Executive Summary

Generative AI is a workplace tool that can offer large efficiency advantages, but so far, its boost to corporate profitability has been elusive. This is largely for two reasons: the first is the widespread, incorrect assumption that Generative AI on its own, without an assist from human originality, will permanently improve profitability; and the second is mistakenly treating AI adoption as a technical problem, rather than as a challenge for workplace culture.

Hupside’s new product, Hupchecker, allows businesses and other organizations to avoid these errors and to realize Generative AI’s promise. Hupchecker provides a revolutionary, easy, and rapid assessment of a person’s use of original thinking to solve problems. In an AI-driven workforce, originality is the critical component of a person’s successful adoption of AI. Hupchecker users will know which employees have the ability to use originality after an AI transformation to help their business or organization create lasting competitive advantage.

AI Adoption Needs Something New

Generative AI has been touted as a workplace efficiency tool, yet as many as 95 percent of organizations report that they have not obtained any financial benefit from their spending on AI tools. This raises the question whether AI transformation is worth doing at all. We believe that the answer is yes, if two things occur. First, businesses and organizations need to appreciate both the limitations of Generative AI and the role of human originality in breaking through these limitations. Second, employers must encourage workers to use AI by showing them how using AI will make them more valuable, rather than less.

Although it sometimes feels like the adoption of AI in the workplace has been occurring for a while, we are still very close to the beginning of the experience when we compare the AI timeline with that of other technology adoption cycles. Far too often during this initial period, businesses and organizations have looked at AI transformation as a technology opportunity: add it to an operation and process to generate cost savings and efficiency. It’s not surprising that people in an organization can be resistant when they are being told to use a technology they don’t understand. Or worse, being told that they are no longer needed because the technology can do their jobs.

While some dismiss this natural reaction to AI adoption as the inevitable result of business efficiency, the reality is this: even state of the art of Generative AI makes people who can use it well more valuable. Unlike other technologies that can be evaluated based only on efficiency and labor savings, Generative AI is a double-edged sword. Its efficiency comes at a price: sameness and declining differentiation. Generative models like GPT or Claude are designed to rely on probability. They identify and provide what’s likely, rather than what’s different. When used across many contexts and users, this probability-driven logic homogenizes output instead of diversifying it. What appears fresh or clever at the individual level is often lost to sameness at large scale.

This means that focusing on AI transformation without focusing on maintaining differentiation through originality creates a strategic dilemma. Although job cuts while pursuing AI transformation can result in almost instant margin improvement (fewer labor costs in the near term), the persistence of that improvement is much less certain. In a world where AI will allow all its users to capture efficiency in the same way, businesses will find that their efficiency gains are matched by their competitors.

That’s why businesses need novelty to build lasting competitive advantage: they need to stand out as distinct and to deliver something that their customers will value. True novelty, not simulated novelty derived from AI and then immediately available to all, requires human participation in the AI-driven process. AI combined with human original intelligence will create competitive advantage. The worker who is not afraid of AI tools but uses them as an assistant to come up with unique answers will be the essential driver of competitive advantage for business.

Getting to that point requires that those leading AI transformation projects treat AI transformation as a “bottom up,” not a “top down,” cultural change. This, in turn, will require two important steps. The first is adopting change management techniques that encourage organizational “buy in,” along with individual excitement and acceptance. The second is adopting an assessment process that surfaces how an individual will work with AI successfully – in other words, their ability to bring added value to the AI output. Hupside offers an assessment tool that does exactly that.

Why Originality Matters

We often imagine AI as a tool that supports creativity. However, the actual relationship is more nuanced. The best users of AI apply it to expand the idea space, and not just to pass on what the AI has told them. People who just rely on the AI output are likely amplifying the largest limitation of AI and LLMs: their reliance on probability.

Businesses are already seeing the effect of AI’s homogeneity in many places: job applications, marketing content, product design and work product generally. Some effects are more insidious: customer dissatisfaction, strategic stagnation, and employee disaffection. The sameness of AI output is undeniable. The more we rely on AI, the more our collective thinking begins to flatten, narrowing the idea space we had hoped to expand.

For many businesses whose competitive advantage can come only through differentiation, using AI without incorporating human originality will result in margin compression, not improvement. Even for the few businesses that can succeed through efficiency alone, AI sameness will eventually cause stagnation.

There is an additional reason why determining an individual’s predilection for original thinking is essential.  The originality competency of a person is a direct predictor of their  likelihood of adopting new tools to solve problems. Established research on creativity in humans (e.g., their desire to find novelty) shows that the higher a person’s predilection to think originally, the more likely they are to use new tools (such as AI) to seek novelty.

Introducing Hupchecker: The First Assessment Tool for AI Transformation

Hupside has developed a new assessment tool – Hupchecker – that identifies the most important personal attribute necessary in an AI environment: an individual’s unique approach to using original thinking to solve problems.

Hupchecker is a software-as-service product. An administrator uses it to create and deliver short and entertaining problem-solving challenges. The challenges have been designed by Hupside’s creativity experts to rapidly surface an individual's or team’s specific approach to the use of original thinking in problem-solving. Hupchecker then measures where each human’s response is located within the “idea space” of all prior answers to the same question. The idea space—the plotting of how answers to a common question compare—reveals who is creating something new. The idea space is a heatmap for finding original thinking.

Hupchecker then uses the measurement of a person’s output in the idea space to calculate the person’s Original Intelligence Quotient (or “OIQ”). An OIQ can be measured against other humans or against the output of one or more of the market leading LLMs. This allows for the concurrent measurement of human performance against humans and AI. It also allows for diagnostics of people who are using AI to create something new, rather than just paraphrasing the answer they got from a prompt.

The OIQ is then used in turn to position the individual within one of ten OIQ specific archetypes that predict how the person will approach problem solving. All this data is provided to the administrator through a graphical control panel, which allows for individual and cross group comparisons.

Hupchecker will take AI transformation in a new direction. It creates a metric against which results can be objectively evaluated and tracked. Technological efficiency is then combined with changes in human performance to create competitive advantage. It will allow businesses and organizations to understand their workers’ originality potential in an AI environment and to employ situation-specific tactics to achieve AI adoption and integration: customize training, match people with the right AI tools, create AI-ready teams, avoid AI induced group-think and AI over-reliance, monitor progress, and benchmark success.  People can be trained and led to AI adoption in ways that will connect to their personal problem-solving approach. True AI transformation that includes the assessment and promotion of human originality will achieve profit growth and succeed.

Hupchecker is a Transparent, Objective, and Independent AI Technology

Hupchecker’s proprietary technology is computationally efficient and self-contained. It is first and foremost a measurement of human originality. It uses external LLMs only to provide a point of comparison. Therefore, it can operate and provide valuable insight even if all currently popular LLMs ceased to operate. This is a significant differentiator from most current AI companies, which rely on these LLMs to create their own commercial products.

Hupchecker identifies an individual’s capacity for original thinking through an objective determination of the location of their output in an idea space. It is an indicator of a person’s originality competency. It is not a measurement of the value of the response. That task is reserved for the human observer. Hupchecker does not judge people; it provides a heatmap of competencies and potential in an AI-driven economy. Built into its operation is a place for administrators to assess and accept each individual response before an OIQ or archetype is rendered.

From Insight to Action: Using Hupchecker for Successful AI Implementation

Change management often divides employees into early, middle, and late adopters. Early adopters support new initiatives, late adopters tend to resist, and the middle group observes how others respond before deciding what to do. Leaders know their approach to the early and late adopters directly influences the rest. With its archetypes, Hupchecker can predict and provide ongoing measurement of how AI adoption is progressing. And, after the AI transformation is complete, whether the organization’s people are continuing to maintain their ability to problem solve through original thinking.

Hupchecker categorizes people into ten archetypes which are directly useful in charting successful AI transformation. For instance, “Expanders”—people Hupchecker identify as strongly inclined toward originality—are typically enthusiastic about using new tools and developing innovative problem-solving strategies with them, making them likely early adopters.

By contrast, “Focal” individuals—people who predominantly rely on traditional methods—are generally more cautious in adopting new tools and techniques, waiting until they feel confident in their use and until these tools gain broader acceptance – in other words, late adopters. Other OIQ types, such as Augmenters and Connectors, are more likely to combine existing concepts with their own original ideas. With suitable training and support, these individuals adopt new methodologies at varying rates and frequently invent practical enhancements that may not have been operationalized by Expanders.

To facilitate effective cultural transformation and AI deployment, businesses can leverage insights from Hupchecker to strategically assign pilot roles, sequence training initiatives, staff change management networks and assemble project teams. Importantly, the organization can make strategic decisions about how to compose teams, fill roles, and operate: the aim is to ensure that each individual occupies a position that aligns with their originality capabilities and potential. For example, this could mean pairing a small number of Expanders with a larger group of Augmenters and Connectors, while providing Focal contributors on the team with structured support and peer mentorship.

A Suggested Playbook for Successful AI Adoption

There are many successful approaches to adapting new tools and approaches in a business or organization. What follows is an example a one playbook for how to use Hupchecker in combination with proven approaches to organizational change to effectuate an AI transformation that results in permanent financial improvement.

Stage Setting

Adopt human-focused guiding principles.

Awareness & Urgency: Communicate the ‘why;’ share pilot outcomes and risks of sameness.

Desire & Coalition: Recruit first adopters across functions; recognize early contributions.

Knowledge & Vision: Offer role‑based learning; translate OIQ insights into team charters.

Ability & Barriers: Pair profiles for peer support; address evaluator bias in QA gates.

Reinforcement & Wins: Celebrate learning milestones; publish quick wins; repeat OIQ snapshots.

Execution

Phase 1 — Prepare

• Define business outcomes and in‑scope workflows.

• Establish objective sponsorship, governance, and escalation routes.

• Communicate the ‘why’ in plain language and set expectations for learning.

• Run a Hupchecker baseline across affected workers.

Phase 2 — Pilot and Learn

Assign pilots to Expanders, capture stories and quick wins.

• Provide initial tool training for Augmenters and Connectors.

• Provide creativity training to targeted team leaders to enhance their ability to generate original output.

• Stand up peer coaching and office hours; address failure modes (over‑trust, hallucinations).

• Track early indicators: adoption, time‑to‑proficiency, rework rates, and experience feedback.

• Run Hupchecker on trained individuals to begin to benchmark progress.

Phase 3 — Scale and Embed

• Build/Form/Assemble/Organize mixed‑profile teams (mix of Expander, Augmenter, Connector and Focal depending upon strategic implementation objectives) for rollout waves.

• Publish a “new rules of success” guide tied to skills, safety, and outcomes—reinforce that the changes were driven by objective metrics that are clear and easy to understand.

• Use Hupchecker to measure shifts in OIQ distribution and expanding‑the‑idea‑space metrics over time; recognize growth.

• Reinforce new norms with communities of practice and continuous‑learning nudges.

• Target continued training to originality competencies.

Phase 4 — Sustain and Improve

• Use Hupchecker to re‑assess OIQ periodically; identify capability lift at the individual and team level.

• Retain and develop high‑impact profiles; offer growth paths for all contributors.

• Refresh guardrails and evaluation methods to avoid AI evaluator bias.

• Publicize outcomes and ROI using business KPIs and example narratives.

Suggested Benchmarking for Progress and ROI

Leading indicators

• Adoption and active use (by workflow).

• Time‑to‑proficiency and reduction in rework/defects.

• Shift in expanding‑the‑idea‑space metrics and OIQ distribution.

• Diversity of ideas in ideation backlogs and shipped features.

• Employee experience and trust in AI guardrails.

• New hiring and promotion assessment includes originality competencies and OIQ type.

• Customer success metrics.

• Time-to-job-completion metrics.

• Time-to-market for innovation.

Lagging indicators

• Cycle time and cost‑to‑serve improvements.

• Win rates and NPS/CSAT changes tied to differentiated offerings.

• Retention of high‑impact talent and internal mobility.

• Revenue/EBIT contribution from AI‑enabled products and processes.

Recommended Additional Reading

Below are some of the articles and research papers that were used in the preparation of this White Paper.

Doshi & Hauser (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content.

Anderson, Shah, & Kreminski (2024). Homogenization Effects of LLMs on Human Creative Ideation.

Liu, Moosavi, & Lin (2023/2024). LLMs as Narcissistic Evaluators.

Wataoka, Takahashi, & Ri (2024). Self‑Preference Bias in LLM‑as‑a‑Judge.

Dell’Acqua et al. (2023). Navigating the Jagged Technological Frontier.

Prosci (2024–2025). ADKAR Model and Methodology.

Kotter Inc. (2024–2025). The 8‑Step Process for Leading Change.

Challapally, Pease, Raskar & Chari (2025). The GenAI Divide – State of AI in Business 2025.

Wenger & Kenett (2025). We’re Different, We’re the Same: Creative Homogeneity Across LLMs.

Liang, W., et al. (2024). The Widespread Adoption of Large Language Model-Assisted Writing Across Society.

Anderson, Shah & Kreminski (2024). Homogenization Effects of Large Language Models on Human Creative Ideation.

Moon, Green & Kushlev (2024). Homogenizing Effectof a Large Language Model on Creative Diversity.

Photo by Redd Francisco on Unsplash