Lessons from digital media as AI takes hold

Because automation and AI is are nothing new to digital media professionals, we bring a different perspective to AI than people who work in fields where the technology is relatively young.

We’ve seen automation profoundly change how we work over the past decade, starting with the launch of automated bidding capabilities on programmatic and social platforms, known by Google as Smart Bidding.

Since then, AI has transformed every aspect of how we work at a blistering speed. But despite the disruption of our industry during this time, the technology has yet to replace human digital media experts. If there is one lesson that professionals should take from this as we move into 2025, it’s that those who skill up in AI are more likely to remain relevant and potentially increase their value. 

Here are some trends to look out for in AI in 2025.

1. AI will drive a step change in productivity in 2025 and beyond

Coming from the perspective of an industry in which AI disruption has been under way for 10 years, we have seen how rapidly it has transformed slow, laborious manual tasks into automated processes. Just a decade ago we needed to track every ad we ran across each platform and adjust bids manually to ensure ad spending aligned with revenue generation.

We’d figure out how much we’d need to pay for a click to get a decent return on ad spend for the conversions we achieved. But soon machine learning algorithms were able to perform this function exponentially faster and significantly more accurately than any human could. Automation of campaign targeting and bid structures made the human role in these functions obsolete.

When a campaign is set up with the right conversion metrics, and taps into the full capability of the machine learning algorithms supported with first-party data, you can be sure you are spending the campaign budget in the optimal way. The time we spend on optimising campaigns and creating tailored ads to target different users has been sliced from hours to minutes, allowing us to take on more clients and spend more of our time on strategic, value-adding work. 

2. It’s not what you know, but how you apply it

The big recent development in AI is the advent of generative AI platforms like Gemini and ChatGPT, which promise to profoundly change how knowledge workers do their jobs. These platforms transform AI output from data and information into knowledge and content. This will challenge what so many professionals — from marketing consultants and lawyers to coders and designers — do in the years to come.

If you want a medical or tax opinion, you can ask ChatGPT and interrogate its output. You’d still be wise to consult a doctor or an accountant (generative AI is known to get things wrong), but you will be well-informed when you talk to them. Likewise, it’s now possible to generate your own images for social media posts, the first draft of a blog post or a script for your website without being a designer, writer or coder.

We are seeing exponential growth in generative AI platforms and co-pilots that target a range of specialist use cases, from psychotherapy to search engine optimisation. Each knowledge worker and professional will need to think carefully about how to position themselves in a world where knowledge is a commodity and creation can be automated. Expertise and knowledge will need to be packaged and applied, rather than sold as is.

3. Human touch required in an AI-driven world

If the second point sounds scary, our experience in the digital media sector offers hope. AI has enabled us to be more efficient and effective,  but if anything, the demand for digital media experts has grown as the complexity of the AI technology has increased. Machines are only as good as the instructions and the data they are given. The importance of the human role in overseeing AI from a quality, governance and ethics perspective cannot be understated. Equally important are creativity, empathy, interpretation and judgment; these are unique human qualities that AI still falls short of in the “Turing” test — a measure of whether a machine can exhibit behaviour that is indistinguishable from that of a human.

4. Expect acceleration of the speed of change in 2025

2025 will be the year that companies across industries that range from manufacturing and agriculture to media, professional services and finance will need to scale up and accelerate their investments in AI, and generative AI in particular. Each business will need to think about how to get maximum value from the technology. Many task-based jobs will fall away, and the way businesses create and capture value will change.

I’m optimistic that AI will, much like the internet or the cloud, create huge opportunities for people and businesses. Think back to the way the mainstreaming of the internet has changed our lives and how we work. Putting powerful generative AI tools into everyone’s hands will unleash new waves of innovation and creativity in the years to come.

As was the case with the internet, businesses and people will show amazing resilience and adaptability in adjusting to the new possibilities and dangers of AI at scale. While there are understandable concerns about data privacy, data ethics and job security, AI will boost economies and open up job opportunities that never existed before.

Enjoy the ride

As AI continues to gain momentum, businesses, industries and professionals all face a leap into the unknown. The important part is to put the governance and ethical guardrails in place to ensure we benefit from the rise of the machines. But along with the caution there is the promise that 2025 will be the year AI truly starts to gain momentum and deliver on the potential it has shown over recent years. Buckle up — it’s going to be an exciting ride.

Grant Lapping is the digital executive at midnight, the innovation agency of future-focused management and digital growth enabler IQbusiness.

The big take-out: Expect the way you work to change, but remember that human expertise remains the key to unlocking value from AI.

Read more: