Summary
AI in Creativity: What Really Changes for Brands
TL;DR: At the HUMAN X Conference, Abhay Parasnis from Typeface, Jessica Powell from AudioShake, and Mikayel Vardanyan from Picsart explained that AI in creativity does not eliminate human value: it shifts the competitive advantage to relevance, personalization, taste, trust, and the ability to orchestrate increasingly complex workflows.
At the HUMAN X Conference, the panel moderated by Nora Ali tackled a crucial question for marketers, media teams, and tech companies: how is creativity changing when content production accelerates at the speed of AI? The answer that emerged is clear. Today, the problem is no longer just creating content quickly. The real issue is creating relevant, contextual content that is consistent with the brand and credible to the audience.
In summary: AI drastically reduces production time and costs, but makes what cannot be easily standardized more important. The most important thing is that value shifts from mere execution to the quality of decisions.
HUMAN X Conference: The New Balance Between Speed and Quality
The discussion involved three complementary perspectives. Abhay Parasnis, for Typeface, brought the viewpoint of AI applied to personalized marketing. Jessica Powell, for AudioShake, demonstrated how AI can transform audio workflows that were previously impossible or too slow. Mikayel Vardanyan, for Picsart, described a broader vision, where multiple agents and models reduce operational friction for creators and teams.
Context matters. This is not a theoretical debate on automation, but a practical reflection on how campaigns, assets, and creative processes are already changing within organizations.
AI in Creativity: Where Friction Disappears and Where Bottlenecks Remain
According to Parasnis, AI has already eliminated two historical frictions: the cost and time of content creation. With increasingly effective models for text, images, video, and audio, production is no longer the main obstacle. The bottleneck shifts elsewhere: to the ability to be relevant in real-time and to truly personalize messages and touchpoints.
This means that an abundance of content does not automatically guarantee effectiveness. On the contrary, it can increase noise.
The new brakes are mainly three:
- relevance at the right moment
- deep personalization on audience, product, and channel
- human judgment on quality, tone, and opportunities
Parasnis emphasizes a crucial aspect: when everyone accesses similar platforms, the advantage is not just in the tool, but in how judgment, trust, and brand voice are applied.
What Does “Taste” Mean in the AI Era
One of the strongest ideas that emerged from the panel is that “taste” does not have a universal definition. Parasnis links it to context: B2B and B2C have different criteria; acquisition and engagement follow different logics. This is precisely why taste is difficult to fully automate.
In this scenario, taste means judgment ability. It means understanding which message to send, to whom, when, and in what form. It also means avoiding communications that are aesthetically correct but strategically irrelevant.
For Typeface, a central element of taste is extreme relevance. A brand that communicates without considering the existing relationship with the customer generates friction, not value. In an ecosystem saturated with content, even a well-crafted asset loses effectiveness if it arrives out of context.
The second component is trust. Parasnis ties the concept of taste to the perception of authenticity and reliability of the message. In an environment where AI-generated content is becoming more common, the brand must convey consistency and correctness, not just creative brilliance.
Key Question: How Do You Build Trust with AI-Generated Content?
Answer: Trust is built with more pertinent, less invasive communications that are more consistent with the real profile of the customer.
Parasnis cited the case of a large U.S. telecom company that sent identical emails to millions of users, despite having a lot of contextual data on plans, purchasing behaviors, and consumption profiles. The intervention was not to “produce more content,” but to send fewer, more targeted content. The reported result was an improvement of about 93% in click-through rate and downstream conversions, thanks to more relevant messages for each segment.
The strategic lesson is clear: effective personalization does not coincide with multiplying outputs. It coincides with better message selection.
How to Avoid AI Content Homogenization
The risk of homogenization was explicitly addressed. Parasnis highlights the problem of “AI slop”: when many use similar tools with similar logics, the results start to resemble each other.
Typeface’s response is based on two levels.
- Brand context and system of record
The idea is to train systems with the specific context of the brand: voice, product catalogs, audience, data, and analytics. In this logic, value does not derive only from the base model but from the layer of context that makes the output distinctive and useful for that specific business.
- Cross-channel orchestration
Personalization cannot stop at a single touchpoint. Parasnis emphasizes that a personalized email leading everyone to the same landing page does not create a truly relevant experience. Therefore, it becomes essential to coordinate email, web, and social in a single personalization logic.
This means that the future of AI in marketing is not just generating assets faster. It is orchestrating consistency between assets, channels, and moments of the customer journey.
AudioShake: When AI Makes the Impossible Possible
Jessica Powell added a very useful perspective by distinguishing between generative AI and other machine learning applications. In the case of AudioShake, the cited technology is source separation, which means separating the components of an existing audio. Simply put, source separation means isolating voice, music, noise, or other elements from the same sound track.
This capability changes workflows in two ways.
The first is productive: it allows for faster work on broadcast, social, and post-production content.
The second is even more interesting: it makes materials usable that were not usable at all before. Powell gives concrete examples, such as clips recorded in noisy environments or content that could not be published due to background music. In these cases, AI not only accelerates but unlocks new creative and distribution possibilities.
Key Question: Where Does Human Judgment Enter AI-Enhanced Creative Workflows?
Answer: It enters when the system produces components or options, but an expert is needed to decide the final form, aesthetic rendering, and creative priorities.
Powell explains that once audio tracks are separated, every subsequent decision remains profoundly human. An audio engineer, a post-production team, or an A&R professional apply years of experience to determine how a mix should sound, how to adapt it to a format, or how to enhance it in a new context.
The most important thing is this: AI expands the scope of human talent, it does not erase its function in high-impact qualitative steps.
Picsart: Less Technical Choice, More Focus on the Problem to Solve
Mikayel Vardanyan addressed an increasingly relevant issue: decision fatigue caused by the proliferation of models. For the end user, understanding which model to use for video, images, or other tasks is often an unnecessary overhead. Picsart tries to reduce this friction by integrating numerous models and having them work behind the scenes based on the use case.
Here emerges another relevant point: with the increase in automation, figures capable of orchestrating are needed. Vardanyan talks about people who can understand process, taste, quality, and the right moment of application. It is not enough to use AI tools. They must be coordinated within a business objective.
Also interesting is his observation of taste as an experimental phenomenon. In many cases, what really works does not coincide with the personal taste of the creator, but with what resonates with specific audiences, such as Gen Z or Millennials. This brings AI-driven creativity back to a very concrete ground: testing, audience response, and continuous adaptation.
The Skills That Really Matter in the AI-Driven Creative Economy
In the end, the panel shifted to teams, roles, and skills. Jessica Powell highlighted the role of the generalist, especially in dynamic contexts like startups and agile teams. A person capable of moving between design, social, communication, and content can generate a great operational advantage, while still being useful to pair with specialists where depth is needed.
Vardanyan reinforced this view with a broader perspective: understanding how the business works horizontally, from product to finance, from user needs to AI tools, becomes a decisive competence. His idea is that the market will increasingly favor hybrid figures and even solopreneurs capable of leveraging automation.
And when, in closing, Abhay Parasnis is asked which skill to choose today in two words, the answer is immediate: “taste and trust.”
What Brands and Creative Teams Should Do Now
A clear direction emerges from this panel.
- speeding up production is no longer enough
- relevance is a primary competitive lever
- personalization must be contextual and cross-channel
- human judgment remains crucial in high-value moments
- the strongest skills combine adaptability, business literacy, and critical thinking
In summary: AI in creativity does not lower the bar. It shifts it. Less weight on pure production, more weight on strategy, consistency, trust, and the ability to interpret what really matters to the audience.
FAQ
What Does AI in Creativity Mean?
AI in creativity means using artificial intelligence systems to accelerate, expand, or make possible the production of content in formats such as text, images, video, and audio. However, at the HUMAN X panel, it emerged that the real value arises when AI is guided by context, taste, and business objectives.
Will AI Replace Creatives?
The discussion does not suggest a linear replacement of creatives. Rather, it suggests a transformation of roles. When creating costs less and takes less time, human judgment, the ability to choose, quality control, and brand consistency become more important.
How Do You Prevent AI Content from All Looking the Same?
The answer indicated in the panel is to use proprietary context, brand data, audience knowledge, and cross-channel orchestration. Without these elements, the output risks resembling that of anyone else using the same base models.
What Skills Are Needed Today to Work in AI-Driven Creativity?
Both generalist adaptability and specialist skills are needed. In particular, the speakers highlighted the value of taste, trust, product understanding, business knowledge, and the ability to work across different functions.
Why Is Personalization So Important?
Because when content production becomes industrialized, what truly differentiates a brand is the ability to send the right message, to the right person, at the right time. Relevant personalization improves experience, trust, and performance.
To delve deeper into art in creativity and its impact, it is useful to explore educational paths like those offered by LABA or specialized institutes like SAE Institute.
Finally, for those who want to stay updated on the latest technological innovations, we highlight the launch of Grok 3 API, Elon Musk’s alternative to GPT-4o, now available for developers, and innovative tools like TurboLearn AI, which transforms PDFs, videos, and audio into personalized notes, quizzes, and flashcards.
