There are moments when technology makes us pause and feel genuine wonder.

Many of us felt that watching recent Artemis coverage, or reading about the rovers already working on Mars. These missions do more than showcase technical brilliance. They remind us that human beings have always pushed forward by building tools that extend what we can do.

And now, increasingly, artificial intelligence is one of those tools.

I recently watched a documentary explaining that once a rover is operating in the Martian environment, it cannot rely on humans on Earth to direct every move in real time. The communication delay is too great. That means some decisions must be made autonomously.

NASA’s recent updates show just how far that has gone. In January 2026, NASA announced that the Perseverance rover completed the first drives on another world that were planned by artificial intelligence. The mission team used generative AI to create waypoints for the rover; work normally carried out by human rover planners. NASA also says that 88% of Perseverance’s driving has been autonomous. 

This is extraordinary.

But the most important lesson is not that AI is replacing the humans. It is that AI is helping humans go further than they otherwise could.

This distinction matters.

AI can take humankind further but only with humans in the loop

When people speak about AI in organisations, the conversation often narrows too quickly into cost, headcount and automation.

But space exploration offers a better lens.

The question is not simply, “What can the machine do?”
The better question is, “What becomes possible when humans and intelligent systems work together well?”

That is what we are seeing in exploration. NASA’s own ethical framework for AI makes clear that its approach is meant to be human-centric, accountable, explainable, secure and robust. In other words, even in some of the most technically advanced environments imaginable, the goal is not unchecked machine action. It is responsible, guided use. 

That should make organisational leaders stop and think.

Because if NASA is not framing AI as a simple “human replacement” story, businesses should be very cautious about doing so.

What about Artemis? How much of the work used AI?

Artemis II is a useful example because it shows the balance clearly.

NASA’s 2026 mission materials emphasise astronauts, mission control, communications networks and manual capability. During the mission, the crew carried out a manual piloting demonstration, while mission controllers monitored the spacecraft and automated systems supported operations. NASA’s public descriptions point to a highly advanced human-machine environment — but not one where human judgment disappears. 

So in 2026, the public evidence suggests this:

AI and automation are clearly part of the wider mission ecosystem at NASA, including autonomy, planning support, data handling and exploration systems. But the astronauts and mission control remain central to operational oversight, safety and decision-making. 

That is exactly the kind of model many organisations need to think about more carefully.

Not: “How do we remove the people?”
But: “How do we design the best possible human–AI partnership?”

The workplace lesson leaders should not miss

Used well, AI can do what good tools have always done:

It can reduce delay.
It can support analysis.
It can extend reach.
It can process information at speed.
It can help people work on things that truly require human judgment.

But used badly, it can create confusion, dependency, poor-quality outputs, ethical blind spots, and serious distrust.

That is why governance matters.

NIST says its AI Risk Management Framework is designed to help organisations manage risks to individuals, organisations and society, and specifically notes the importance of defining human roles and responsibilities in operational AI use. 

This is not an abstract issue.

When leaders expand AI use without thinking through training, accountability and decision rights, they do not create transformation. They create fragility.

What a thoughtful AI–human partnership looks like

When leaders within organisations decide to introduce AI, or expand the processes AI will cover, thought should be given to what could be achieved if there is a fully considered partnership between people and technology.

Practical steps should include:

1. Start with purpose

Be clear about why AI is being introduced.
Is it improving accuracy? Reducing repetitive work? Supporting faster decisions? Freeing up capacity for better human work?

If the purpose is fuzzy, adoption will become fuzzy too.

2. Identify what should be automated, augmented and retained by humans

Not every task should be handed over.

Some tasks can be automated.
Some should be AI-supported but human-reviewed.
Some should remain firmly human because they involve ethics, nuance, trust, judgment, relationships or accountability.

3. Train people properly

This is non-negotiable.

NASA’s science and AI materials stress training, education, skill development and adaptation as part of responsible AI implementation. People cannot be expected to use complex systems well if they are given access without preparation. 

Training should include:

• how the tool works 
• what it is good at 
• what it is poor at 
• how to check outputs 
• when to override it 
• when  to escalate concerns 

4. Give people room to experiment

Real confidence does not come from a slide deck announcing a rollout.

It comes from practice.

People need the opportunity to test the tool, make mistakes safely, compare outputs, challenge assumptions and learn what good use looks like in their actual role.

5. Keep meaningful human oversight

If AI influences decisions that affect quality, safety, fairness, customers, staff or careers, someone should remain clearly responsible for that judgement.

That is not old-fashioned. That is good leadership.

6. Be transparent

People need to know:

• where AI is being used 
• what it is doing 
• what data it draws on 
• what the limits are 
• how final decisions are made 

Trust grows when people understand the system they are being asked to work with.

7. Build governance early

Do not wait until something goes wrong.

Set expectations around:

• review processes 
• quality standards 
• documentation      
• privacy      
• accountability      
• ethical risk 

NASA’s and NIST’s frameworks both point in this direction: trustworthy AI does not happen by accident. It has to be designed and governed. 

8. Strengthen the human capabilities that matter even more in an AI age

As AI grows more capable, human value does not disappear.

In many cases, it becomes more important.

Judgement.
Communication.
Ethical reasoning.
Context.
Discernment.
Leadership.
Relationship-building.
Courage.

These are not “soft extras.” They are part of what makes technology usable, safe and effective.

This is why I remain hopeful

I do not think the most interesting AI story is replacement.

I think it is amplification.

Space exploration gives us a glimpse of what is possible when intelligent systems are used to extend human capability, not erase it. Mars rovers can traverse terrain in ways that would be impossible if every move had to wait for Earth. Artemis shows us that even in the most advanced missions, human expertise, communication and judgment still matter profoundly. 

This is the deeper lesson for organisations.

AI may help us go faster.
It may help us go further.
It may help us solve harder problems.

Only if people are trained to use it well, trusted to question it, and empowered to make operational and ethical decisions about its use.

This is the partnership worth building.

And perhaps that is the real wonder of this moment:
not that machines are becoming more capable,
but that human beings now have the chance to become more capable with them.

AI adoption at work should start with one question: what problem are we solving?

There is a great deal of noise around AI adoption.

Some organisations are rushing to implement tools because competitors are doing it. Others are buying systems before they have defined the problem. Some are hoping AI will solve productivity issues that are actually caused by poor processes, duplicated work, unclear decision-making, overloaded teams, or weak communication. That is where many adoption efforts begin to go wrong.

AI adoption should not start with, “What tool can we buy?” It should start with, “What problem are we solving, for whom, and what would better look like?” This approach is not only more strategic; it is also more humane, more effective, and more likely to produce value for both organisations and employees. Guidance from NIST, OECD, ILO and EU-OSHA all points in the same broad direction: successful AI use depends on governance, clarity of purpose, worker consultation, human oversight, and ongoing measurement of both benefits and risks. (NIST)

The organisations that benefit most from AI are unlikely to be the ones that deploy it fastest. They are more likely to be the ones that deploy it thoughtfully.

Before adoption: define the real problem first

Before any system is introduced, leaders need to pause and ask a few difficult questions.

What is the actual issue? Is it excessive admin? Slow turnaround times? Poor access to knowledge? Inconsistent customer responses? Reporting burdens? Low-quality handovers? Decision bottlenecks? Staff fatigue caused by repetitive tasks?

If the problem has not been defined properly, AI may simply accelerate the wrong thing.

This matters because evidence suggests the benefits of AI are real, but not automatic. OECD findings show that many workers report better performance with AI, and many also say it can improve enjoyment of work. At the same time, those benefits depend on how the technology is introduced and governed. (OECD)

A good starting point is to identify:

• the problem to be solved
• the tasks involved in that problem
• which of those tasks are repetitive, time-heavy, or information-heavy
• which parts require judgement, empathy, context, ethics, or accountability
• what risks would arise if the system got it wrong

That last point matters more than many organisations admit.

Not every process is suitable for automation. In some areas, speed is helpful. In others, speed without judgement creates risk. ILO guidance has repeatedly stressed that the impact of AI depends on whether it is used to replace human contribution or complement it. Whether adoption leads to harm or benefit depends heavily on task design, management choices, and whether humans are retained to perform, review, or oversee critical parts of the work. (International Labour Organization)

So one of the most important questions before adoption is this:

Where will human input remain?

This should be discussed explicitly, not assumed.

Human input should remain wherever work requires:

• ethical judgement
• contextual decision-making
• empathy and relationship management
• interpretation of nuance
• exception handling
• accountability for final decisions
• safeguarding, risk, or legal responsibility

NIST’s AI Risk Management Framework places strong emphasis on trustworthiness, governance, and human-AI configurations, including oversight. In practice, that means organisations should be clear about where people remain “in the loop,” where they review outputs, where escalation is needed, and who is accountable when AI output is wrong, biased, incomplete, or unsafe. (NIST)

This is not anti-AI. It is good design.

Consultation is not a soft extra. It is part of successful adoption.

Too many organisations communicate AI adoption after the key decisions have already been made. That is often when fear starts to grow.

Employees may wonder: Is this here to help me, monitor me, speed me up, reduce my role, or quietly replace parts of my job?

Those are not irrational reactions. They are predictable reactions when change is introduced without enough clarity, involvement, or trust.

OECD research indicates that training and worker consultation are associated with better outcomes for workers. OECD and EU-OSHA materials also point to worker involvement as important for adoption, trust, safety, and wellbeing. (OECD)

That means discussion should happen before rollout, not just after concerns arise. Staff often know where processes break down, where the workload is unrealistic, and where automation may help or harm. They can also identify practical risks leaders may miss.

Useful questions to ask employees before adoption include:

• Which parts of your work feel unnecessarily repetitive?
• Which  tasks most drain time and energy?
• Where  would assistance be useful?
• Where would automation create anxiety or risk?
• What would you never want removed from the human part of the role?
• What  support or training would help you use new systems confidently?

That kind of dialogue improves implementation because it improves design.

During adoption: training should be ongoing, practical, and role-specific

One-off training is rarely enough.

If organisations want people to use AI well, they need more than a launch webinar and a policy document. They need ongoing, practical development. OECD has highlighted that AI literacy and broader training provision remain insufficient relative to need, while the World Economic Forum’s Future of Jobs 2025 report identifies both AI-related capabilities and human capabilities such as resilience, flexibility, and leadership as increasingly important. (OECD)

Training should include at least four areas.

First, tool competence: what the system does, what it does not do, and how to use it properly.

Second, judgement: how to review outputs critically, spot errors, and avoid over-reliance.

Third, role redesign: how the human job changes when some tasks are supported by AI.

Fourth, wellbeing and workload awareness: how to prevent “digital piling-on,” where AI is introduced but old processes, old expectations, and old workloads remain in place.

This last point is often missed.

AI can reduce friction, but poorly implemented AI can also create new forms of pressure: constant adaptation, increased monitoring, faster work expectations, ambiguity about standards, and a sense that people must now produce more in less time because the tool exists. EU-OSHA has warned that digital and AI-based management systems can create psychosocial risks if poorly implemented, and that worker involvement helps reduce stress and improve wellbeing. (healthy-workplaces.osha.europa.eu)

So training should not only teach staff how to use the system. It should help managers understand what responsible use looks like.

What does successful AI adoption actually look like?

Successful adoption is not simply “the system is live.”

Nor is it “people are using it.”

A better definition would be this: AI adoption is successful when it improves outcomes, supports better work, preserves appropriate human judgement, and does not damage trust, wellbeing, or accountability.

That means success measures should go beyond cost reduction or speed.

They should include:

• improved quality and consistency
• time saved on low-value tasks
• reduced duplication or admin burden
• clearer workflows
• appropriate human oversight
• confidence in use
• lower error rates
• stronger trust in leadership communication
• evidence that workload has improved rather than intensified
• employee wellbeing indicators

Wellbeing belongs on this list because poor wellbeing is not a side issue. It affects performance, absence, retention, trust, and the sustainability of change. WHO states that wellbeing in the workplace influences health and productivity, and that negative work environments or excessive job strain can lead to mental and physical health problems. In Great Britain, HSE reports that work-related stress, depression and anxiety continue to account for a large share of work-related ill health and lost working days. (World Health Organization)

If AI adoption increases output while also increasing exhaustion, confusion, fear, or emotional strain, it should not be described as fully successful.

It may be efficient in one narrow sense, but ineffective in a broader organisational one.

After adoption: measure, review, and adjust

The work does not end at implementation.

After adoption, organisations should review what is happening in reality, not just what was predicted in the business case. NIST’s framework is built around ongoing governance, measurement, and management of risk, not a one-time decision. (NIST)

That means asking:

• Is the system solving the original problem?
• Has it improved work, or just shifted the burden elsewhere?
• Where are people still having to correct outputs?
• Are staff more confident, or more anxious?
• Has workload reduced, or has pressure risen?
• Are managers using the tool responsibly?
• Are there unintended consequences for fairness, trust, or decision quality?

There should also be space for employees to report concerns safely and honestly. The most useful post-adoption review is not a glossy success summary. It is an honest assessment of what is working, what is not, and what needs redesign.

The best AI adoption is strategic and human-centred

The most useful conversation about AI at work is not “How much can we automate?”

It is “How can we solve real problems better, while protecting the quality of work and the people doing it?”

This means starting with purpose.
It means identifying where human input remains essential.
It means consulting employees before rollout, not just announcing decisions to them.
It means ongoing training, not one-off exposure.
It means defining success in human as well as operational terms.
It means recognising that employee wellbeing is not a nice-to-have metric on the edge of the dashboard. It is one of the clearest indicators of whether change is actually working.

AI can help organisations become more effective.
It can also help people do better work.

But only when adoption is not reduced to procurement, pressure, or speed.
Only when it is designed with judgement.
Only when it is reviewed honestly.
And only when the humans in the system still matter.

Panic Applications in the Age of AI

How to stay calm, employable, and in control when workplace systems start to change

AI is increasingly being introduced into everyday workplace systems: reporting, scheduling, document drafting, customer support workflows, research, analysis, and routine “admin-heavy” processes. For most organisations, the stated goal isn’t job losses; it’s speed, efficiency, quality, and competitiveness. Yet even when job cuts aren’t the intention, role redesign and redundancies can still happen as work is reorganised around what AI can automate or accelerate. (TechRadar)

That mismatch “this is meant to help productivity” vs “this might threaten my job”  is where many mid-career professionals begin to feel unsettled. Not because they’re irrational, but because uncertainty is stressful, and the rules of the game can feel like they’re changing mid-match.

One of the most common reactions I see at this point is what I call panic applications: applying widely and quickly “just in case,” without a clear strategy. It can feel productive. It can also lead to the same stress in a different building.

This article is your calm alternative.

What’s actually changing (and why it can feel personal)

A helpful way to think about AI is: it changes tasks before it changes job titles.

Many roles won’t vanish overnight but, the mix of tasks inside them can shift rapidly. The World Economic Forum reports that employers expect 39% of workers’ core skills to change by 2030, which tells us the pressure is not limited to “tech roles.” (World Economic Forum)

That’s why the workplace can feel strange during AI adoption:

• You  may be asked to use new tools without clear guidance.
• Work  that once took hours may be expected in minutes.
• Leaders  may promise “augmentation” while quietly reviewing headcount.
• Colleagues may worry in silence which fuels rumours and anxiety.

This is exactly the kind of environment that creates job insecurity, and research consistently links job insecurity with poorer psychological wellbeing (including higher anxiety). (PMC)

So if you’ve noticed your mindset wobble distracted focus, low-level dread, irritability, compulsive scrolling, sudden urgency to “do something” that’s not weakness. That’s a nervous system trying to create certainty.

The trap: panic applications

Panic applications usually start with a thought like:

“If I apply everywhere, I’ll be safe.”

The trouble is that urgency doesn’t automatically produce better decisions. It often produces more action, less clarity.

Vignette 1: “Escape at any cost”

Amira hears that AI is being rolled into her team’s workflows. Someone mentions “efficiency” and “restructures” in the same sentence. That evening she updates her CV and applies to 26 roles, anything with a familiar job title and a slightly higher salary. She gets an offer quickly and accepts, relieved.
By month two, the relief is gone: the new role has the same relentless pace, unclear priorities, and constant urgency just with different systems and new faces. Amira didn’t find safety. She found the same stress, repackaged.

Vignette 2: “The confidence collapse”

James is mid-career and respected, but AI makes him doubt himself. He assumes others are “ahead.” He applies for roles beneath his level because he wants certainty. He lands interviews, but his story is messy: he can’t clearly explain what he wants, what he offers, or why he’s leaving. The rejection emails feel like proof he’s “falling behind,” and his confidence drops further.

Panic applications don’t fail because people aren’t capable. They fail because the strategy is missing.

A calm, evidence-informed reframe

Here’s the reframe that reduces fear and improves outcomes:

1. AI is changing the value chain: routine tasks are easier to automate;      higher judgement work becomes more valuable. (TechRadar)
2. The advantage goes to people who can work well with AI not necessarily      those who know the most about AI. (TechRadar)
3. Uncertainty impacts wellbeing and performance, so you need a plan that stabilises      your mindset while strengthening employability. (PMC)

That’s exactly what the Thrive Forward pathway is designed to do.

The Thrive Forward way: move with Clarity, Confidence, Strategy, Decision

1) Clarity: do a role/task audit (not a job-title identity crisis)

Before you apply anywhere, map your role into tasks. Write down your main tasks and label them:

• Automate: AI can do most of this reliably
• Accelerate: AI can help, but you must steer and verify
• Anchor:     human-led value (judgement, relationships, leadership, accountability)

This reduces panic because it replaces “I’m doomed” with specifics:

• What is actually changing?
• What stays human-led?
• Where can I move up the value chain?

2) Confidence: strengthen your “AI-resilient” value story

Employers don’t hire panic. They hire clarity.

Your story needs to answer:

• What outcomes do you deliver?
• What problems do you solve?
• How  do you work smarter now (including responsible AI use where appropriate)?
• What  do you want next — and why?

This is where many mid-career professionals regain confidence: not by becoming “technical,” but by being able to articulate value in a changing landscape.

3) Strategy: choose targeted moves, not scattered applications

The World Economic Forum consistently highlights skill gaps as a major barrier for organisations. (World Economic Forum)
Translation: employers are looking for people who can step in and perform not people sending 70 applications hoping one sticks.

A Thrive Forward strategy focuses on:

• 2–3 target role families (not 20)
• a clear “ideal environment” (culture, workload, autonomy, progression)
• a small number of high-quality applications
• warm outreach conversations (your best shortcut to hidden roles)

4) Decision: choose well, not fast

When you’re anxious, any offer can feel like rescue. Thrive Forward decisions are slower and smarter.

You evaluate offers using real criteria:

• workload expectations and resourcing
• manager style and clarity of priorities
• role scope (is it truly different or just a rebrand?)
• growth and skill development
• sustainability (does your body relax or brace?)

A final vignette: what “AI-resilient” looks like

Sana notices AI tools are being introduced to speed up reporting. Instead of panicking, she does a task audit and realises the reporting output is exposed but,  stakeholder advising is not. She learns to use AI for first drafts and scenario summaries, then applies her judgement to refine and verify. She shares an improved workflow with her manager.
Her visibility rises, her work becomes more strategic, and she becomes harder to replace  because she is now operating where human value is clearest.

That’s the goal: calm competence + clear positioning + deliberate movement.

If AI is arriving in your workplace, here’s your next best step

If you feel the urge to start applying everywhere, pause and do this instead:

• Write down your top 10 tasks.
• Mark: Automate / Accelerate / Anchor.
• Identify one “Anchor” strength you want your next role to use more.
• Draft  a 2-sentence value statement (outcome + strength + impact).

You don’t need panic. You need a plan.

If you’d like support building yours, I help mid-career professionals become redundancy-ready with out rushing using the Thrive Forward pathway: Clarity, Confidence, Strategy, Decision.

I’ve led teams, managed budgets, and delivered results. Yet when I first became Head of Department, I quickly realised what was missing: I’d had no real preparation for leading people.

The in-house support was helpful, but not enough—so I invested in a management diploma and used it to transform how I led.

Fast-forward to today: the data shows many women still don’t get the structured support they need at the moments that matter. That’s where tailored coaching comes in.

• Progress at the top is fragile. The FTSE Women Leaders Review (2025) reports women now hold 45.7% of leadership roles across FTSE 350 Executive Committees and their direct reports—a slight decrease on last year.
   
• Executive pipelines have slipped. Women Count 2024 shows women hold fewer than one-third of Executive Committee positions—the first decline in eight years.
   
• The “broken rung” persists. Globally, for every 100 men promoted to manager, only ~89 women are promoted, with worse outcomes for women of colour.
   
• Pay gaps remain—and widen with age. UK mothers earn substantially less than fathers each week, and the graduate pay gap emerges within five years of leaving university.
   
• Macroeconomic stakes are high. PwC’s Women in Work 2025 highlights that raising female participation and progression would significantly boost UK productivity.
   

 “Even as board representation improves, the leadership pipeline remains leaky and uneven. Women face headwinds at every stage.”
 

Coaching isn’t a perk—it’s a performance intervention. Research shows workplace and executive coaching improves leadership behaviours, goal attainment, engagement, and wellbeing.

• A 2023 meta-analysis reports significant positive effects on organisational outcomes like performance and goal achievement.
   
• Executive coaching studies consistently show moderate, reliable effects across multiple trials.
   
• Organisations that embed coaching cultures see wider benefits in retention and performance.
   

 “Well-designed coaching tied to clear goals delivers measurable results—for leaders and for the organisations they serve.”
 

1. Mastering the first step up – Breaking through the “broken rung” with skills in delegation, feedback, and strategic thinking.
    
2. Strategic visibility without burnout – Gaining recognition while managing workload sustainably.
    
3. Negotiation and influence – Building confidence to secure resources, pay equity, and influence.
    
4. Career-life design – Protecting momentum during maternity, caring, and perimenopause transitions.
    
5. Resilience and energy management – Preventing burnout in high-stakes, visible roles.
    

• Diagnostic clarity – 360 feedback and measurable goals.
   
• Personalised leadership plans – Linking habits directly to business outcomes.
   
• Evidence-based skill sprints – Negotiation scripts, sponsorship asks, boundary-setting.
   
• System alignment – Engaging sponsors and managers to embed change.
   
• Measured impact – Tracking promotions, pay movement, team engagement, and wellbeing.
   

• Protect the pipeline by investing at first-line management and mid-career inflection points.
   
• Pair coaching with systemic levers such as flexible work and sponsorship.
   
• Measure what matters—promotion rates, retention, and pay equity progress.
   

 “Without tailored coaching, organisations risk losing female leaders at the very moments that shape the diversity of their future leadership.”
 

The Market Reality:

• Gains in women’s leadership remain fragile.
   
• The “broken rung” still limits early-career promotions.
   
• Pay and progression gaps widen with age and motherhood.
   

Why Coaching Works:

• Evidence shows coaching improves leadership, performance, and wellbeing.
   
• Tailored programmes prepare women for transitions and prevent burnout.
   
• Organisations benefit through retention, stronger pipelines, and measurable ROI.
   

• FTSE Women Leaders Review (2025). UK Government / Hampton-Alexander successor review.
   
• The Pipeline (2024). Women Count 2024: Holding Women Back.
   
• McKinsey & LeanIn.org (2024). Women in the Workplace.
   
• Office for National Statistics (2024). Gender Pay Gap in the UK.
   
• Financial Times (2025). Graduate earnings gap analysis.
   
• PwC (2025). Women in Work Index.
   
• Jones, R.J., Woods, S.A., & Guillaume, Y.R.F. (2023). Meta-analysis of workplace coaching effectiveness. Journal of Occupational & Organisational Psychology.
   
• Grover, S., & Furnham, A. (2021). Executive coaching outcomes: A systematic review. Frontiers in Psychology.
   
• CIPD (2022). Coaching and Mentoring in Organisations.

 Like many women stepping into leadership for the first time, I had to build the toolkit on my own. Today, tailored coaching provides that toolkit by design—so women don’t have to “figure it out alone,” and organisations don’t lose momentum (or talent) at the exact moments that determine the diversity of their future leadership.

If you’d like a focused conversation about your context—whether you’re a woman leader preparing for your next step, or an HR/OD leader building a stronger pipeline—I offer targeted 1:1 programmes and cohort-based sprints aligned to your goals and metrics.

Contact: info@bjsandiford.com