based on feedback tainted by the interviewer effect, match interviewers' demographics to those of your participants.
- Are you evaluating impact through an equity lens? Be aware of the point of view of your model and don't rely only on randomized control trials.
Even when you find a source of potential bias in your model, you can't always change the data set or circumstances you're working with. Two mitigations can help to counter this potential:
- “Blind” applications or resumes by removing biasing characteristics before an algorithm is applied to them.
- Instead of using a proxy for a missing data set, find data sets from outside your organization that describe the group you are trying to understand.
One key trait that separates harmless algorithms from “weapons of math destruction” is transparency. To foster transparency, include a “data biography” with algorithm-based models. This information should be available to anyone interacting with the model and should include the following:
- The elements of your equation: What + What = What?
- How the data is used.
- Who created the mathematical model.
- Data sets and specific definitions.
- How the model was evaluated (and reevaluated and adjusted).
Chapter 15: DEI Principles to Live By
In the face of rising AI, a rapidly transforming workforce, workspaces in transition, and other unknowns that we are yet to encounter, crafting a DEI strategy can feel daunting. How can you know that the policies you implement today will be relevant even a year from now?
When creating, implementing, and communicating your strategy, these 10 core DEI principles can help it survive the unknown:
1. Don't over-automate the underestimated. Pre-plan transition safety nets.
While AI means the elimination of some roles, it also means there will be increased demand for other skills. Identify underestimated individuals who have these skills, or train for them, starting now.
2. Don't create a recruiting strategy without a retention strategy.
Your recruiting strategy may attract employees from diverse backgrounds, but what is your strategy to retain them? Retention, equity, and inclusion strategies must address at least the following key elements:
- Performance evaluations that are standardized and safeguarded against bias.
- Pay equity ensured by pay gap audits, based on total compensation, and conducted over time, not just once.
- Promotion process that is transparent and open to all.
- Professional development/ongoing education with equal access to opportunities for employees of all backgrounds.
- Flexibility/leave/remote work policies determined with employee input and including paid leave for all genders.
- Vendor/procurement policy that includes awareness of vendors' DEI policies.
- DEI evaluation and reporting policy, incorporated into your holistic business strategy, that includes how you track and report progress toward DEI goals.
3. Act on data, not best guesses.
Your strategy should be driven by data, not best guesses. Crafting a data-smart DEI strategy means collecting data that answer the following two questions:
- What are the quantifiable differences between different populations' workplace experiences?
- What do employees, of all backgrounds, want?
4. Think long term and set expectations.
As you implement your DEI strategy, keep a long-term perspective. Remember that making up for what may be years, or decades, of institutional bias takes time.
5. DEI is not kickball. Work with data-smart experts.
DEI practitioners must have a range of skills and knowledge, from supply chain management to statistical analysis, to be successful. If you do not have a DEI expert in-house, find a qualified consultant—one who measures lead indicators—to work with you.
6. DEI does not belong to HR.
DEI is the job of everyone in the company. Make DEI part of every job description, and make sure that leaders and non-HR employees visibly communicate their approach to equity and inclusion.
7. Evaluate—and reevaluate—with an intersectional lens.
New forces that shape the future of work will affect underestimated groups differently. Transform your DEI strategy in alignment with an evolving workplace by finding out what's working, what's not, and adjust accordingly—with an intersectional lens.
8. DEI is a choice; it is not “in your DNA.”
Implying that DEI naturally occurs within your organization can make employees feel like their concerns will not be taken seriously and minimizes the effort that you have chosen to invest in your DEI strategy. Instead, communicate the steps you have taken to ensure that equity is integrated into your business practices.
9. DEI is not charity.
Be aware not to communicate about your DEI initiative in a way that implies it is like a charity campaign. Doing so can make underestimated employees feel like “problems” to be “fixed.” Your DEI implementation is about your company reaching its full potential.
10. You're going to mess up. It's okay—as long as you make it okay.
Odds are good that you will make missteps in your journey toward an equitable workplace. If you do, three steps are essential to helping everyone move forward:
- Take accountability: Name your mistake. The first step to regaining trust and confidence is showing that you see what you have done wrong.
- Acknowledge any harm done. Without apologizing for someone else's feelings, acknowledge any hurt you may have caused.
- Commit to (educated) action. In stating what steps you will take to do better in the future, it is important to note how you plan to educate yourself in doing so.
Chapter 16: Hold the Door
With the processes and methods in this book—the practical steps toward equity that you can take now—you are in a position to be an agent for change. When you open the door to a future of inclusion, others will follow.