The Growing Complexity Of Patient Feedback
Healthcare now operates across more digital touchpoints than ever, and a single hospital can receive thousands of patient comments each week, each filled with emotion, nuance and context. When staff manually review this volume of feedback, two challenges quickly surface.

The first is scale: Human teams can't keep up. The second is subjectivity: The same comment can be interpreted differently by different readers. As a result, early signals of dissatisfaction, confusion about instructions, long wait times or unclear communication often slip through unnoticed. Operational issues that could be resolved early instead linger until they begin affecting ratings and outcomes.

AI is becoming the bridge between overwhelming feedback volume and meaningful insight, helping healthcare organizations understand patient sentiment with far greater depth and speed.

How AI Is Changing Patient Feedback Interpretation
The rise of AI in healthcare is turning feedback analysis from a manual, time-consuming process into a more precise, data-driven one. Advances in natural language processing and machine learning now help organizations understand not just what patients say, but how they feel and why.

NLP And Sentiment Understanding
Natural language processing (NLP) enables AI to interpret tone, emotion and context. Instead of labeling a comment as simply "positive" or "negative," NLP can detect multiple sentiments within a single statement.

Generative AI tools crept closer to the uncanny valley this year as tools like OpenAI’s Sora or Google’s Nano Banana, have flooded the market. Marketers have higher expectations than ever for AI to streamline marketing processes and save dollars. The novelty of generative AI has worn off and marketers want controllable, consistent tools that can be easily integrated into the creative ecosystem.

“If you’re using AI to automate a certain task, but then you have to move files from 15 different softwares, you’re not actually making your life more efficient,” said Freddy Dabaghi, chief transformation officer at Crispin.

This year has brought winners, losers and middleground contenders based on their integration capabilities, user interface and of course, final output. To get a better sense of how these tools performed this year, Digiday pulled together a 2025 agency generative AI report card ranking based on output, user experience and capabilities. This list is by no means exhaustive as generative AI tools continue to flood the market.

Digiday reached out to Google, OpenAI and Midjourney for comment. Google and OpenAI defended the creative potential of their respective generative AI tools, while Midjourney did not respond by press time.

Nano Banana: A
As with most things Google touches, its generative AI tools have become the gold standard, according to marketers. Google brought Nano Banana Pro, its image generation and editing tool built on its Gemini LLM, to market in November.

At least two agency execs say Nano Banana Pro is the go-to starting point for creative workflows, who largely complimented the image generation tool for its higher precision than competitors, like Midjourney, Canva or Adobe Firefly. There’s less of what marketers call AI sheen — which is described as inconsistencies and a “too perfect” quality that people notice — and there’s consistency in what the image generation tool produces. Images look hyper-realistic, which makes it more difficult for users to tell the difference between what’s real and what’s AI generated.

86% of U.S. buyers use or plan to use generative AI in video ads, showing why brands must adopt AI workflows to stay competitive on speed and cost.
AI-powered YouTube campaigns deliver 17% higher ROAS than manual ones, proving that automation outperforms when paired with strong creative strategy.
Mondelez is slashing production costs by up to 50% using AI with Publicis and Accenture, showing how brands can reinvest savings into sharper storytelling and bigger ideas.
Generative AI has made video faster and cheaper to produce. That part is obvious.

What follows is less straightforward; feeds are filling up with automated content that looks complete but feels interchangeable.

Now, creative judgment carries more weight than output.

Taste, intent, and restraint shape how the work lands.

In an interview with DesignRush, Chris Marcus, Chief Creative Officer at Colormatics, explains how brands can use AI while protecting creative judgment and human presence.

A new bill making its way through the state House Committee on Communications and Technology aims to address how artificial intelligence is being used in health care. House Bill 1925 would set standards for AI use in the doctor’s office and require safeguards for decision-making.

The Pa. House Committee on Communications and Technology held a session for H.B. 1925 through numerous panels from health care professionals, insurance agencies, technology representatives and those who are impacted by AI in the doctor's office.

Last week, President Donald Trump signed an executive order that bars states from enacting their own AI regulations and leaves oversight to the federal government.

"I think it's very likely that executive order is unconstitutional," said Rep. Venkat. "Health care, in particular, is well-regulated at the state level. So, I'm very confident that our effort is constitutional and is the right way to go for Pennsylvania."

Agentic AI differs from generative AI in ways that makes it particularly well-suited for marketing. Unlike generative AI models, which respond to isolated prompts, agents are software systems powered by AI that can act on marketers’ behalf to plan and complete multi-step workflows.

Marketing leaders are already beginning to tap agentic AI’s capabilities to streamline and improve process-heavy marketing campaigns and free marketers to innovate. Working together—and with human counterparts—agents can compress multi-week workflows into hours, enhance the accuracy of decision-making, and enable new levels of scalability, all while helping reduce errors and ensure adherence to brand.

“Perhaps most importantly, agentic AI gives marketers the gift of time,” says Nanivadekar. “By automating repetitive tasks and streamlining workflows, it frees marketers to focus on strategic thinking, creative direction, and human connection—all elements where human expertise remains irreplaceable.”

As Real Social’s Westley puts it, “By helping me do the more administrative tasks of my job better, AI can let me really focus on identifying influencers and building relationships—the parts of my job that differentiate the work most and I truly enjoy.”

Vibe coding is increasingly part of how software engineers work with LLMs. Right now, “good” vibe coding is mostly judged by speed. Who gets to a working result fastest?

As LLM usage scales inside companies, a familiar evaluation pattern will likely emerge. Not whether something works, but how its cost grows. The same question we ask of algorithms will start being asked of vibe coding workflows, even if no one calls it Big-O at first.

The push toward something like “vibe complexity” comes from a very practical tradeoff. Teams that do not use vibe coding at all move slower and leave obvious productivity gains unused. Teams that rely on it too heavily, especially through one-shot prompting and repeated retries, accumulate large token bills with little visibility into why. In practice, most teams oscillate between these extremes, and cost only becomes visible after it has already grown.

Vibe coding makes engineers faster, but iteration is not free. Each prompt and correction consumes tokens, and tokens map directly to cost. A workflow that relies on many fast iterations may look productive, but its effective complexity can be closer to O(n²) than O(n) once you factor in retries, clarifications, and prompt sprawl.

Two engineers can reach the same outcome at similar speed. One does it with many loosely scoped prompts, effectively linear in iterations but quadratic in correction. The other reaches the same result with a small number of well-structured prompts that encode constraints up front, closer to constant or linear token growth.

2025 has been a strong and eventful year of progress in LLMs. The following is a list of personally notable and mildly surprising "paradigm changes" - things that altered the landscape and stood out to me conceptually.

1. Reinforcement Learning from Verifiable Rewards (RLVR)
2. Ghosts vs. Animals / Jagged Intelligence
2025 is where I (and I think the rest of the industry also) first started to internalize the "shape" of LLM intelligence in a more intuitive sense. We're not "evolving/growing animals", we are "summoning ghosts".
3. Cursor / new layer of LLM apps
What I find most notable about Cursor (other than its meteoric rise this year) is that it convincingly revealed a new layer of an "LLM app" - people started to talk about "Cursor for X". As I highlighted in my Y Combinator talk this year (transcript and video), LLM apps like Cursor bundle and orchestrate LLM calls for specific verticals:

They do the "context engineering"
They orchestrate multiple LLM calls under the hood strung into increasingly more complex DAGs, carefully balancing performance and cost tradeoffs.
They provide an application-specific GUI for the human in the loop
They offer an "autonomy slider"
4. Claude Code / AI that lives on your computer
5. Vibe coding
6. Nano banana / LLM GUI

Key Points
A study found that advertisements created entirely by AI achieved a 19 percent higher click-through rate, while AI-modified versions of existing ads showed no improvement or even performed worse.
The findings suggest that AI image generation is most effective when producing completely new advertising content from scratch, as modifications to existing visuals actually decreased purchase intent.
Adding transparency labels such as "AI-generated" to advertisements reduced click-through rates by approximately 31.5 percent, regardless of what the images actually showed.

Effective communication is the backbone of high-quality health care. Whether between clinicians and patients or among interdisciplinary teams, the ability to exchange information clearly, empathetically and efficiently can determine the success of diagnosis, treatment and patient outcomes. In recent years, medical improvisation — an adaptation of improvisational theater techniques for clinical settings — has emerged as a powerful tool to strengthen communication skills among healthcare professionals. By focusing on presence, adaptability, active listening and collaborative problem-solving, medical improvisation helps clinicians navigate complex, unpredictable interactions with greater confidence and compassion.

Medical improvisation is grounded in the core principles of theatrical improv: “Yes, and …,” active listening, perspective-taking and a willingness to embrace uncertainty. These concepts translate naturally into healthcare, where professionals frequently encounter ambiguous information, emotionally charged situations and the need to make rapid decisions with limited data.

Unlike scripted simulations that focus on specific clinical tasks, medical improv workshops emphasize relational skills, such as how to pay attention, how to respond thoughtfully and how to build on what others say. These abilities are not only essential for patient communication but also for fostering teamwork and trust among colleagues.

Freelance marketing specialist and copywriter Maria Breaux doesn't buy the hype around AI-generated content. After more than a decade working across education, pharma, and tech—writing for brands like Pinterest, Marin Software, and Common Sense Media—she's watched too many marketers mistake efficiency for effectiveness.

Her approach? Compassionate marketing. "You're writing to a human being who may be suffering from something," she says. It's a framework that treats audiences as humans first, metrics second. And it's delivered results: 125% of fundraising goals at Common Sense Media and campaigns that consistently beat industry benchmarks.
...
For Maria, the issue is ethics. In regulated spaces like pharma or healthcare where she's spent years writing copy, authenticity isn't optional.

"People can tell right away when it's insincere," she explains. "It's just a few words that are off."

She uses ChatGPT strategically—as an advanced thesaurus for word alternatives and tone parameters. She uses Grammarly for spell-checking. But she draws a hard line: AI is a writing assistant, not a writer.

For many websites, 2025 has been a rude awakening to the AI era. Retailers, news publications, and marketing agencies saw drops in traffic of 20–40 percent, with much of that decline coming from a loss of organic search traffic. The likely culprit: the new AI-generated summaries appearing at the top of search results, providing many users with the answer they sought without any additional clicks.

But Kelly Cutler, an associate professor at the Medill School of Journalism and lecturer in the Kellogg Executive Education program, says that the search reset is merely the latest seismic shock in an always-changing internet ecosystem.

“This industry has seen these shifts many, many times throughout the last 20–25 years,” Cutler says. “So this is not something where I would be panicking or assuming that SEO is dead or search engines are over. The sky is not falling. It’s okay, but marketers need to evolve.”

As such, businesses and publishers will need to reexamine their strategies and adapt to the ways AI is changing search, much as they did for the rise of Google or the growth in mobile browsing.

From focusing on new metrics and SEO strategies to pushing hard into personalization and alternative sources of traffic, it’s time to experiment, Cutler says.

“Engagement is changing,” she says. “I know that’s not what anyone wants to hear, but it’s a fact. I do think in some ways this brings us back to where we should be, which is that it’s all about the user.”

Here are three tips for navigating the new search environment.

Track more-nuanced metrics
Beneath the alarming dip in search traffic, some sites have reported a silver lining: while fewer people are visiting, those who end up on a site are more engaged. For retailers, that means higher conversions to sales; for content publishers, that means more articles read or videos watched.

If AI summaries are detouring visitors by providing the “easy answers” that previously drove a lot of casual web-browsing, the remaining traffic is at least made up of more-motivated customers, Cutler says.

For years, websites relied on clicks from Google to bring in readers, buyers, and leads. That basic pattern is now shifting. More search pages show AI-generated summaries and instant answers that keep people from leaving the results page. When users get what they need at the top of the screen, they often stop there. Studies from Pew Research Centre, Bain & Company, and several industry trackers show that this has become common in many search types, from simple questions to early research steps before a purchase.

The change is creating pressure for marketers and publishers to think differently about how people find information. The click is no longer a sure thing. And when the click disappears, the usual way of measuring performance starts to change as well.

People click less when AI summaries appear
Pew Research Centre found that users are less likely to click links when an AI-generated answer appears in their results. Its study looked at how users respond to Google’s AI Overviews and found the pattern of when the summary appears, fewer people visit the websites listed below it.

This lines up with broader data from Bain & Company, which says around 80% of consumers now rely on zero-click results for at least 40% of their searches. It also estimates that many sites have seen a drop of 15% to 25% in organic traffic due to these behaviours.

What used to be a straight line – search → click → website – has become uneven. AI summaries sit above the links, take up space, and often deliver the information that users want. That means the position of a search result matters less than it once did. Even a top result can see lower traffic if the answer box covers the main points before the link appears on screen.

A new AI model can estimate how old your blood vessels really are using data from a simple wearable sensor, revealing an age gap that may offer critical insight into your future cardiovascular risk.

In a study published in Communications Medicine, researchers introduced AI-PPG age, a deep learning-based biomarker derived from photoplethysmography (PPG) signals - those optical pulses captured by common wearable devices like smartwatches.

Generative AI tools are pulling heavily from community-driven platforms, prompting a shift in how companies think about search, social, PR, content, and product messaging.

Reddit, Wikipedia, YouTube, Facebook, and other open forums make up the majority of sources LLMs cite, per Semrush; Reddit alone appears in 40.1% of all genAI citations. That means AI relies far more on community conversation than on brand-owned pages.

The outcome is a new cross-channel reality: AI frequently amplifies community consensus over brand-produced material.

Because AI assistants integrate conversations, reviews, and public commentary directly into responses, teams that once operated independently—SEO, social, communications, product marketing—now influence visibility in shared ways. AI does not separate on-site content, social chatter, or earned media; it blends them into a single response that users often treat as authoritative.

Why it matters: Traditional operational silos break down when AI tools merge information across platforms. Search teams can no longer optimize solely for on-site material; Social teams can no longer focus only on engagement; Communications teams cannot assume press hits live in a separate lane. What appears inside AI responses often depends on how all of these elements show up in public channels simultaneously.

This creates a shared responsibility for visibility. If social conversations lack depth, PR coverage is thin, product pages fail to clarify use cases, or if community discussions overlook a brand entirely, AI tools will reflect those gaps.

AI can produce more ads, but brands still need guardrails
If the latest flare-up over AI slop made anything clear, it is that brands cannot treat volume as a proxy for quality. Coca-Cola’s Christmas AI campaign — now an annual Rorschach test for AI-generated advertising — offers the latest case study. Viewers quickly flagged the telltale glitches: shape shifting trucks, continuity errors and uncanny animals.

Executives, for their part, recognize the limits of a laissez-faire approach. A recent study from Bynder of 1,800 business leaders found that nine in 10 believe human oversight remains essential when using AI to protect brand identity and maintain compliance. Understandably, those concerns aggregated around governance (see chart). Speed and cost savings may be winning over CFOs, but brands still need real guardrails if they want AI-generated work to resonate rather than appeal.

“Marketers need to rethink how they use gen AI in advertising,” said Jim McGorty, creative director at brand experience agency onepointfive. “Missteps don’t just weaken creative — they expose cultural blind spots and erode brand trust. And in a world where audiences already assume their feeds are distorted, that skepticism is only accelerating.”

Question: What are some simple AI tools marketers can build today without coding — and how can they make daily work easier?

You don’t need to be a developer to start building useful AI tools for your marketing stack. With the right platforms, marketers can automate repetitive tasks, create better content, and make smarter decisions — all without writing a line of code. Here are a few practical tools you can build today:

Chatbots for lead gen and support
Build with: Chatfuel, ManyChat
Why it matters: These no-code platforms let you create chatbots that answer FAQs, qualify leads, and support customers in real-time — on your site or social channels.

Email subject line generators
Build with: ChatGPT, Jasper
Why it matters: Use AI to brainstorm and test high-performing subject lines that boost open rates and cut down on creative fatigue.

Social media post creators
Build with: Canva Magic Write, Lately
Why it matters: These tools generate and schedule social content using AI suggestions, helping you maintain consistency and speed up campaign workflows.

Personalized content engines
Build with: Persado, Crayon
Why it matters: Serve relevant content to site visitors based on behavior and intent — no developer required.

Auto-generated dashboards
Build with: Google Looker Studio, Tableau (with templates)
Why it matters: Connect your data sources and let AI build visual reports you can share with stakeholders, without wrangling spreadsheets.

AI-powered design tools
Build with: Canva, Adobe Express
Why it matters: Create professional-looking visuals and ad creatives fast, even if you’re not a designer.

Customer feedback analyzers
Build with: MonkeyLearn, Thematic
Why it matters: Drop in your survey or review data and let AI surface themes, sentiment, and opportunities for improvement.

These lightweight tools don’t just automate tasks — they also free up your team to focus on strategy, creativity, and growth. With low lift and high ROI, they’re a smart starting point for marketers experimenting with AI.

The new tools, which the company announced on Wednesday, include Ads Advisor, which creates personalized suggestions for Performance Max campaigns, and Analytics Advisor, which is built into Google Analytics. Both agents are expected to roll out to English-language accounts in coming weeks.

“The lesson we learned the hard way is that these underlying language models underneath many applications, even the ones we build ourselves, are not always up to date,” said A. Lee Judge, cofounder and CMO of Content Monsta during “Zero to Launch: Using AI for Campaign and Content Creation at Scale,” a session at the MarTech Conference earlier this month.

Here’s how marketers are weaving automation and LLMs into their creative and operational processes.

In addition to language translation, AI plays an important role in interpreting labs and imaging results into plain language for patients to understand, says Tom Gillette, CIO of Mount Sinai Medical Center in Miami Beach.

LinkedIn’s chief marketing and strategy officer, Jessica Jensen, has built her leadership philosophy around one core principle — curiosity. She likens it to a kaleidoscope that shifts patterns with every turn, arguing that in today’s volatile business environment, leaders must do the same: adapt, question, and reimagine what success looks like. “Leadership isn’t about having all the answers,” she tells Campaign India. “It’s about reinventing success, remaining resilient as patterns change, embracing humour, and making work both meaningful and enjoyable.”

Smartly's 2026 Digital Trends Report shows 92% of marketers agree AI transforms customer engagement while precision-first marketers waste 27% less budget.

Artificial intelligence technologies are improving early diagnosis accuracy in various diseases, enabling faster treatment and better patient outcomes.

Innovative AI applications are helping physical therapists personalize treatment plans and track patient progress more effectively.

Mental health services are integrating AI chatbots that provide timely emotional support and resources to users in need.

Robotic systems powered by AI are enabling more precise and adaptive physical rehabilitation, improving patient mobility and recovery times.

Experts discuss the ethical considerations of expanding AI use in healthcare, emphasizing patient privacy, data security, and bias prevention.