top of page
Search

The Power of Perceived Effort PART 2: Proof of effort vs judgement

4 days ago
9 min read
The judgement underneath the polish
The judgement underneath the polish

RECAP: In Part 1, we explored how professional polish became a proxy for something much harder to observe: the quality of the thinking behind it. This “effort heuristic” is particularly pervasive when the underlying quality of work is difficult to judge. AI is now disrupting that relationship. It can also produce compelling surface signals without guaranteeing equally strong judgement underneath. We now expect polish more than ever yet can no longer assume it signals quality. If polish is becoming a hygiene factor rather than a reliable signal of expertise, the question for Part 2 is what should replace it?


Perhaps the biggest mistake professionals could make in response to AI would be trying to manufacture visible effort simply to reassure clients that enough work was involved. Producing 80 slides when 20 would do is not rigour. In fact, this is becoming the essence of the AI productivity conundrum. If someone produces content in minutes that takes someone else hours to process, have we improved efficiencies or created busy work? 

The better approach is to make the judgement behind the work more visible.


Showing how hard you worked vs how hard you thought


Research on “operational transparency” has shown that people can value services more highly when otherwise invisible effort becomes visible, even when the underlying result was unchanged. In an AI-enabled professional environment, however, simply showing activity may no longer be enough.


The new equivalent of operational transparency may be intellectual transparency.

  • Tradeoff logic between business objectives, stakeholder needs, realpolitik and contextual domain considerations.

  • Assumptions that were challenged and the evidence that caused us to reject our original hypothesis.

  • What we still cannot confidently know but may be able to discover.


That is harder to fake because it demonstrates selection, rejection and contextual judgement rather than merely production. AI is exceptionally good at generating options so expertise increasingly needs to reveal why options were chosen and discarded.


Consulting models may need to look different


This is particularly important for consulting. Consulting has traditionally bundled several things together: access to expertise, analysis, intellectual property, experienced judgement and a highly polished artefact communicating the outcome.


AI begins to unbundle them as a polished artefact, generic framework, competent first-pass synthesis and even the appearance of strategic sophistication, which are all effectively becoming widely accessible and dramatically cheaper. However, that does not necessarily make ‘expertise’ cheaper. It simply exposes which part of the bundle was most valuable.


Consider the classic consulting 2x2 matrix. For decades, part of its persuasive power came from compression. Someone had apparently taken a messy problem and distilled it down to two dimensions that explained it. Today an AI system can propose countless plausible 2x2s before your coffee arrives so the rectangle is no longer the IP. The IP is knowing which axes most usefully explains the problem to facilitate an effective solution.

Likewise, anybody can ask an AI system for “five emerging consumer trends”. The value comes from knowing which trend is real, which is noise, which matters for this organisation, what evidence supports it and what decision it should change.


This suggests a useful shift in how professional work is often presented. Instead of showcasing only the finished recommendation, occasionally expose the intellectual scar tissue behind it. These are signals of thought, not simply signals of production.


  • What did you initially believe? What surprised you?

  • What did experience allow you to notice that a generic analysis would miss?

  • What alternatives were seriously considered?

  • Where is the recommendation vulnerable?


This creates a tension for consultants and knowledge workers - whether internally or externally your job is often “selling” increased certainty I.e. often unequivocal confidence packaged with clarity. This often leads to an oversimplification of the work. Opening the hood to your “process” may feel counter to this objective. Often to some degree you may be right, but if you are knocking out generic but polished and confidently delivered solutions this will become less valuable. Your competitive value may increasingly lean on your brand rather than the distinctiveness of your product/ service.


It’s important here to emphasise the balancing act between oversimplify vs. bogging things down by showing your workings vs. creating an experience of discovery.


The value of an idea is only realised when they are successfully implemented. We are now saturated with ideas and the challenge is more choosing the right one to follow through on. Obviously, this is about choosing the most effective solution but it also about choosing the idea people can get behind and helping motivate them to do so.


This is the value of the discovery experience which builds momentum for successful implementation i.e. manifesting shared belief and excitement to get around an idea - otherwise known as the ‘buy in’. That’s the human experience AI will never replace with volume of polished ideas - that moment everyone gets on the same page and our human superpowers really kicks in - sharing a conviction and vision to collaborate towards a common goal.


This sees experts acting more like curators guiding the solution seeking process rather than gatekeepers to accessing information. With knowledge synthesis and idea generation being taken over by the machines, lets face it, this discovery experience may also be one of the last bastion of human intellectual fulfillment in age of AI.


Removing effort vs. transferring it


This also changes what “effort” should mean. The old assumption was that a substantial deliverable reflected substantial effort by its creator but that’s no longer a useful metric.

Emerging workplace research has coined the term “workslop” for AI-generated material that looks finished but lacks enough substance or context to meaningfully advance the task. In research conducted by BetterUp Labs with Stanford Social Media Lab, 40% of surveyed US desk workers reported receiving such output in the previous month. Recipients estimated spending close to two hours resolving each incident.


That introduces an excellent new test of professional quality: Where did the effort go?

A 50-page AI-generated report may have taken the sender ten minutes and the recipient an hour to work out what matters. A five-page synthesis may have taken the sender two hours but saved the executive one.


Which is the more effortful piece of work? More importantly, which is more valuable?


When competent becomes common, distinctive becomes valuable


There is another consequence of AI raising the baseline. Generative AI can improve individual creative performance but there is evidence that it may simultaneously pull outputs towards one another.


A Science Advances experiment found that participants given generative AI ideas produced stories judged to be more creative, better written and more enjoyable, particularly among weaker writers. However, AI-assisted stories were also more similar to each other than stories produced without AI assistance.


There is a professional analogue worth considering. If everyone asks broadly similar systems to improve their executive summaries, sharpen their propositions, suggest their frameworks and generate their “three key implications”, we may get better average work while also getting considerably more sameness.


This changes the competitive landscape. When competent communication becomes abundant, distinctiveness starts carrying more information. This might be a genuinely surprising observation, a proprietary dataset, and contentious point of view, a recommendation subtly shaped by years of domain experience or a client-specific insight that would make no sense in another organisation.


These become disproportionately valuable because they are evidence that something more than generic pattern completion has occurred.


Six practical shifts for communicating ideas in the AI era


For anyone producing reports, presentations, proposals, thought leadership or strategic recommendations, I suspect a few principles will become increasingly important:


  1. Keep the polish, but stop relying on it to prove quality. Standards will rise, not fall. Clear writing and thoughtful design remain valuable because they reduce cognitive load. Treat them as hygiene factors rather than your primary evidence of expertise.


  2. Make choices visible. Show what was prioritised, rejected or changed. Selection is increasingly a stronger expertise signal than generation but don’t bog things down with complexity.


  3. Increase specificity. Generic fluency is cheap. Context is not. Names, numbers, observations, examples, constraints and proprietary evidence make work harder to interchange with somebody else’s.


  4. Expose uncertainty intelligently. Do not manufacture rough edges, but resist the temptation to make every argument artificially complete. Knowing what cannot yet be concluded is itself a form of expertise.


  5. Design for the recipient’s effort, not the creator’s effort. Before sending something, ask whether AI has actually reduced work or merely moved it downstream.


  6. Use the time AI saves to improve the thinking it cannot automatically validate. Challenge the obvious interpretation. Test the assumption. Speak to the customer. Look for the contradiction. Interrogate the recommendation. That is where the productivity dividend becomes a quality dividend.

 

Unbundling the commoditisation of expertise


This, ultimately, may be the more interesting consequence of generative AI. The debate is often framed as whether AI can replace human expertise and it is now undoubtedly making some of the historical signals of expertise vastly easier to produce. That does not mean these things no longer matter but that they no longer tell us everything they once did.


The professionals who prosper will not necessarily be those who resist AI in order to demonstrate their human effort nor those who use AI to produce the largest quantity of polished work. They will be the people who use cheaper execution to create more room for expensive judgement. More interrogation, context, original evidence, transparent difficult choices and responsibility for the answer.


For design, that means moving beyond decorating information towards deciding what genuinely deserves attention. For writing, it means moving beyond immaculate sentences towards specificity, argument and distinctive point of view. For consulting, it means moving beyond demonstrating the quality of the deliverable towards demonstrating the quality of the decisions embedded within it.


Perhaps the ultimate quality test becomes very simple. Take the polished slides away. Does the idea still stand up? If the answer is yes, the polish has done its proper job. It has helped good thinking travel. If the answer is no, we may simply be looking at a very convincing signal of effort that never actually happened.


The value of curiosity (and its origins in uncertainty)


On the surface these articles about perceived effort in the age of AI explore the games we play and business theatre required to not only communicate ideas but to communicate our value. At a deeper level, they touch on the existential crisis artificial intelligence represents to once ‘automation proof’ knowledge worker roles such as creative designers, researchers and strategic consultants.


Well-trodden territory in AI related blogs, so please forgive if you have heard it all before, but it might be worth reflecting here on what makes us humans so special when it comes to accumulating knowledge, answering questions, developing ideas and proposing solutions with confidence.


Machines are getting good at all of these.  The widely accessible supply of ‘solutions’ has never been greater. What is the comparative advantage and inherent value of a knowledge worker with this unprecedented ‘hyperinflation of certainty’?


As pointed out by others, perhaps as answers get cheaper, asking the right questions gets more valuable. As data becomes cheaper (and overwhelming), effectively communicating the right part of the answer gets more valuable.


Sure, it’s the whole ‘let’s stay positive AI is just a tool’ argument but it does bear consideration. Used right, AI is less artificial and more ‘augmented intelligence’. Despite the traditional effort required to produce answers, our special spark was more about coming up with the right questions - the power of our curiosity, which answers then serve. Don’t forget to let that part of your value shine through your packaged polish.


Einstein often asserted that the most important thing is to never stop questioning and that the key to finding the right solution is framing the right problem. He quipped that if given an hour to find a solution his life depended on, he would spend 55 minutes determining the proper questions to ask. Looking to another pinnacle of human thinking, Feynman was quoted as saying “I would rather have questions that can’t be answered, than answers that can’t be questioned.”


This is a time where ‘uncertainty’ feels like it’s becoming intolerable. Polished outputs and confident answers are expected almost immediately.  Let’s embrace admitting some ‘uncertainty’ and seeking ‘discovery journeys’ as fundamental to curiosity and developing truly great ideas that are unique and will work.


Most importantly, the value of ideas is only realised with follow through. Ideas and answers now proliferate cheaply and look convincing - inevitably just producing more noise and analysis paralysis. We need capable humans to communicate the right answers effectively to motivate a collective of people to believe in a common idea, pull together and act on it.  A very human power - for now (and forever we can only hope).


Authors

Tai Rotem is a consulting partner at MYMAVINS with several decades experience in consumer, financial services, public health, and social research.


Reach out to him at Tai@mymavins.com.au

 






Mitzi Cruz is the creative operations manager at MYMAVINS with seasoned leadership experience managing large-scale web development and design teams. Combining a career in creative direction and operations management with a background in Psychology, she translates research into compelling visual narratives and campaign-ready assets.






Bibliography


 
 
 

Comments

bottom of page