Artificial intelligence is often discussed through two competing narratives.
One says AI will transform productivity, innovation, medicine, education and business, creating opportunities we are only beginning to understand. The other warns that AI will eliminate jobs, concentrate power, undermine human judgement and potentially create risks that society cannot adequately control.
In Superagency: What Could Possibly Go Right with Our AI Future, Reid Hoffman and Greg Beato deliberately challenge the pessimistic side of that debate.
Their central question is refreshingly provocative: instead of focusing almost exclusively on what could go wrong with artificial intelligence, what could go right?
Published in January 2025, Superagency argues that widespread access to AI could increase rather than diminish human agency. Hoffman and Beato define “superagency” as the widespread empowerment that emerges when millions of people simultaneously gain access to a transformative technology. (Reid Hoffman)
For small-to-medium business owners and leaders, that idea deserves serious attention.
The strategic question is not simply:
“How can we use AI?”
It is:
“How can we use AI to make our people, decisions, customers and business more capable?”
That distinction sits at the heart of Superagency, and it may prove to be one of the most important leadership questions of the next decade.
Table of Contents
- What Is Superagency Really About?
- Human Agency Before Artificial Intelligence
- AI as an Intelligence Amplifier
- From AI Automation to AI Augmentation
- Why Experimentation Matters
- The Democratisation of Expertise
- AI and the Future of Work
- What Superagency Gets Right
- Where SME Owners Should Exercise Caution
- Why AI Strategy Is Really Business Strategy
- The AGENCY Framework for SME Owners & Leaders
- Practical Recommendations
- Key Takeaways
- FAQs
- Conclusion
What Is Superagency Really About? (Business Advisor Perth)
Superagency is not fundamentally a book about algorithms, large language models or computer science.
It is a book about human capability.
Hoffman and Beato argue that technological progress has repeatedly expanded what individuals can accomplish. Writing extended memory. Printing expanded access to knowledge. Cars increased mobility. Computers amplified calculation. The internet dramatically reduced the cost of accessing and distributing information.
AI, they argue, represents another step in this progression.
But it may be qualitatively different.
A calculator helps us calculate.
A search engine helps us find information.
Generative AI can increasingly help us think, analyse, create, communicate, learn, decide and act.
That potentially makes AI less like another software application and more like a broadly accessible cognitive amplifier.
The official description of Superagency captures this ambition clearly: AI tutors could personalise education, researchers could accelerate medical discovery, and AI advisers could help individuals navigate increasingly complex systems. (Superagency)
The implications for SMEs are substantial.
An SME traditionally operates with significant resource constraints. It cannot employ a specialist for every problem. It may not have dedicated strategy, research, data analytics, marketing, HR, technology or commercial-analysis teams.
AI potentially changes the economics of accessing some of those capabilities.
That does not make expertise obsolete. It makes basic access to capability dramatically cheaper.
Human Agency Before Artificial Intelligence (Leadership Development Perth)
Agency is essentially our ability to make choices, pursue goals and influence outcomes.
Hoffman and Beato’s argument is that technology can expand that ability.
This matters because much of the AI debate assumes an implicit contest:
human versus machine.
But that may be the wrong framing.
The more commercially useful question is:
What can a capable person achieve when augmented by increasingly capable technology?
This is particularly relevant to SME leadership.
Consider an owner preparing a strategic plan.
Without AI, the process might require extensive research, external consultants, management workshops and considerable analysis.
With AI, that same owner can potentially explore industry trends, challenge assumptions, model scenarios, examine competitor positioning, structure strategic alternatives and prepare questions for advisers far faster.
That does not mean AI should determine the strategy.
It means the owner can arrive at the strategic planning process better informed and better prepared.
The human remains responsible for judgement.
The machine expands the information and analytical capability available to that human.
That is agency amplification in practice.
AI as an Intelligence Amplifier (Business Improvement Perth)
One of the most valuable ways SME owners can interpret Superagency is to stop thinking about AI simply as a productivity tool.
Productivity matters, but it is only one dimension.
AI can potentially amplify:
Thinking. Exploring alternatives, challenging assumptions and structuring complex problems.
Research. Synthesising large amounts of information rapidly.
Communication. Improving proposals, reports, presentations and customer communication.
Analysis. Interrogating financial, operational and commercial information.
Creativity. Generating alternatives for products, services, marketing and problem-solving.
Learning. Accelerating the acquisition of knowledge.
Decision preparation. Helping leaders identify the questions they should be asking before making important decisions.
This complements an important principle from *Atomic Habits*: better systems can produce better outcomes repeatedly.
AI becomes strategically interesting when it moves beyond occasional prompting and becomes embedded into how work is performed.
From AI Automation to AI Augmentation (Business Improvement Perth)
Many businesses begin their AI journey by asking:
“What can we automate?”
That is understandable.
Automation can reduce administration, eliminate repetitive tasks and lower costs.
But Superagency points towards a more ambitious question:
“What can our people become capable of doing that they could not do before?”
This is augmentation rather than simple automation.
Imagine a salesperson who can analyse a customer’s history before every meeting.
A financial controller who can interrogate management accounts conversationally.
A manager who can rapidly prepare alternative rostering scenarios.
A marketing employee who can test ten campaign concepts rather than two.
A CEO who can conduct an initial strategic analysis before engaging the Board.
A customer-service employee who can instantly access relevant product and customer information.
These applications do not necessarily remove humans.
They can make humans substantially more effective.
That distinction matters strategically because businesses focusing only on labour replacement may miss AI’s larger opportunity: capability multiplication.
Why Experimentation Matters (Strategic Planning Perth)
Hoffman’s philosophy is strongly oriented towards experimentation and iterative deployment.
In discussing Superagency, he has described his approach as “smart risk taking” rather than blind optimism, arguing that getting AI into people’s hands, learning from actual use and improving safeguards iteratively can help society discover both opportunities and problems. (TechCrunch)
There is a useful SME lesson here.
Businesses do not need to predict precisely what AI will look like in five years before beginning.
In fact, waiting for certainty may itself become a strategic risk.
A better approach is controlled experimentation:
identify a real business problem, select an appropriate AI application, establish safeguards, test it, measure the result, learn from it, then scale what works.
That philosophy closely resembles the disciplined experimentation behind effective business model innovation.
AI capability will probably be developed through use, not observation.
The Democratisation of Expertise (Business Growth Perth)
One of Superagency’s most powerful implications for SMEs is the democratisation of capability.
Historically, sophisticated analytical resources were concentrated inside large organisations.
Large companies could employ economists, analysts, strategists, lawyers, researchers, programmers, designers and specialist consultants.
Small businesses could not.
AI does not eliminate the need for specialists, particularly where professional judgement, accountability, regulation or deep expertise matter.
But it can reduce the cost of accessing a useful first layer of capability.
A small manufacturer can analyse operational data.
A family business can explore succession scenarios.
A founder can pressure-test a business model.
A sales manager can analyse customer segmentation.
A small professional-services firm can research markets that previously required substantial manual effort.
This could partially narrow the capability gap between SMEs and larger corporations.
But there is an important qualification.
Access to intelligence is not the same as possessing judgement.
Knowing what question to ask, whether an answer is credible and what action to take remains enormously important.
Indeed, as information becomes cheaper, judgement may become more valuable.
AI and the Future of Work (Leadership Development Perth)
Superagency rejects the simplistic assumption that AI necessarily leads to mass human redundancy.
The authors acknowledge job disruption but emphasise the possibility that AI will complement people, create new capabilities and reshape jobs rather than simply eliminate them. (Superagency)
For business leaders, however, optimism should not become complacency.
Some tasks will disappear.
Some roles will shrink.
Others will become more productive.
New roles will emerge.
And the value attached to particular capabilities will change.
The practical SME question therefore becomes:
Which tasks should AI perform, which should people perform, and which are most powerful when performed together?
This requires thoughtful job redesign.
An employee spending ten hours each week assembling information manually may instead spend two hours validating AI-assisted analysis and eight hours acting on the findings.
That is not simply productivity improvement.
It changes the nature of the job.
Leaders therefore need to think about AI alongside roles, responsibilities and accountability.
What Superagency Gets Right (Business Advisor Perth)
The book’s greatest strength is that it provides a necessary counterweight to AI fatalism.
Hoffman and Beato remind us that societies frequently underestimate the adaptability of human beings.
New technologies create disruption, but they also create opportunities that are difficult to imagine before widespread adoption.
Their second major strength is placing agency rather than technology at the centre of the discussion.
That is particularly useful for business owners.
The objective is not AI adoption for its own sake.
The objective is improving what customers, employees, managers and businesses can accomplish.
Third, Superagency encourages participation rather than passive observation.
That is strategically important.
Businesses that experiment with AI develop organisational knowledge. Employees discover what works. Managers understand limitations. Processes evolve.
A company waiting indefinitely for the technology to mature may eventually discover that competitors have accumulated years of practical learning.
This reinforces the argument I have previously made about artificial intelligence for SME owners: AI should increasingly be treated as a strategic management issue rather than simply an IT initiative.
Where SME Owners Should Exercise Caution (Governance & Boards)
The book’s optimism is also where SME readers need to apply judgement.
Hoffman’s proximity to the technology industry gives him considerable insight, but it also means readers should recognise the perspective from which the argument is being made.
AI adoption does create genuine risks.
These include:
- inaccurate or fabricated outputs,
- confidential information leakage,
- cybersecurity exposure,
- intellectual-property uncertainty,
- privacy concerns,
- poor-quality automated decisions,
- employee overdependence,
- bias,
- weak accountability,
- erosion of critical thinking,
- regulatory risk.
And the wider AI-risk debate has become more, rather than less, contested since Superagency was published. Some AI researchers and executives now argue for substantially stronger safeguards and slower frontier development, while Hoffman continues to favour responsible advancement rather than stopping development. (Reuters)
That does not invalidate Superagency.
It makes governance more important.
The correct SME response is neither:
“AI is dangerous, so we will avoid it.”
nor:
“AI is extraordinary, so everyone should use it for everything.”
A better principle is:
Adopt aggressively where the opportunity is meaningful, govern carefully where the consequences of error are significant.
AI Must Strengthen Human Judgement, Not Replace It (Leadership Development Perth)
There is another risk that deserves more emphasis than Superagency gives it.
People may become cognitively lazy.
If AI writes every report, develops every argument, summarises every document and proposes every decision, leaders risk gradually outsourcing the thinking that develops judgement.
That would represent the opposite of superagency.
The strongest AI users will therefore not necessarily be those who delegate the most thinking to machines.
They may be those who use AI to challenge and strengthen their own thinking.
Ask AI:
What assumption am I making?
What evidence contradicts my position?
What alternative strategy have I overlooked?
What would a competitor do?
What questions should the Board ask?
Where might this analysis be wrong?
Used this way, AI becomes an intellectual sparring partner rather than an intellectual substitute.
That distinction is crucial.
Why AI Strategy Is Really Business Strategy (Strategic Planning Perth)
A common mistake is delegating AI entirely to IT.
AI is too important for that.
Its implications touch customers, pricing, products, operations, employees, decision-making, competitive advantage, intellectual property and potentially the entire business model.
The Board and leadership team therefore need to ask:
Where could AI materially improve customer value?
Where could it increase productivity?
Which decisions could improve?
Which processes could be redesigned?
Which capabilities could become cheaper or faster?
Could competitors use AI to undermine our existing advantage?
Could AI create entirely new products or services?
These are strategy questions.
AI is simply the catalyst.
The AGENCY Framework, A Practical AI Framework for SME Owners & Leaders
The central idea of Superagency can be translated into a practical six-part framework SMEs can use to move from random experimentation to disciplined AI adoption.
A — Assess the Opportunity
Start with the business, not the technology.
Identify processes that are repetitive, slow, expensive, information-heavy or dependent on scarce expertise.
Ask:
Where could better intelligence materially improve performance?
Prioritise commercial problems rather than fashionable applications.
G — Govern the Risk
Define boundaries before scaling.
Determine what information employees may provide to AI systems, what requires human review, which decisions cannot be automated and who remains accountable.
Governance should enable intelligent experimentation rather than suffocate it.
E — Experiment Small and Fast
Run controlled pilots.
Choose one workflow.
Establish the current baseline.
Introduce AI.
Measure time, cost, quality and outcome.
Keep what works. Modify what partly works. Stop what does not.
This is substantially better than launching an organisation-wide “AI transformation” without understanding the economics.
N — Nurture Human Capability
Use AI to make people better.
Train employees to question outputs, verify important facts, improve prompts, recognise limitations and combine machine intelligence with human context.
The objective is not merely AI literacy.
It is AI-enhanced professional capability.
C — Connect AI to Customer and Commercial Value
Every significant AI initiative should ultimately connect to something economically meaningful.
Can we serve customers better?
Respond faster?
Reduce errors?
Improve margins?
Develop better products?
Make better decisions?
Increase capacity?
Create a stronger competitive advantage?
If not, AI risks becoming technological theatre.
Y — Yield, Measure and Scale
Measure outcomes.
Useful metrics may include:
- hours saved,
- turnaround time,
- error reduction,
- revenue improvement,
- gross-margin improvement,
- customer satisfaction,
- employee capacity,
- conversion rates,
- decision quality.
Then scale successful applications systematically.
ASSESS → GOVERN → EXPERIMENT → NURTURE → CONNECT → YIELD
That is how SMEs can turn AI experimentation into genuine AGENCY.
A Simple AGENCY Self-Assessment
Score your business from 1 to 5 on each dimension:
Assess: Have we identified our highest-value AI opportunities?
Govern: Do we have clear AI policies, controls and accountability?
Experiment: Are we actively testing AI against real business problems?
Nurture: Are we developing employees’ ability to work effectively with AI?
Connect: Are AI initiatives connected to measurable customer or commercial value?
Yield: Are we measuring results and scaling successful applications?
Maximum score: 30
6–12: AI Observer
Your organisation is largely watching rather than learning.
13–18: AI Experimenter
Activity exists, but it is fragmented or insufficiently commercial.
19–24: AI Integrator
AI is becoming embedded into meaningful workflows.
25–30: AI Amplifier
AI is systematically expanding organisational capability.
The purpose is not the score itself. The value lies in identifying which dimension is preventing AI from producing meaningful business results.
10 Practical Recommendations for SME Owners & Leaders
- Start using AI personally. Leaders cannot govern intelligently what they do not understand.
- Identify your ten most time-consuming knowledge tasks. They are obvious places to investigate.
- Select three high-value AI experiments. Avoid organisation-wide transformation initially.
- Establish an AI-use policy. Address confidential information, customer data, intellectual property, verification and accountability.
- Require human review for consequential decisions.
- Train employees to challenge AI, not simply accept its answers.
- Measure outcomes rather than usage. Ten thousand prompts mean nothing if performance does not improve.
- Redesign workflows rather than simply inserting AI into inefficient existing processes.
- Review your strategy and competitive position. Ask how AI could change your industry, customer expectations and value proposition.
- Make AI a recurring leadership and Board discussion. The technology is evolving too quickly for an annual review.
Key Takeaways
- Superagency provides a deliberately optimistic counterweight to AI pessimism.
- Hoffman and Beato’s most important idea is that AI can amplify human agency, not merely automate work.
- For SMEs, AI could partially democratise access to capabilities historically available mainly to larger organisations.
- The biggest opportunity may be augmentation rather than employee replacement.
- AI experimentation creates organisational learning and should begin before perfect certainty exists.
- Access to more intelligence makes human judgement more important, not less.
- AI must be connected to strategy, customers, productivity and commercial outcomes.
- Responsible adoption requires governance, privacy, cybersecurity and human accountability.
- SMEs should measure AI by business results rather than the number of tools they adopt.
- The ultimate question is not whether your business “uses AI”, but whether AI makes your people and business demonstrably more capable.
Frequently Asked Questions
What is Superagency about?
Superagency argues that widespread access to AI can expand individual and collective human agency, giving people greater ability to learn, create, solve problems and achieve goals. (Superagency)
Who wrote Superagency?
The book was written by Reid Hoffman and Greg Beato and published in January 2025. (Simon & Schuster)
What does “superagency” mean?
Hoffman describes it as widespread empowerment arising when millions of people gain simultaneous access to a breakthrough technology. (Reid Hoffman)
Is Superagency optimistic about artificial intelligence?
Yes. It deliberately concentrates on AI’s potential benefits while acknowledging issues including disinformation, employment disruption and other risks. (Superagency)
Is the book relevant to SME owners?
Very much so. Its central concept can be applied to productivity, decision-making, customer service, strategy, innovation, employee capability and competitive advantage.
Should SMEs use AI primarily to reduce staff?
Not necessarily. A more strategic starting point is identifying how AI can augment employee capability, improve workflows and create customer or commercial value.
What are the major risks for SMEs using AI?
Confidentiality, inaccurate outputs, cybersecurity, privacy, intellectual property, bias, overreliance and unclear accountability are among the most important.
Should an SME have an AI policy?
Yes. Even a relatively small organisation should establish practical rules governing approved tools, confidential information, customer data, verification and accountability.
Should AI make important business decisions?
AI can contribute analysis and alternatives, but consequential decisions should retain appropriate human oversight and accountability.
How should an SME start implementing AI?
Start with a genuine business problem, run a controlled experiment, measure the outcome and scale successful applications. The AGENCY Framework provides a practical structure for doing this.
Conclusion, AI Should Make Us More Human, Not Less (Business Advisor Perth)
Superagency is valuable precisely because it challenges the dominant question surrounding artificial intelligence.
Instead of asking only:
What could AI destroy?
Hoffman and Beato ask:
What could AI enable?
That is an important question.
Their optimism should not be accepted uncritically. AI creates genuine commercial, societal and governance risks, and developments since the book’s publication have intensified rather than settled the debate about how rapidly increasingly powerful AI should be developed. (Reuters)
But SMEs cannot build strategy around fear alone.
Nor should they build it around technological enthusiasm alone.
The better path is informed experimentation, disciplined governance and relentless focus on human and commercial value.
The companies that benefit most from AI may not be those with the most AI tools.
They will be businesses that learn how to combine machine capability with human judgement, creativity, relationships, experience and accountability.
That is where the concept of superagency becomes commercially powerful.
The future competitive advantage may not belong to AI or to humans. It may belong to organisations that learn how to combine the best of both.
For SME owners and leaders, the time to begin developing that capability is not when AI has stopped changing.
It is now.
For further practical guidance on integrating AI into strategy and business performance, see my article on Artificial Intelligence for SME Owners, or contact me if you would like to discuss how AI, strategy, leadership and business improvement can be applied practically within your organisation.
