How AI can best augment creatives

This exploration reveals not just technological possibilities, but fundamental questions about the nature of creative work itself.

The evolution of creative partnerships

The advertising industry provides an illuminating case study. Since Bill Burnback co-founded Doyle Dane Bernbach (DDB) in 1949, the copywriter-art director pairing has dominated creative agencies. This unique collaboration model suggests something profound about human creativity: we perform better when we have thought partners. The question becomes whether AI can fill a similar role for the billion knowledge workers globally who may lack such partnerships.

Move 37’s founding hypothesis centred on this concept. What would happen if everyone had an AI thought partner for work? The vision wasn’t human replacement, but rather extending the collaborative model that already proved successful in advertising to a broader creative landscape. This perspective shifts the conversation from fear to possibility.

The attention revolution

The timing proved fortuitous. The seminal paper ‘Attention Is All You Need’ emerged in April 2017, coinciding with Move 37’s inception. This research defined a new neural architecture that had not existed before, fundamentally changing everything that followed. The transformer architecture enabled the Large Language Models that now power creative workflows across industries.

Early explorations involved smaller language models and LSTM networks, technologies that seem ancient by today’s standards. When GPT-2, GPT-3, and GPT-4 arrived, they transformed theoretical possibilities into practical realities. This evolution illustrates how quickly the AI landscape changes and why early adoption, despite initial limitations, provides crucial learning advantages.

The reality of AI product development

Implementing AI in creative workflows requires rethinking traditional product development principles. AI introduces non-deterministic technology into environments that traditionally relied on predictable outcomes. This creates interesting tensions with established design thinking frameworks around what’s desirable, feasible and viable.

The challenge intensifies because even the creators of Large Language Models don’t fully understand how they’ll perform in specific contexts. Success requires combining subject matter expertise, local data, robust guardrails and proper evaluations within LLM-based systems. This creates highly niche-specific contexts that demand experimentation and prototyping rather than theoretical planning.

Client perspectives and implementation challenges

Enterprise adoption patterns reveal distinct phases. Initially, executives experienced the magic of ChatGPT and wondered about business applications. The gap between that consumer experience and enterprise implementation remains substantial. Over time, organisations develop better intuition about viable use cases, often through observing industry examples and press releases.

The most successful implementations combine individual productivity initiatives with organisational transformation efforts. Companies like realestate.com.au demonstrate this dual approach: encouraging safe prompting experimentation, while simultaneously pursuing product development initiatives. This prevents the uneven knowledge distribution that occurs when some team members advance rapidly while others remain beginners.

The inverted 80-20 rule

Jeremy Somers from Not Content articulated a crucial insight about AI-augmented creative workflows. Traditional productivity follows an 80-20 rule, but AI inverts this relationship. Humans handle the first 10% of any project through relationship building, understanding requirements and high-level strategic thinking. AI manages 80% of the production volume. Humans then contribute the final 10% through curation, selection and refinement.

The critical insight: while humans perform only 20% of the work, they contribute 80% of the value. This framework helps teams understand how to structure AI-augmented workflows and addresses concerns about human relevance in creative processes.

The ownership question

An unexpected challenge emerges around intellectual property ownership of prompts and AI-generated content. Creative professionals develop sophisticated prompting techniques that tap into their expertise, taste and experience. These become deeply personal workflows that professionals feel ownership over, similar to traditional creative methods.

This creates interesting dynamics when professionals leave organisations or share knowledge with team members. The question of whether prompts constitute company IP or personal intellectual property remains largely unresolved. The principle that emerges: individuals must own their outputs completely, being able to defend every word and remember it weeks later.

Critical thinking in the AI age

Perhaps the most significant challenge involves maintaining critical thinking capabilities. Large Language Models naturally produce averaged, middle-ground outputs without specific context and expertise. The risk lies in accepting this banal output rather than pushing for specificity and uniqueness.

The solution involves using AI as a thought partner rather than a content generator. This means leveraging the Socratic method, asking better questions and bringing personal expertise into the collaboration. The goal becomes thinking with AI rather than having AI think for you.

Collective intelligence perspective

Reframing Artificial Intelligence as collective intelligence provides a useful perspective. AI systems contain the collective human knowledge and experience, but this remains broad and generic without specific input. Users must bring their unique experiences, expertise and perspectives to create something interesting and valuable.

This perspective emphasises human agency and responsibility. The democratisation of intelligence and insight creates unprecedented opportunities for research, discovery and lateral thinking. However, realising these benefits requires active engagement rather than passive consumption.

Organisational change management

Successful AI adoption requires addressing very human concerns. Three common patterns emerge: protective behaviour around prompts, fear of being discovered using AI tools and paralysis from feeling behind. Most people apologise for their perceived lack of progress, not realising that AI knowledge remains unevenly distributed globally.

Effective approaches include regular prompt sharing sessions, open discussions about experiences and creating safe environments for experimentation. These social elements prove as important as technical implementation for successful organisational adoption.

The agency question

Individual and organisational agency emerge as crucial factors. Some people naturally embrace AI tools and explore possibilities, while others feel constrained, despite having access to the same technologies. This difference often relates to personality traits and how individuals view their work in relation to their identity.

Organisations must consider whether they create environments that support or hinder AI adoption. This involves both individual permission-giving and structural support for experimentation and learning.

Future implications

The rapid pace of change means that much existing technology needs rebuilding from scratch, rather than incremental AI additions. Large Language Models serve as fundamental building blocks for new systems rather than add-on features for existing ones.

This creates both opportunities and threats. The barriers to creating sophisticated applications continue decreasing, enabling individuals and small teams to build what previously required large organisations and significant resources. This democratisation of capability represents a fundamental shift in how innovation occurs.

The question of how AI can best augment creatives doesn’t have simple answers. Success requires balancing technological capabilities with human expertise, individual agency with organisational support and efficiency gains with creative quality. The most promising approaches treat AI as a thought partner that amplifies human capabilities, rather than replacing them.

The future of creative work likely involves deeper human-AI collaboration, where the uniquely human elements of experience, taste, judgment and wisdom become more valuable, not less. The challenge lies in developing these collaborations thoughtfully, ensuring that AI enhances, rather than diminishes human creative potential.

As we navigate this transformation, the focus should remain on empowering creative professionals to do their best work, using AI to handle routine tasks, while preserving and enhancing the essentially human elements that make creative work meaningful and valuable.