[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fXyVQ5xMNjHbYCuzmjY84osWtzIe0xaPDpzVkKrZV16c":3},{"item":4},{"id":5,"idKnowledge":6,"idDomain":7,"idCluster":8,"kindOverride":9,"slug":10,"title":11,"description":12,"bodyMarkdown":13,"bodyHtml":14,"author":15,"date":16,"createdAt":17,"topics":18,"image":29,"hasDownload":30,"fileName":9,"youtubeId":29,"domainCrumb":31,"clusterCrumb":34},"209","AB2C75C6-79ED-FF48-9C7C-919786388F6F","9338921B-ED02-F043-A322-3D31F60B951F","DD55E419-2BF7-5A47-8EF3-F919AD7E4FE2","","how-ai-changes-the-role-of-managers","How AI changes the role of managers","AI is reshaping management from task supervision to strategic judgment. Here's what that shift looks like in practice — and how to lead through it.","Your team is spending hours every week pulling data, formatting reports, and chasing down numbers that a machine could surface in seconds. Meanwhile, the decisions that actually require human judgment — how to respond to a struggling employee, whether to enter a new market, how to handle an operational crisis — keep getting pushed to \"when there's time.\" There never is. This article maps exactly how AI is redistributing the work of management, what responsibilities fall away, what new ones emerge, and how you can navigate the transition without losing your team in the process.\n\n## What work does AI actually take off a manager's plate?\n\nThe honest answer: more than most managers expect, and less than AI vendors claim.\n\nAI tools are genuinely good at high-volume, pattern-based work that used to eat management time:\n\n- **Automated reporting**: Instead of a logistics manager spending Friday afternoon compiling a weekly performance report from four spreadsheets, the system generates it automatically — on-time delivery rate, cost per shipment, driver utilisation — and flags anything outside tolerance. The manager opens a summary, not a raw dataset.\n- **Anomaly detection**: A production manager no longer needs to scan hundreds of output rows to find the line running 12% below target. The AI flags it, often before end-of-shift.\n- **Demand forecasting**: A retail operations manager who used to build next-quarter stock projections by hand — adjusting for seasonality, promotions, and supplier lead times — now reviews and challenges a model-generated forecast instead of constructing one from scratch.\n- **Routine recommendations**: An AI system trained on historical order patterns can recommend reorder points, flag a supplier's declining reliability score, or suggest which customer accounts are at churn risk. The manager decides what to do about it.\n\nThe shift is from *producing* information to *judging* information. That sounds like a small change. It is not.\n\n\n\n## What responsibilities grow — and get harder?\n\nWhen AI handles the operational layer, management bandwidth does not simply free up and sit idle. It gets redirected to tasks that are structurally harder:\n\n### 1. Validating AI-driven insights\n\nAn AI forecast is only as good as the data and assumptions behind it. Managers must now ask: *Is this model seeing what I know about this market?* A demand forecast built on last year's data will not reflect a competitor who just entered your territory. A churn-risk score built on purchase frequency misses the customer who paused buying because of a temporary budget freeze, not dissatisfaction.\n\nManagers who simply trust AI output without interrogating it will make worse decisions than managers who never had AI at all — because the false confidence is invisible.\n\n### 2. Owning accountability when AI is wrong\n\nA classic operational error has a clear owner. An AI-assisted decision is murkier. When the AI-recommended staffing plan leaves you understaffed during a peak weekend, who is accountable? The manager who approved it. AI does not absorb accountability — it transfers it, silently, to whoever signs off.\n\nThis means managers need a new habit: *explicitly documenting why they agreed with or overrode an AI recommendation*, so that learning accumulates and accountability stays visible.\n\n### 3. Coaching employees through uncertainty\n\nTeam members who used to own the weekly report now feel displaced. The analyst who spent Tuesday building the revenue model watches AI do it in four minutes. That is a real identity crisis, not just a workload adjustment. Managers are now expected to coach people through a redefinition of their value — which requires emotional intelligence, not just project management skill.\n\n### 4. Continuous process improvement\n\nAI surfaces patterns. It does not fix the processes behind them. If the anomaly detection keeps flagging the same production bottleneck, someone has to own redesigning the workflow. That someone is the manager, working with their team — not the algorithm.\n\n### 5. Change management at every level\n\nImplementing AI tools into an existing team is a change management exercise, not an IT project. Resistance, confusion, and quiet workarounds are the norm, not the exception. Managers who treat AI rollout as \"the tech team's problem\" watch adoption fail silently.\n\n## What does a \"new\" manager's week actually look like?\n\nHere is a concrete before\u002Fafter for a mid-sized operations manager at a manufacturing company:\n\n**Before AI integration:**\n- Monday: compile weekend production report from three systems (3 hours)\n- Tuesday: prepare supplier performance review manually (2 hours)\n- Wednesday: attend KPI meeting where half the time is spent correcting data\n- Thursday: field team questions about why targets were missed\n- Friday: build next week's shift schedule based on gut feel and last week's attendance\n\n**After AI integration:**\n- Monday: review auto-generated weekend report, investigate the two flagged anomalies (45 minutes)\n- Tuesday: challenge the AI-generated supplier risk scores with market context the system doesn't have\n- Wednesday: KPI meeting focused entirely on decisions, not data validation\n- Thursday: one-on-one sessions with team members on skills development and process ideas\n- Friday: review AI-proposed shift schedule, adjust for the two things the model doesn't know (an employee's upcoming training day, a known peak from a trade show)\n\nThe volume of *hours* may look similar on paper. The cognitive mode is completely different.\n\n\n\n## How do you validate AI-driven insights without second-guessing everything?\n\nThis is the practical skill gap most managers have right now, and almost nobody is training for it. Here is a working framework:\n\n1. **Ask where the data came from.** Is it real-time or a batch update from 48 hours ago? Does it include all relevant data sources, or just the ones that were easy to connect?\n2. **Ask what the model was optimised for.** A forecasting model optimised for average accuracy will systematically underperform at the extremes — exactly where you need it most.\n3. **Ask what it cannot see.** Competitor moves, regulatory changes, one-off events, and relationship context are usually invisible to the model. Add that context before acting.\n4. **Test the recommendation against your own judgment.** If the AI says \"reduce inventory by 15%\" and your sales team is telling you a large order is coming, that tension is worth investigating — not suppressing.\n5. **Document your decision.** \"I approved the AI recommendation because X\" or \"I overrode it because Y\" — three sentences in your project log. This builds organisational learning over time.\n\n## How should managers coach their team through this shift?\n\nThe employees most at risk are not the lowest performers — they are the analytical, detail-oriented people who built their identity around being the ones who knew the numbers. Here is what actually helps:\n\n- **Name the shift explicitly.** Tell your team: \"Your value is no longer in producing data. It is in knowing what to do with it.\" Most managers assume their team understands this. Most teams do not.\n- **Reassign, don't just remove.** When AI takes over report generation, give the analyst a new owner role — reviewing model assumptions, catching edge cases, communicating findings to stakeholders. The role evolves; it doesn't disappear.\n- **Celebrate overrides, not just approvals.** When a team member spots that the AI forecast is wrong because of a market factor the model missed, that is a win. Recognise it visibly, or people will stop questioning the machine.\n- **Invest in prompt literacy.** The ability to ask an AI system the right question — with the right context, constraints, and format — is now a core operational skill. Train for it deliberately.\n\n## What new skills do managers need to develop?\n\nThe capability gap is real. Managers who thrive with AI share a recognisable profile:\n\n- **Critical data literacy**: not knowing how to build a model, but knowing how to challenge one\n- **Judgment under ambiguity**: acting decisively when the AI gives a probability, not a certainty\n- **Contextual intelligence**: holding the organisational and market context that no dataset can fully encode\n- **Psychological safety as a leadership practice**: creating an environment where people say \"I think the AI is wrong here\" without fear\n- **Accountability discipline**: owning outcomes from AI-assisted decisions as fully as from manual ones\n\nNone of these are new concepts. AI makes them load-bearing in a way they were not before.\n\n## Checklist: Is your management team ready for AI-augmented operations?\n\n- [ ] Managers can articulate *why* they agreed with or overrode an AI recommendation — not just what they decided\n- [ ] AI outputs are regularly challenged in team meetings, not just presented\n- [ ] Team members whose roles changed due to AI have been given new, meaningful responsibilities\n- [ ] There is a documented process for escalating suspected AI errors\n- [ ] Managers are spending measurably less time on data collection and more on exception handling and people development\n- [ ] AI recommendations are treated as inputs to decisions, not as decisions themselves\n- [ ] Change management support was part of your AI rollout, not an afterthought\n\n## FAQ\n\n**Will AI make managers redundant?**\nNot in the foreseeable future — but it will make *some managers redundant*. Specifically, managers whose primary value was coordinating information flow and producing reports. Managers whose value lies in judgment, accountability, and people development become more important, not less.\n\n**How do I get my team to trust AI without over-trusting it?**\nModel the behaviour yourself. Ask questions about AI output in public. Override a recommendation in front of the team when you have good reason to, and explain why. Trust calibration is a cultural practice, not a policy.\n\n**What if our AI tools are giving us bad recommendations?**\nTreat it as a data quality and model design problem, not a technology problem. Bad recommendations usually trace back to incomplete data, misaligned optimisation targets, or missing business context. Fix those, rather than either abandoning the tool or ignoring the bad outputs.\n\n**How does this connect to broader AI strategy in our organisation?**\nManager-level AI adoption rarely happens in isolation. It works best when it is part of a deliberate design of [how people and AI collaborate across the organisation](https:\u002F\u002Floggix.com\u002Fen\u002Fblog\u002Fhow-to-design-effective-collaboration-between-people-and-ai) — with clear boundaries, escalation paths, and feedback loops built in from the start.\n\n**Is this different for small vs. large companies?**\nThe principles are the same; the stakes are different. In a 12-person company, the manager *is* the team, so the shift to judgment-focused work is even more immediate. In a large organisation, the challenge is more about scaling the cultural change consistently across management layers.\n\n---\n\nFor organisations where these decisions are tied to real operational systems — ERP data, production logs, customer records, supplier feeds — the quality of AI-generated insights depends entirely on the quality of the underlying data infrastructure. If your systems are fragmented, your AI recommendations will be too. Loggix helps businesses build the integrated, custom software foundation — from tailored FileMaker environments and ERP connections to API integrations and embedded AI workflows — that makes AI-augmented management actually work in practice, not just in theory. If you are mapping out what that looks like for your organisation, that is a conversation worth having.","\u003Cp>Your team is spending hours every week pulling data, formatting reports, and chasing down numbers that a machine could surface in seconds. Meanwhile, the decisions that actually require human judgment — how to respond to a struggling employee, whether to enter a new market, how to handle an operational crisis — keep getting pushed to "when there's time." There never is. This article maps exactly how AI is redistributing the work of management, what responsibilities fall away, what new ones emerge, and how you can navigate the transition without losing your team in the process.\u003C\u002Fp>\n\u003Ch2>What work does AI actually take off a manager's plate?\u003C\u002Fh2>\n\u003Cp>The honest answer: more than most managers expect, and less than AI vendors claim.\u003C\u002Fp>\n\u003Cp>AI tools are genuinely good at high-volume, pattern-based work that used to eat management time:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Automated reporting\u003C\u002Fstrong>: Instead of a logistics manager spending Friday afternoon compiling a weekly performance report from four spreadsheets, the system generates it automatically — on-time delivery rate, cost per shipment, driver utilisation — and flags anything outside tolerance. The manager opens a summary, not a raw dataset.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Anomaly detection\u003C\u002Fstrong>: A production manager no longer needs to scan hundreds of output rows to find the line running 12% below target. The AI flags it, often before end-of-shift.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Demand forecasting\u003C\u002Fstrong>: A retail operations manager who used to build next-quarter stock projections by hand — adjusting for seasonality, promotions, and supplier lead times — now reviews and challenges a model-generated forecast instead of constructing one from scratch.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Routine recommendations\u003C\u002Fstrong>: An AI system trained on historical order patterns can recommend reorder points, flag a supplier's declining reliability score, or suggest which customer accounts are at churn risk. The manager decides what to do about it.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>The shift is from \u003Cem>producing\u003C\u002Fem> information to \u003Cem>judging\u003C\u002Fem> information. That sounds like a small change. It is not.\u003C\u002Fp>\n\u003Ch2>What responsibilities grow — and get harder?\u003C\u002Fh2>\n\u003Cp>When AI handles the operational layer, management bandwidth does not simply free up and sit idle. It gets redirected to tasks that are structurally harder:\u003C\u002Fp>\n\u003Ch3>1. Validating AI-driven insights\u003C\u002Fh3>\n\u003Cp>An AI forecast is only as good as the data and assumptions behind it. Managers must now ask: \u003Cem>Is this model seeing what I know about this market?\u003C\u002Fem> A demand forecast built on last year's data will not reflect a competitor who just entered your territory. A churn-risk score built on purchase frequency misses the customer who paused buying because of a temporary budget freeze, not dissatisfaction.\u003C\u002Fp>\n\u003Cp>Managers who simply trust AI output without interrogating it will make worse decisions than managers who never had AI at all — because the false confidence is invisible.\u003C\u002Fp>\n\u003Ch3>2. Owning accountability when AI is wrong\u003C\u002Fh3>\n\u003Cp>A classic operational error has a clear owner. An AI-assisted decision is murkier. When the AI-recommended staffing plan leaves you understaffed during a peak weekend, who is accountable? The manager who approved it. AI does not absorb accountability — it transfers it, silently, to whoever signs off.\u003C\u002Fp>\n\u003Cp>This means managers need a new habit: \u003Cem>explicitly documenting why they agreed with or overrode an AI recommendation\u003C\u002Fem>, so that learning accumulates and accountability stays visible.\u003C\u002Fp>\n\u003Ch3>3. Coaching employees through uncertainty\u003C\u002Fh3>\n\u003Cp>Team members who used to own the weekly report now feel displaced. The analyst who spent Tuesday building the revenue model watches AI do it in four minutes. That is a real identity crisis, not just a workload adjustment. Managers are now expected to coach people through a redefinition of their value — which requires emotional intelligence, not just project management skill.\u003C\u002Fp>\n\u003Ch3>4. Continuous process improvement\u003C\u002Fh3>\n\u003Cp>AI surfaces patterns. It does not fix the processes behind them. If the anomaly detection keeps flagging the same production bottleneck, someone has to own redesigning the workflow. That someone is the manager, working with their team — not the algorithm.\u003C\u002Fp>\n\u003Ch3>5. Change management at every level\u003C\u002Fh3>\n\u003Cp>Implementing AI tools into an existing team is a change management exercise, not an IT project. Resistance, confusion, and quiet workarounds are the norm, not the exception. Managers who treat AI rollout as "the tech team's problem" watch adoption fail silently.\u003C\u002Fp>\n\u003Ch2>What does a "new" manager's week actually look like?\u003C\u002Fh2>\n\u003Cp>Here is a concrete before\u002Fafter for a mid-sized operations manager at a manufacturing company:\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Before AI integration:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Monday: compile weekend production report from three systems (3 hours)\u003C\u002Fli>\n\u003Cli>Tuesday: prepare supplier performance review manually (2 hours)\u003C\u002Fli>\n\u003Cli>Wednesday: attend KPI meeting where half the time is spent correcting data\u003C\u002Fli>\n\u003Cli>Thursday: field team questions about why targets were missed\u003C\u002Fli>\n\u003Cli>Friday: build next week's shift schedule based on gut feel and last week's attendance\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cstrong>After AI integration:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Monday: review auto-generated weekend report, investigate the two flagged anomalies (45 minutes)\u003C\u002Fli>\n\u003Cli>Tuesday: challenge the AI-generated supplier risk scores with market context the system doesn't have\u003C\u002Fli>\n\u003Cli>Wednesday: KPI meeting focused entirely on decisions, not data validation\u003C\u002Fli>\n\u003Cli>Thursday: one-on-one sessions with team members on skills development and process ideas\u003C\u002Fli>\n\u003Cli>Friday: review AI-proposed shift schedule, adjust for the two things the model doesn't know (an employee's upcoming training day, a known peak from a trade show)\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>The volume of \u003Cem>hours\u003C\u002Fem> may look similar on paper. The cognitive mode is completely different.\u003C\u002Fp>\n\u003Ch2>How do you validate AI-driven insights without second-guessing everything?\u003C\u002Fh2>\n\u003Cp>This is the practical skill gap most managers have right now, and almost nobody is training for it. Here is a working framework:\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>Ask where the data came from.\u003C\u002Fstrong> Is it real-time or a batch update from 48 hours ago? Does it include all relevant data sources, or just the ones that were easy to connect?\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Ask what the model was optimised for.\u003C\u002Fstrong> A forecasting model optimised for average accuracy will systematically underperform at the extremes — exactly where you need it most.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Ask what it cannot see.\u003C\u002Fstrong> Competitor moves, regulatory changes, one-off events, and relationship context are usually invisible to the model. Add that context before acting.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Test the recommendation against your own judgment.\u003C\u002Fstrong> If the AI says "reduce inventory by 15%" and your sales team is telling you a large order is coming, that tension is worth investigating — not suppressing.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Document your decision.\u003C\u002Fstrong> "I approved the AI recommendation because X" or "I overrode it because Y" — three sentences in your project log. This builds organisational learning over time.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Ch2>How should managers coach their team through this shift?\u003C\u002Fh2>\n\u003Cp>The employees most at risk are not the lowest performers — they are the analytical, detail-oriented people who built their identity around being the ones who knew the numbers. Here is what actually helps:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Name the shift explicitly.\u003C\u002Fstrong> Tell your team: "Your value is no longer in producing data. It is in knowing what to do with it." Most managers assume their team understands this. Most teams do not.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Reassign, don't just remove.\u003C\u002Fstrong> When AI takes over report generation, give the analyst a new owner role — reviewing model assumptions, catching edge cases, communicating findings to stakeholders. The role evolves; it doesn't disappear.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Celebrate overrides, not just approvals.\u003C\u002Fstrong> When a team member spots that the AI forecast is wrong because of a market factor the model missed, that is a win. Recognise it visibly, or people will stop questioning the machine.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Invest in prompt literacy.\u003C\u002Fstrong> The ability to ask an AI system the right question — with the right context, constraints, and format — is now a core operational skill. Train for it deliberately.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>What new skills do managers need to develop?\u003C\u002Fh2>\n\u003Cp>The capability gap is real. Managers who thrive with AI share a recognisable profile:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Critical data literacy\u003C\u002Fstrong>: not knowing how to build a model, but knowing how to challenge one\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Judgment under ambiguity\u003C\u002Fstrong>: acting decisively when the AI gives a probability, not a certainty\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Contextual intelligence\u003C\u002Fstrong>: holding the organisational and market context that no dataset can fully encode\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Psychological safety as a leadership practice\u003C\u002Fstrong>: creating an environment where people say "I think the AI is wrong here" without fear\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Accountability discipline\u003C\u002Fstrong>: owning outcomes from AI-assisted decisions as fully as from manual ones\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>None of these are new concepts. AI makes them load-bearing in a way they were not before.\u003C\u002Fp>\n\u003Ch2>Checklist: Is your management team ready for AI-augmented operations?\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> Managers can articulate \u003Cem>why\u003C\u002Fem> they agreed with or overrode an AI recommendation — not just what they decided\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> AI outputs are regularly challenged in team meetings, not just presented\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> Team members whose roles changed due to AI have been given new, meaningful responsibilities\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> There is a documented process for escalating suspected AI errors\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> Managers are spending measurably less time on data collection and more on exception handling and people development\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> AI recommendations are treated as inputs to decisions, not as decisions themselves\u003C\u002Fli>\n\u003Cli>\u003Cinput disabled=\"\" type=\"checkbox\"> Change management support was part of your AI rollout, not an afterthought\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>FAQ\u003C\u002Fh2>\n\u003Cp>\u003Cstrong>Will AI make managers redundant?\u003C\u002Fstrong>\nNot in the foreseeable future — but it will make \u003Cem>some managers redundant\u003C\u002Fem>. Specifically, managers whose primary value was coordinating information flow and producing reports. Managers whose value lies in judgment, accountability, and people development become more important, not less.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>How do I get my team to trust AI without over-trusting it?\u003C\u002Fstrong>\nModel the behaviour yourself. Ask questions about AI output in public. Override a recommendation in front of the team when you have good reason to, and explain why. Trust calibration is a cultural practice, not a policy.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>What if our AI tools are giving us bad recommendations?\u003C\u002Fstrong>\nTreat it as a data quality and model design problem, not a technology problem. Bad recommendations usually trace back to incomplete data, misaligned optimisation targets, or missing business context. Fix those, rather than either abandoning the tool or ignoring the bad outputs.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>How does this connect to broader AI strategy in our organisation?\u003C\u002Fstrong>\nManager-level AI adoption rarely happens in isolation. It works best when it is part of a deliberate design of \u003Ca href=\"https:\u002F\u002Floggix.com\u002Fen\u002Fblog\u002Fhow-to-design-effective-collaboration-between-people-and-ai\">how people and AI collaborate across the organisation\u003C\u002Fa> — with clear boundaries, escalation paths, and feedback loops built in from the start.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Is this different for small vs. large companies?\u003C\u002Fstrong>\nThe principles are the same; the stakes are different. In a 12-person company, the manager \u003Cem>is\u003C\u002Fem> the team, so the shift to judgment-focused work is even more immediate. In a large organisation, the challenge is more about scaling the cultural change consistently across management layers.\u003C\u002Fp>\n\u003Chr>\n\u003Cp>For organisations where these decisions are tied to real operational systems — ERP data, production logs, customer records, supplier feeds — the quality of AI-generated insights depends entirely on the quality of the underlying data infrastructure. If your systems are fragmented, your AI recommendations will be too. Loggix helps businesses build the integrated, custom software foundation — from tailored FileMaker environments and ERP connections to API integrations and embedded AI workflows — that makes AI-augmented management actually work in practice, not just in theory. If you are mapping out what that looks like for your organisation, that is a conversation worth having.\u003C\u002Fp>\n","Jeroen","2026-07-24",1784901664000,[19,20,21,22,23,24,25,26,27,28],"AI","management","business strategy","organizational intelligence","AI adoption","decision-making","change management","leadership","process improvement","human-AI collaboration",null,false,{"title":32,"slug":33},"AI and Organizational Intelligence","ai-and-organizational-intelligence",{"title":35,"slug":36},"How to design effective collaboration between people and AI","how-to-design-effective-collaboration-between-people-and-ai"]