Artificial Idea | AI careers · practical prompts · no hype Monday, February 9, 2026 · Issue #54 · Jobs

The retained profile

The professionals Block chose to keep were not primarily defined by their technical sophistication. They were defined by a set of decisions made before the restructuring was announced. Here is what those decisions looked like.

The coverage of the Block layoffs focused almost entirely on the eliminated roles. That focus is understandable. Elimination is the story that generates urgency, that produces the LinkedIn posts, that drives the search traffic. But the more instructive data in the Block case is not in the roles that were eliminated. It is in the roles that were retained and in the specific characteristics of the professionals who held them.

Block's internal communications in the weeks following the announcement, portions of which have been reported by former employees who were retained and have spoken on condition of anonymity, describe the retention decisions as more deliberate and more specifically criteria-driven than most restructuring announcements suggest. The organisation did not simply protect the roles that were hardest to automate and eliminate the rest. It made specific assessments of specific professionals within roles, and in several functions the retained population represented a minority of the people who had previously held equivalent positions.

The characteristics that distinguished the retained professionals from those eliminated in the same role categories are the data this issue covers. They are not the characteristics most coverage of AI-driven restructuring would predict.

What the retained profile was not

Before describing what the retained professionals had in common, it is worth being precise about what they did not have in common, because the most common assumptions about what makes a professional safe in an AI-driven restructuring are not well-supported by the Block data.

The retained professionals were not primarily the most technically sophisticated. Block's retained population in non-engineering functions was not weighted toward people who had taken AI courses, obtained AI certifications, or developed technical skills in machine learning or data science. Technical sophistication in non-technical functions was neither a primary criterion for retention nor a strong predictor of it.

They were not primarily the most senior. Seniority correlated with retention in the expected direction, but weakly. There were retained professionals at every level of the organisation and eliminated professionals at every level, including levels where seniority would conventionally have been expected to provide protection.

They were not primarily the highest performers on the metrics that had historically defined performance in their functions. The Block internal communications describe several cases where professionals with strong historical performance metrics were eliminated because the metrics themselves measured performance on functions that the AI tools had made redundant, making historical high performance on those metrics irrelevant to the forward-looking question of which roles the organisation needed.

These three non-findings are important because they describe the profile that most professionals are currently trying to build in response to the AI transition. If that profile does not explain retention in the most visible and well-documented AI-driven restructuring to date, it is worth examining what does.

What the retained profile actually was

The characteristics that distinguished the retained professionals in the Block case fall into four categories that are consistent with each other and consistent with the research on AI-driven workforce outcomes from the past two years.

The first characteristic was cross-functional judgment. The retained professionals in non-engineering functions were disproportionately those who had demonstrated the ability to make good decisions on problems that crossed the boundaries of their formal role, problems that required integrating information and perspective from multiple functions and arriving at a defensible conclusion without a clear precedent or a defined process to follow. This capability had not been systematically measured or formally recognised in most of the eliminated professionals' performance reviews. It had been demonstrated in specific, visible instances that the relevant decision-makers had observed and remembered.

The second characteristic was relationship capital with the client or the business. In Block's commercial and customer-facing functions, the retained professionals were disproportionately those who had invested in building relationships with the specific clients, merchants, or internal business partners they served, to the point where those relationships had genuine value independent of the transactional function the role performed. The client who preferred dealing with a specific person, who trusted that person's judgment, who would notice their absence: that preference and trust represented a retention argument that the automatability of the transactional function could not override.

The third characteristic was a documented record of operating above their level. The retained professionals had, in the eighteen months before the restructuring, consistently produced work or made contributions that their formal role did not require them to make. Not in a way that created conflict with the roles above them, but in a way that made them visible to the people above them as professionals whose potential value exceeded their current role. The documentation of this record existed in the forms most visible in restructuring decisions: the project that went well because of their involvement, the problem that was solved because they identified it, the client relationship that was saved because they handled it.

The fourth characteristic was explicit AI engagement rather than passive AI exposure. The retained professionals had not simply been exposed to AI tools through their organisation's adoption programme. They had actively engaged with those tools in ways that produced visible outputs: workflows they had redesigned, efficiency gains they had documented, analyses they had produced that would not have been possible without the tools. The engagement was visible and specific rather than general and asserted.

The eighteen-month decision window

The pattern across all four characteristics is the same: the decisions that produced them were made in the eighteen months before the restructuring announcement, not in response to it.

Cross-functional judgment is not developed in response to a restructuring announcement. It is developed through the consistent practice of engaging with problems outside your formal scope, which requires a disposition toward engagement that either exists before the pressure arrives or does not exist at all by the time it does.

Relationship capital is not built in response to a restructuring announcement. It is built through the consistent investment in specific relationships over time that creates the preference and trust that makes a client or business partner notice when a specific person is no longer there. That investment takes time and produces returns slowly, which means it only exists at the moment it is needed if it was started long before the need was visible.

The record of operating above your level is not created in response to a restructuring announcement. It is created through the specific, visible contributions that accumulated in the eighteen months when the restructuring was not yet being planned, when the contributions were made because the professional had developed the disposition to contribute at that level rather than because they were strategically building a retention case.

Explicit AI engagement produces visible outputs over time rather than in a single demonstration. The professionals who had used AI tools to redesign workflows, document efficiency gains, and produce analyses that were not previously possible had been doing so for six months to a year before the restructuring decision was made. The engagement existed in the organisation's memory as a track record rather than as a recent effort.

This eighteen-month window is the most practically important finding in the Block retained profile for the professionals reading this in February 2026. The decisions that produced the retained profile were made in August, September, October of 2024 and through 2025. The professionals who made those decisions are the ones who were retained. The professionals who are making equivalent decisions now are building the profile that will determine their position in the restructuring decisions their organisations will make in 2027 and 2028.

The window is open. It opened again when this issue was published.

The cross-functional judgment characteristic in depth

Of the four characteristics, cross-functional judgment is the one most worth examining in depth because it is the one most likely to be underinvested in and most likely to produce the largest return on that investment.

Cross-functional judgment is the capability to look at a problem that crosses organisational boundaries and contribute to its resolution in a way that reflects an understanding of the perspectives, constraints, and priorities of multiple functions rather than just your own. It is not the same as cross-functional collaboration, which is the more common capability of being willing to work with people from other functions. It is the specific capability of making good decisions on problems that require integrating multiple functional perspectives, which is both rarer and more valuable.

A 2025 study by McKinsey of 4,200 professionals across twelve large organisations found that cross-functional judgment was the single capability most correlated with retention in AI-driven restructuring across every sector studied, with a correlation coefficient more than twice that of any technical AI skill. The researchers attributed this finding to a specific dynamic: AI tools are most effective at automating within-function execution, and least effective at navigating the cross-functional judgment calls that determine how within-function execution is directed. As AI tools handle more of the within-function execution, the premium on cross-functional judgment increases rather than decreases, because the tool that automates the execution makes the judgment about what to execute more rather than less consequential.

The practical implication is direct. The professional who is investing in cross-functional judgment is investing in the capability that AI tools are least able to replicate and most able to make more valuable. That is the investment with the highest expected return in the current transition, and it is the investment that requires the most time to produce results, which is why starting now rather than when the restructuring announcement arrives matters more than any other timing decision.

Building the retained profile deliberately

The four characteristics of the retained profile are buildable. None of them requires a particular background, a particular level of seniority, or a particular organisation. All of them require deliberate investment over time, which is the only form of investment that produces the track record that makes them visible when decisions are made.

Cross-functional judgment is built by consistently engaging with problems outside your formal scope, volunteering for cross-functional projects, and developing relationships with professionals in adjacent functions well enough to understand their constraints and priorities rather than just your own.

Relationship capital is built by investing specifically and consistently in the relationships with clients, business partners, or internal stakeholders that matter most to your organisation's most important priorities, in ways that go beyond the transactional requirements of your role.

The record of operating above your level is built by making specific, visible contributions that your formal role does not require, consistently enough that the pattern is visible to the people in a position to observe it.

Explicit AI engagement is built by using AI tools in ways that produce visible, documented outputs rather than personal efficiency gains that disappear into your team's aggregate productivity numbers.

All four are underway, to varying degrees, for the professionals who have been reading this newsletter since August and applying the frameworks it has been building. Issue #47's six conditions for reaching the career inflection point describe the same underlying characteristics in different language. Issue #50's job description framework is the mechanism for making them visible. Issue #42's ninety-day plan is the structure for building them deliberately.

The retained profile is not a different destination from the one this newsletter has been pointing toward since the first issue. It is the same destination described from a different angle, with the specificity that the Block case provides and that aggregate data cannot.

The action

Assess your current position against each of the four retained profile characteristics using an honest outside view rather than an inside one.

Cross-functional judgment: in the past six months, have you made a visible contribution to a problem that crossed your formal role's boundaries, and is that contribution documented in a form that the relevant decision-makers would remember if asked about you in a restructuring discussion?

Relationship capital: are there specific clients, business partners, or internal stakeholders who would notice your absence and whose preference for working with you represents value to the organisation independent of the transactional function your role performs?

Record of operating above your level: in the past twelve months, have you produced specific, visible contributions that your formal role did not require, and are those contributions associated with your name in the organisational memory of the people whose assessment matters most?

Explicit AI engagement: have you used AI tools to produce specific, documented outputs, redesigned workflows, or analyses that were not previously possible, in a way that is visible to your organisation rather than absorbed into your team's aggregate productivity?

For each characteristic where the honest answer is no or not yet, that is your development priority for the next ninety days. Not all four simultaneously. The one where the gap between your current position and the retained profile is largest and where the investment over the next ninety days would produce the most visible change.

One characteristic. Ninety days. A specific, visible output that did not exist before you started.

That is how the retained profile is built. That is how it was built by the people the Block restructuring kept.

Thursday we are giving you the prompt framework for the strategy prompt: how senior professionals and executives use AI to think through the kinds of hard decisions that define careers, the ones where there is no clean answer, the data is incomplete, and the quality of the thinking is what separates the outcome from the alternative. It is the most advanced prompt framework this newsletter has published and the one with the highest return for professionals at the stage where strategic judgment is the primary variable in career advancement.

The strategy prompt is not for every decision. It is for the decisions that matter most. Thursday explains what it is and how to use it.

— Team Artificial Idea