Key takeaways:

1. There is a growing recognition of the need to balance human insight with AI capabilities. While AI can enhance efficiency and generate content, the human element remains crucial for emotional connection and trust in communication.
2. Trust in communication is shifting from broad, weak ties (like social media influencers) to strong ties within smaller, more personal networks. Brands must adapt by being authentic and relevant to build trust in these intimate communities.
3. The integration of AI into communication strategies offers opportunities for hyper-personalization and precision targeting. However, brands need to be aware of the potential pitfalls, such as biases in data and the importance of maintaining meaningful human connections.

Participating in the discussion were:

  • Paul Holmes, founder and editor of PRovoke Media
  • Sarah DaVanzo, global chief innovation officer and head of strategic services and foresight at Porter Novelli 
  • Sylvia Park-Ekecs, global head of consumer experience and design at Sanofi
  • Erin Lanuti, chief innovation officer at Omnicom Public Relations Group
  • Michael Maslansky, CEO at Maslansky & Co
  • Katie Earlham, director of creative strategy at 72Point
  • Annalie Killian, strategic partnerships, DDB
  • Jonathan Brown, chief growth officer at Trend Hunter
  • Imari Oliver, founder and CEO of Bond & Play
  • Tim Hsu, head of business development, US at Leonardo.AI

The conversation has been edited for length and clarity.

Paul Holmes (PH): I think most people in communications are taking an incrementalist approach to transformation, focusing on doing things a bit faster, better, or cheaper—often emphasizing the cheaper aspect. However, Sarah is discussing a reinvention of processes based on new technologies, which differs from what I’m hearing from other agencies. I’m interested in hearing Sarah's thoughts and the reactions from everyone in the room.

Sarah DaVanzo (SD):  So let's start with you, Imari. What was your thinking about smart cities and how you are reframing that?

Imari Oliver (IO): You talk about language, and there's always bad baggage with it. When people mention smart cities, there's fear mongering, like the idea of 15-minute cities where the government wants to control everyone. But they’re not really smart; 50% of corporate data is often useless. You see cameras everywhere, yet crime persists. In New York, I wait two minutes for a traffic light with no traffic—shouldn't smart systems optimize that?

Cognitive cities involve actionable intelligence, always understanding and adapting. It’s about using data effectively, not just storing it. We want cities responsive to their residents. Ultimately, technology should serve people, as they are the final arbiters of its effectiveness.

PH: Let me ask you a quick question relevant to redesigning communication processes. When building a cognitive city, can you transform an existing one, or is it easier to ignore the mistakes of current cities and build a new cognitive city from scratch?

 

IO: We're building our own operating system for communities, allowing integration into any city. With around 20,000 incorporated cities in the U.S., if each adopts our model, it’s beneficial for us, akin to creating our own Amazon Web Services.

I also wanted to address gentrification displacement. Growing up, I've seen how new developments change the culture. We can help investors while upskilling long-term residents, creating new opportunities and bringing them along with growth.

SD: The term "cognitive" in Smart City prompts reflection—it's different. You mentioned humanity; it's a brain, a metaphor for things right?

Michael Maslansky (MM): We constantly battle concepts like generative engine optimization, a new language in the SEO revolution. As you've noted, 'he who controls the language controls the debate.' Establishing a term that generates positive associations is powerful, but if it evokes negative ones, it can lead to demonization.

Terms like "smart" in smart grid or smart home can be loved in some markets and hated in others. People instinctively associate different meanings with terms. If a term like "cognitive" has positive associations, it’s beneficial, but if not, it can fail. In the AI era and amidst filter bubbles, differing political views complicate communication, raising questions about which terms prevail and whether AI improves or worsens this dynamic.

Sylvia Park-Ekecs (SP): Ultimately, human plus AI is better. AI mimics human intelligence, but it operates on the principle of "garbage in, garbage out." Its effectiveness depends on clean data, which many large companies have neglected. Now, there's a scramble to re-architect data standards.

This raises a dilemma between business profitability and societal impact, affecting human behavior and communication. Different generations use language differently; for example, my first language is Korean, and my mom would correct my usage.

Communication also varies with mobile devices, where brevity is key. It all comes down to targeting your audience, which involves complex ecosystems of data, IT, engineering, costs, and product teams.

Jonathan Brown (JB): When we consider business development, it's clear that people are motivated by self-interest and what affects their lives. We care about technology only in terms of its impact on us. For example, I don’t care how my devices connect; I want to know how it benefits my life.

Most people don’t care about smart cities until they see how it affects them personally. If living in a smart city helps optimize my life, then I can support it. Ultimately, it comes down to what it means for you. We all naturally think about ourselves, and that’s not a bad thing.

Michael Maslansky (MM): When we frame messages, we categorize them by what it is, what it does, and what it means. Many tech companies fail because they focus on complex terms that lack meaning until they catch on, often forgetting the "what it means."

For example, with smart grids, people didn't care until we framed it around reducing power outages. It’s crucial to connect concepts to their significance. The downside of competing over labels is that it often leads to a lack of real meaning, leaving people unsure of what matters.

SP: That’s very true. In healthcare, unless you’re sick, you won’t take action. When I feel unwell, my behavior hasn’t changed: I might ask a friend for advice or try over-the-counter remedies until I finally contact my doctor or search on Google.

The challenge is how we communicate with patients using data about their behaviors and searches. Most people don’t think about healthcare until they have a problem; they just want to live their lives.

Jonathan Brown (JB): Until it affects you personally, it’s easy to stay within your bubble. In our work, we push clients to look beyond their own categories. We often become path-dependent in our expertise, but if we can find other categories where companies align with our values, the specific category becomes irrelevant.

For example, a client in South America asked about customer service trends in banks. I suggested they visit banks themselves but also look at air travel, hospitality, and retail, since those are the experiences people compare.

SD: That’s a classic product innovation methodology—looking laterally at other categories. Communication innovation can also benefit from innovative product development methods like design thinking or agile strategies. What we’re discussing is relevance, which hinges on data and precision.

Erin, from your perspective with Omnicom and big data, you might have insights on how we achieve this. One trend is the demand for relevant, tailored communication. How does the Omnicom ecosystem address this?

Erin Lanuti (EL): We have a massive dataset of consumer behavior and linguistic intelligence—what people say, do, and where they are. For four years, we’ve used this data, and with the advent of AI, we now create synthetic audiences and digital twins. This allows us to have deeper conversations, such as understanding the emotions of an Alzheimer's patient.

We can pre-test our messaging and creative work, validating with human insight. By leveraging real-life behaviors and social data, we can bring audiences to life in ways we never could before. This approach enables our teams to communicate with much greater precision.

MM: There’s a key distinction between where data will help and where it may mislead. In the political space, there was an assumption that American Blacks and Hispanics behaved a certain way, supported by data and research. However, Democrats faced a surprise when people actually voted differently.

While large datasets are valuable for understanding past behaviors, they are inherently backward-looking. They reflect existing products and conversations, but may not accurately predict responses to new ideas or changing environments.

It’s crucial to determine where data is truly useful and where it might give false positives, so we can apply it effectively without heading in the wrong direction.

EL: I have two thoughts on this. First, in politics, we can now analyze human behaviors—how people vote and their habits—and overlay that with what they say. This helps us identify disconnects and will greatly enhance our public affairs efforts.

Second, when you mentioned using Google or ChatGPT, it’s worth noting that one in two or three times, these engines can be wrong. This raises a real concern about the arbiter of truth for individuals with medical conditions who rely on these tools.

SP: China opened the world’s first AI hospital this year, staffed entirely by robots, with no human doctors or nurses. So you know, different countries have different rules. While many surgeons use robotics, they still bring experience and human insight. In contrast, AI robots are already being used in elderly care.

PH: AI also offers the potential to reduce bias in healthcare. Women’s complaints are often disregarded compared to men’s, and African Americans face different treatment in the system. If programmed correctly, AI could filter out these biases. Personally, I think an AI hospital sounds great; I’d prefer being treated by a robot over a biased human doctor.

Tim Hsu (TH): To add to that, I talk to a wide range of companies, many of which are trying to use AI. Some are risk-averse and avoid it due to legal concerns, while others recognize it will change workflows and want to start learning now. Our stance is that AI is a tool to amplify creativity, but it can’t replace human input.

Creative directors, for example, guide AI more effectively than I can. While I can generate a great video quickly, someone with a film or art background can elevate it further. Agencies that avoid AI are entitled to their choice, but AI is already here, and we must learn to work with it. The human element is essential for storytelling and artistic vision.

SD: Can I be provocative and suggest that maybe it’s not about merging humans and AI, but rather keeping them in distinct lanes? What you describe regarding the warmth and energy of human communication, is very different from AI which simplifies complex medical language for patients.

While AI can deliver useful messages, we still need humans for other types of communication. Just as people enjoy both stick shift and automatic cars for different reasons, we could have distinct communication styles—keeping AI and human interactions clearly defined as complementary rather than merged.

TH: To that I would say that in the creative field, AI is often used as a visual storyboard to generate ideas cheaply. Traditional photo shoots can cost tens of thousands, and if something goes wrong, you face significant setbacks. With AI, you can quickly generate multiple ideas, test them, and identify which ones actually resonate.

Often, we think we know what people want, but they may not even know themselves. AI helps prevent costly mistakes in campaigns, especially in sensitive areas like cancer treatment messaging. It allows us to save time and money while refining our approach.

EL: So we created something called AI optics. While AI is a tool, I challenge you to see it as a stakeholder and gatekeeper. It flattens the brand funnel, allowing for faster conversions, and controls your brand narrative and reputation.

This has major implications for communication innovation. We need to rethink our strategies for engaging with this new stakeholder, replacing "reporter" with "large language model." AI is influencing how consumers perceive healthcare, products, and services.

With AI analytics, we can see the priority order of what it considers most important about your offering. How do you convey your message if it differs from the AI's perspective?

SP: I have a different perspective on this. We use Gen AI to generate images and reports, particularly in advertising. In manufacturing, humans struggle to read over 10,000 pages of clinical trial information.

By combining AI with marketers' tools, we’ve achieved 97-99% efficiency. However, I agree that a human must make the final decision due to potential hallucinations in AI output. I see AI as a tool to save time, enhance efficiency, and provide options.

EL: But I would argue that most brands are misinformed. While efficiency is important, consider a world where search traffic declines due to zero-click searches. If traffic isn't reaching your pages, how will you engage and tell your story through large language models like ChatGPT or Google?

We’re seeing multi-billion dollar brands with only 5% visibility in AI engines. This translates to millions in lost revenue if you’re not part of the ad consideration.

PH: So I wonder if there’s a category error in how we’ve been trained to think. Many public relations campaigns aim to drive traffic to their websites, but from your perspective, the goal should be to drive people to the right answer, regardless of where that is—be it a third-party site, media outlet, or search results.

As long as people find the answer they need, it doesn't matter if they visit your site or another. This suggests that the traditional focus on media presence may not be as important as we think.

KE: Hard numbers suggest that around 61% of content comes from earned media, while 40% is from owned media, with social contributing very little. We're seeing a resurgence of interest from brands regretting their lack of long-term PR strategies, particularly regarding trustworthiness.

Many decisions, like weekend plans, are influenced by earned media from travel journalists and similar sources. We aim to invest in the publishing industry, focusing our campaigns on engaging readers rather than targeting everyone indiscriminately.


EL: This is a major existential issue. AI relies on earned media to convey trust, and we’re seeing a resurgence of traditional influence through platforms like YouTube and Reddit. However, traffic for news outlets has dropped by 40-50%, leading to declining revenue.

There’s a critical need to rethink revenue models. On the flip side, OpenAI has invested in Axios for local media because AI engines require clean earned media. This could lead to AI starting to find better sources.

JB: There’s an interesting nuance here. Currently, there’s an adversarial relationship between publishers and AI, with publishers feeling that AI pulls their content without attribution. If publishers continue to dwindle and produce less content, it undermines AI’s ability to access reliable references.

This lack of recognition of their interdependence could weaken AI's promise. Instead of viewing each other as adversaries, there’s a need for greater integration between these entities.

IO: This presents an opportunity to scale a new model of value creation. Instead of just reporting news, we could create a network of citizen journalists, offering diverse perspectives. For example, during a hurricane, we could gather 5,000 different views rather than just one.

By reimagining reporting, news sources could allow users to explore perspectives, like asking what a hurricane looks like from a child's or a senior's viewpoint. Tools that empower people to report news could transform content creation.

Instead of fighting AI, the focus should be on how to empower people to tell their stories better, creating new content and revenue streams. This approach recognizes that everyone views the world through their own lens.

EL: I agree and there’s also a strong emphasis on returning to human interaction and the importance of live, in-person meetings and training.

SP: It really depends on the age group too. Kids today interact very differently, especially between girls and boys. Most, not all, prefer online interaction, and many struggle with face-to-face communication. This has increased the demand for occupational and other therapists, as kids often don’t know how to interact in person.

For example, in research, younger generations in a room often struggle to stay quiet and tend to interrupt themselves. I wonder how this will impact their long-term social skills. While they may seem engaged playing video games with friends nearby, their interactions are changing. I’m not sure what the future holds for this.


SD: The generational differences in communication are a catalyst for trends. We have five generations in the workplace—Alpha, Z, Y, X, and Boomers. Boomers and Xs typically write formal emails with sign-offs and bullet points, while younger generations prefer Slack, WhatsApp, and Teams, using emojis and GIFs.

This is the first time in history we have such a diverse age range at work, leading to fractured communication. As we navigate this, we need to consider linguistic experts who understand generational, cultural, and ideological differences, as communication modalities are evolving.

MM: We don’t have extensive data on communication modalities, but generational differences are clear. Younger generations show significant antipathy toward corporations and capitalism. This crisis of trust is notable, and it's surprising we haven’t seen more AI-related scandals involving companies.

A new study reveals that a majority of Americans expect to be taken advantage of by companies. This skepticism affects consumer interactions, leading them to interpret everything negatively. If consumers expect deception, how can companies establish trust and fulfill promises? Especially in Europe, where there’s a quick response to negative experiences, businesses must communicate reliability and set appropriate expectations.


IO: The key is finding ways for everyone to win. By giving people opportunities to discover new things and earn rewards, you empower them. Tools like Canva simplify design, allowing users to create without starting from scratch. This connects to your point about returning control to everyday people, enabling them to generate value and content.

Brands are often protective, but they should embrace this shift. For example, young entrepreneurs on TikTok have generated millions by offering solutions people didn’t know they needed. AI allows us to tap into smaller markets that might not seem profitable but can yield significant returns.

Education becomes crucial in this new landscape, as understanding language and concepts is essential for effective communication. Brands need to educate their communities about the evolving world and reimagine their roles as resources and thought leaders, rather than just sellers. The focus should be on the value they provide beyond just selling products.

SP: It's interesting from an education perspective that many educators agree traditional schooling is changing. Kids don’t need to learn in the same way anymore. Brands will still exist, but the focus should be on storytelling that evokes human emotions, creating new value.

Earlier, Sarah mentioned hyper-personalization, which is supported by data. As I listened, I thought about clicks and data usage. Instead of a revenue model based on clicks, we might track data usage to understand engagement better. This could be a new approach to consider.


SD: A recurring theme in our conversation is influence—how we influence ideas, people, brands, and businesses. We're discussing the human versus AI dynamic, emphasizing that trust and relevance are crucial for effective influence. You need a clear point of view and something to say, especially today, as communication requires engagement from both senders and receivers.

As we wrap up, this discussion highlights the value of diverse perspectives in communication innovation. Loose networks, like the acquaintances we've made here, can enhance communication by bringing together eclectic groups that wouldn't typically intersect. This exchange is vital for business models and overall functionality. How do you see the value of these loose networks in fostering better communication?

EL: Brands must engage authentically in these smaller communities, which can be costly, but trust and influence are stronger here.

As we consider fragmented audiences, it's essential to build trust and relationships. While the downside is that people may live in bubbles, we still need to create bridges, as weak ties facilitate connections.