Peter Drucker

AI agents are being deployed into organizations without any framework for accountability. Enterprises are experimenting wildly—60-89% have tried agentic AI—but no one has solved the fundamental question: who is responsible when an AI agent makes a decision? Organizations treat AI as either a tool (no accountability) or a worker (inappropriate accountability).

A simple diagnostic framework—not software, but a structured inquiry—that forces organizations to answer three questions before deploying any AI agent: What specific contribution is this agent supposed to make? How will we know if it is making that contribution? Who is accountable when it fails? Publish this as a short, plain-language pamphlet and test it with three or four organizations willing to be guinea pigs, then revise based on what actually happens.

Pension funds now own American business, but they are managed by financial intermediaries with no connection to the workers whose retirement depends on them. The workers own but do not control; the managers control but do not own. This is the worst possible outcome—responsibility without authority for the workers, authority without responsibility for the fund managers. The pension crisis is not primarily financial; it is a crisis of governance.

A prototype for direct worker representation on pension fund investment committees—not as union representatives bargaining for benefits, but as owners with fiduciary responsibility. Find one mid-sized pension fund willing to experiment, help them design the selection process and the decision rights, document what happens, and publish the results.

Remote work has made the fundamental problem of knowledge work visible: no one knows how to measure contribution. The tools that exist—time tracking, activity monitoring, productivity software—measure effort and busyness, not results. They are the knowledge-work equivalent of counting how many times a factory worker touches a machine. Meanwhile, managers either micromanage through surveillance or abdicate entirely.

A method, not a tool, for knowledge workers to define their own contribution objectives quarterly, track their own results weekly, and report upward monthly in a standard one-page format. The format would require answering: What did I commit to contribute? What did I actually contribute? What prevented greater contribution? What will I contribute next? Pilot this with remote teams at three different organizations and publish what works.

Aging populations are treated as a burden—a demographic crisis requiring solutions. But the real problem is that we have made people over sixty-five useless. We force them out of organizations, deny them meaningful work, and then wonder why healthcare costs explode and loneliness kills. The labor force participation rate for people 60-64 varies from 21% to over 70% across countries—proof that this is a policy choice, not a biological necessity.

A matching service connecting experienced retired professionals with small organizations—nonprofits, small businesses, municipal agencies—that cannot afford full-time senior talent but desperately need it. Not volunteering (which strips dignity) and not consulting (which is too expensive and too disconnected). Part-time, paid, ongoing relationships with real responsibility. Start in one metropolitan area with fifty retired professionals and fifty small organizations.

Middle managers are in crisis—40% seeking purpose, 22% wanting leadership skills, caught between executives who demand results and workers who demand meaning. Organizations have spent decades eliminating middle management as 'overhead' while loading those who remain with administrative burden and emotional labor. But middle managers are where strategy becomes execution; without effective middle management, organizations cannot function.

A six-month peer learning cohort for middle managers from different organizations, meeting monthly, with a simple curriculum: each month, one participant presents a real decision they face, the group analyzes it together, and everyone commits to one change in their own practice. No lectures, no content delivery, no certificates. Just structured peer accountability for actually improving. Run three cohorts of fifteen people each in the first year and document the method.

Hospitals are adopting management software, AI analytics, and efficiency tools—but they are not becoming more effective. The tools optimize pieces while the whole deteriorates. Doctors burn out, nurses quit, administrators multiply, and patients wait. The hospital is the most complex organization humans have created, and we are trying to manage it with tools designed for factories and offices.

A diagnostic method for hospital leaders to identify their three highest-leverage decisions—not their biggest problems, but the decisions that, if made well, would most improve patient outcomes and staff effectiveness. Then help them redesign the information flow around those three decisions only. Start with contribution, not with data. Work with two or three hospitals willing to spend six months on this and publish what we learn.

Small businesses are the backbone of employment, but their founders are aging and most have no succession plan. The existing tools—software for talent assessment, leadership pipelines, competency tracking—are designed for large organizations. A fifty-person manufacturing company or a family restaurant cannot use them. When the founder dies or retires without a successor, the business dies and the jobs disappear.

A simple, paper-based succession planning workbook for small business owners, designed to be completed in four sessions over four months. It would force the owner to answer: Who could run this business if I were hit by a bus tomorrow? What do they need to learn? How will I teach them? What will happen to my family? Test this with fifty small business owners in one region, refine it based on what actually gets completed, and publish it freely.

Nonprofits are being strangled by accountability theater—impact measurement frameworks, logic models, outcome metrics—that measure what funders want to see rather than what the organization actually accomplishes. The 2025 nonprofit sector crisis shows organizations spending more time proving their value than creating it. Real accountability requires knowing what contribution you intend to make and honestly assessing whether you made it.

A counter-framework for nonprofit boards: instead of requiring detailed impact metrics, require the executive director to answer three questions annually in plain language: What did we set out to accomplish this year? Did we accomplish it? What did we learn that changes what we should do next year? Recruit fifteen nonprofit boards to adopt this for two years and compare their organizational health to matched organizations using traditional accountability frameworks.

The creator economy is breaking its people—burnout, financial instability, algorithmic dependence, no separation between person and product. Creators have exchanged the boss for the algorithm, which is worse. They have no colleagues, no institution, no accumulation of capability. Each day starts from zero. This is precisely what happened to knowledge workers before management existed: each person alone against the chaos.

A cooperative structure for mid-tier creators—those earning $50,000-$200,000 annually—that provides three things: shared back-office services (accounting, legal, insurance), peer accountability groups meeting monthly, and a pooled emergency fund for members who need to take breaks. Not an agency (which extracts value) and not a platform (which creates dependence). A mutual aid society for digital craftspeople. Start with thirty creators in one vertical and document what actually helps.

Trust in institutions has declined twenty-two percentage points since 1979 and continues falling. The standard responses—transparency initiatives, accountability frameworks, stakeholder engagement—do not work because they assume the problem is information. But the problem is relationship. People do not trust institutions they have no relationship with; and modern institutions have systematically eliminated relationships in favor of efficiency.

A research project documenting specific cases where institutional trust was rebuilt, analyzing exactly what was done and by whom. Not survey data about trust levels, but narrative case studies of trust repair: the hospital that regained its community's confidence, the company that recovered from scandal, the government agency that earned back credibility. Publish these as short, readable cases and identify the common elements.