To follow up from last week, much of my work is shifting away from how work gets done.
Part of that shifts involves responsibly maintaining human-in-the-loop quality control. So again, I get to explore which parts of knowledge work are shifting to AI and which are staying with me. What’s different this week is I’ll be exploring ways to empower AI to improve the how of work.
I asked ChatGPT AI what traditionally human skills it excels at and is getting better at engaging than humans.
Here’s part of what it said:
Here’s a breakdown of how I can support each area:
-
CollaborationI help teams coordinate ideas, align priorities, summarize discussions, clarify responsibilities, and maintain shared context across projects and workflows.
-
CommunicationI can adapt tone, structure, and complexity for different audiences — from executives and clients to technical teams and general users — while improving clarity and reducing ambiguity.
-
Creating, devising, and simulating change management plansI can help design rollout strategies, stakeholder communication plans, training approaches, adoption roadmaps, resistance-mitigation strategies, and scenario simulations for organizational or operational change.
-
Creative thinkingI generate original ideas, alternative approaches, brainstorming prompts, naming concepts, strategic pivots, and unconventional solutions by combining information across domains.
-
Emotional intelligenceI can recognize emotional tone in communication, suggest empathetic responses, improve conflict-sensitive messaging, and help tailor interactions to different personalities and stakeholder concerns.
-
Ethical reasoningI can evaluate tradeoffs, identify potential harms or biases, compare decisions against ethical principles or policies, and support responsible decision-making frameworks.
-
Finding solutions for well-defined problemsI excel at analyzing structured problems, identifying constraints, evaluating options, and producing step-by-step solutions, workflows, or recommendations.
-
Information literacyI can organize, summarize, compare, and synthesize large amounts of information while helping assess credibility, relevance, consistency, and gaps in sources or arguments.
-
Logical reasoningI can identify patterns, analyze cause-and-effect relationships, test assumptions, evaluate arguments, and work through structured reasoning processes.
-
Monitoring change processesI can help define KPIs, track milestones, analyze feedback, identify adoption risks, monitor implementation progress, and recommend adjustments during transitions or transformations.
-
Project managementI assist with planning, prioritization, task breakdowns, timelines, documentation, meeting summaries, stakeholder updates, workflow design, and operational coordination across projects.
Next, I asked what ChatGPT thinks of its own limitations.
Here’s part of what it said:
Collaboration
I can coordinate information, but I do not truly participate in human relationships or organizational dynamics. Humans should verify alignment, resolve interpersonal tensions, assess political implications, and make final decisions about priorities and tradeoffs.Communication
I can draft strong messaging, but I may miss cultural nuance, hidden sensitivities, sarcasm, legal implications, or context not explicitly provided. Humans should review high-stakes communications before sending.Creating, devising, and simulating change management plans
I can model likely outcomes and generate structured plans, but I cannot fully predict human behavior, organizational resistance, market shifts, or operational realities. Humans should validate feasibility, stakeholder readiness, timing, and business constraints.Creative thinking
I can generate many ideas quickly, but novelty does not guarantee usefulness, originality, or strategic fit. Humans should curate ideas, assess practicality, and ensure alignment with goals, audience, and brand identity.Emotional intelligence
I can recognize common emotional patterns in language, but I do not actually feel emotions or fully understand lived experience. I may misinterpret subtle emotional cues or interpersonal context. Humans should handle sensitive conversations, conflict resolution, counseling, and relationship decisions.Ethical reasoning
I can compare ethical frameworks and identify risks, but ethics often depend on human values, legal standards, culture, and context-specific judgment. Humans must make final accountability decisions, especially in high-impact situations.Finding solutions for well-defined problems
I perform best when the problem is clearly defined and the data is reliable. If assumptions are flawed, requirements are incomplete, or real-world conditions differ, my solutions may fail. Humans should verify assumptions, test outputs, and validate implementation.Information literacy
I can summarize and synthesize information, but I may sometimes rely on incomplete, outdated, biased, or incorrect sources. I can also present plausible-sounding inaccuracies confidently. Humans should fact-check critical information and evaluate source quality independently.Logical reasoning
I can reason through structured problems, but I can still make logic errors, overlook edge cases, or follow faulty premises to flawed conclusions. Humans should review reasoning chains in important analytical, scientific, legal, or financial contexts.Monitoring change processes
I can help track metrics and patterns, but I cannot directly observe real-world operations unless humans provide accurate updates and data. Humans should interpret qualitative signals, verify metrics, and intervene when unexpected issues emerge.Project management
I can organize plans and workflows efficiently, but I cannot enforce accountability, negotiate priorities, manage people directly, or respond autonomously to real-time operational disruptions. Humans should remain responsible for leadership, execution oversight, resource allocation, and final approvals.An effective overarching principle is:
“AI can accelerate judgment support, but humans must retain judgment responsibility.”