Editorial
Research Report
The AI Transformation 100: Concrete Ideas for Fixing How We Work
The most consequential truth of the AI age is that autonomous systems only know what has been captured in data. Numerous studies of algorithmic bias demonstrate that when historical data excludes certain groups or misrepresents them, the resulting systems replicate and scale those distortions. If a contribution was never credited, never documented or never entered into the record, the AI will not know it existed.
When we train models on datasets that have already erased the contributions of innovators and change-makers, we reproduce that erasure, automate and scale it.
This is what I call the Double Death.
Table of Contents
- Double Death: Erased Once in Life, Again in Data
- The Difference Between Stepping Back and Being Pushed Out
- How Recognition Becomes a Gatekeeper to Opportunity
- The Damage of Private Praise and Public Silence
- The Organizational Blind Spot AI Can Make Permanent
- The Problem With Auditing Erasure After the Fact
- Who Gets to Exist in AI’s Version of the Future?
Double Death: Erased Once in Life, Again in Data
The first death is being made invisible in the present, through social exclusion, professional omission or the quiet decision not to credit. The second death is being erased from the future, because the training data for the next generation of intelligence contains no record that you were ever there.
The first death can sometimes be contested, corrected or recovered from. But the second cannot. Once the data moves forward without you, the systems built on it will not find you. They will not miss you because they will not know how to look.
We often mistake invisibility for neutrality. It is not. Scholarship (found here and here) on structural power has long shown that what appears neutral often reflects embedded social choices about whose experiences are recorded and whose are ignored. If the same people are repeatedly counted in and others are repeatedly left out, that behavior eventually becomes part of the structure. Research (found here and here) shows that in AI environments, this process accelerates from interpersonal slight to permanent, automated erasure from history itself, because algorithmic systems reproduce the patterns present in the data used to build them.
Related Article: Colonialism in Code: Why AI Models Speak the Language of Empire
The Difference Between Stepping Back and Being Pushed Out
There is a fundamental distinction between being invisible by design and being silenced by force.
Think of the roles of an author and theater director. Both create worlds and shape narratives from behind the scenes, then step back to let others take the stage. Their invisibility is chosen. The essence of that role is precisely the act of giving the stage to someone else; part of the work is the disappearance.
In organizational dynamics, the paradox deepens when presence can no longer be withheld without cost, or when the work demands a specific voice and that voice steps forward. What can follow, in some groups and cultures, is an instruction to diminish. Then, labels arrive quickly and may accumulate slowly: too loud, too much or too busy.
What makes this even more damaging is when it filters through the organization. Other team members hear it, repeat it and reinforce it, until what began as one person's discomfort with another's presence becomes a collective behavior. At that point it is no longer a personality conflict. It is workplace bullying and can lead to organizational trauma. That is what can be encoded into the business decision-making culture long before it is ever encoded into a system.
How Recognition Becomes a Gatekeeper to Opportunity
We rarely acknowledge that these characterizations say more about the observer than the observed. The person may not have been too loud at all. The observer just did not want them to be heard or to occupy that space and found language to justify their demand to shrink.
When those labels are believed, internalized and acted upon, the result can be systemic self-erasure: a person denying themselves the right to exist fully within the systems they helped build.
Invisibility carries a direct transactional cost. Sociological research on recognition and institutional credit shows that visibility is often a prerequisite for professional legitimacy, promotion and long-term opportunity. This is not only a matter of dignity; it is a matter of material consequence.
Consider the wedding planner who coordinates every detail of their client’s special day: who wakes the bride, who ensures the dress is pristine, who resolves the crisis at the venue before anyone else knows there was one. At the reception, the public thanks go to everyone visible. The planner is omitted.
When the extra work of an individual is consistently not seen, it can stunt their reputation and close future opportunities before they open. Most people require evidence of their accomplishments to build credibility: a reference, a credit, a public acknowledgement. To deny that evidence is to affect their livelihood and their sense of what they are capable of and what they deserve.
The Damage of Private Praise and Public Silence
One of the most insidious forms of this dynamic is the public-private divide: the person whose contributions are recognized enthusiastically behind closed doors, in quiet conversations between decision-makers, while in public spaces they are treated as if they do not exist. They watch others celebrated for equivalent work, enjoying the perks that come with recognition of worth. The impacted individual may remain in the struggle because the systems of credit were not built to support them fairly.
Invisibility is also weaponized personally when used to avoid confronting our own inadequacies by diminishing others' achievements. We can find reasons to attribute hard work to luck or to circumstance or to perceived advantage, rather than acknowledging the genuine effort and cost behind someone's success.
For some, this allows them to stay comfortable without having to face what their accomplishments mean and what we desire for ourselves. That individual could be an ally or resource, which would mean a loss of growth opportunity on both ends.
The Organizational Blind Spot AI Can Make Permanent
At scale, this becomes something more serious than individual envy. Technological tools amplify erasure rather than amplify voices when the underlying datasets already reflect social hierarchies and exclusions. Psychological wounds become weapons turned outward, used to make others smaller so that our own size feels adequate. Invisibility and competition, operating together, function as forms of systemic violence, not visibly or dramatically, but cumulative and structural in their effect.
For leaders, this is a risk inventory. When exclusion filters through an organization and technology scales it, we are facing Epistemic Violence, the algorithmic erasure of legitimate expertise. We are also generating Invisible Suffering that digital dashboards will not automatically show. Human Scaffolding Gaps leave our workforces unprotected just when AI integration requires the most from them.
And, remember, if you are deploying governance frameworks on top of a culture that is already erasing people, you are not managing risk. You are masking it. Many organizations will not see the problem coming. Complaints, investigations and the departure of key staff often come long after the damage has already been done. They are the consequences of decisions that felt inconsequential at the time.
The Double Death is not a future risk; it is a present one. Think about every uncredited contribution, every omitted name and every dataset built from sources that belong to an individual. When the data is uploaded into an AI system without that uncredited, omitted and inaccurate data source, the second death begins to take shape.
Related Article: Cashmere Economics: When Algorithms Determine Opportunity
The Problem With Auditing Erasure After the Fact
We cannot audit our way out of an erasure that wasn’t even documented as an error. We cannot correct absences that the system wasn’t designed to identify.
Webinar
Jun
25
The Hidden Cost of Fragmented Customer Communication
Discover why growing businesses are rethinking the systems, workflows and communication habits shaping customer experience.
Register
Webinar
Jun
30
How Modern Marketing Is Exposing the Limits of Legacy CMS
Why marketers are rethinking CMS workflows that slow publishing, personalization and campaign execution.
Register
Webinar
Jul
9
Why Some Dealers Are Pulling Ahead With AI
AI results vary. Learn where dealers win and where they get stuck.
Register
Webinar
Jul
14
Content Leaders Collective: Proving Content’s Business Impact
Join us as top content leaders look beyond the buzzwords to share how they actually prove ROI and scale what works.
Register
Research shows a governance approach that treats lived human experience is required, particularly insights from communities historically marginalized by technological systems, as essential evidence in governance and model evaluation. This means testing models on sanitized benchmarks and against real names, real accents, real histories and the real patterns of omission that have shaped the data those systems were built on and for. It means asking, before deployment, whose contributions a system will include, and whose it will categorize as invisible by design.
Emerging governance tools like model cards, algorithmic impact assessments and independent auditing frameworks have been proposed (found here, here and here) to improve transparency and accountability in automated systems.
Who Gets to Exist in AI’s Version of the Future?
People and communities that have already experienced the first death are uniquely positioned to identify where the second death is being built into an infrastructure. Their knowledge is diagnostic. Governance frameworks that fail to incorporate this insight can continue to automate the same erasures they claim to prevent.
Invisibility was never neutral, and in the era of AI automation, it can be permanent. The decisions being made now — about what counts as data, whose work is credited and whose experience is treated as a signal rather than noise — are decisions about who will exist in the future that AI is helping to build.
That is a governance responsibility, and it belongs to all of us.
Learn how you can join our contributor community.