Want to have a job in an AI world? Start preparing now
The Drum’s founder reports back from WPP’s Stream conference in Athens, where AI’s impact on marketing and media jobs was a hot topic.
/ Abobe Stock
‘In conclusion’ seems the preferred way for ChatGPT to end its articles. To me, it’s a tell that something has been produced using AI, but I’ve been seeing it a lot more often at the end of reports, blogs and articles that cross my computer screen.
It is a sign that the great AI tsunami is already inundating adland. It has already inundated the conversations at every marketing summit.
It has also been said that it is a prime concern of UK prime minister Rishi Sunak, who is planning a summit of his own to work out how best to regulate AI. Leading executives from OpenAI, Google DeepMind and Anthropic are expected to join the two-day conference next week.
The gathering is being held in Bletchley Park, home of Allied code-breaking during WWII, and will aim to ensure a system where AI is used for the good – as opposed to the detriment – of humanity. Delegates will be asked to endorse a statement in advance. It is still being drafted but may include a warning that AI risks causing ‘catastrophic harm’ if left unchecked.
I have just returned from WPP Stream, an ‘unconference’ where most agreed with that warning. There was another anxious subtext to all of the talks of improved productivity, efficiency and capability, however – the impact AI may have on marketing and media jobs.
We heard estimates of how creative production costs could be reduced by 85%, how commercials could be made using a fraction of crew and cast, as well as how much of media buying could be automated.
I could go on, but suffice to say there is no high ground to protect against the scale of this wave. The impact has already been profound despite the fact we are nowhere near high tide.
The Bletchley Park meeting is likely to focus on the development of artificial intelligence that surpasses human intelligence. That prospect could lead to potentially catastrophic consequences, as these advanced AI systems may not be aligned with human values.
But AI, even in the current large language model form, is already highly disruptive. As day turned to night at the WPP Stream event in Athens, some of the conversations also turned dark. Many of the roles people currently have will soon be replaced with AI.
The speed of change is perhaps its biggest short-term threat. Any ecosystem faces collapse when the environment changes faster than its ability to evolve. And I doubt the economic issues will top the agenda at Bletchley, which will first want to mitigate against any threat that the tech poses to humanity itself.
So how can people and companies prepare themselves for this new environment? First, they need to lean into AI, embrace it at every opportunity and get to know the tools that will have the most impact. Take every chance to experiment, play with and use them. Understanding AI is key to ensuring it is as much an opportunity as a threat.
But large language models also have a serious flaw – they need original human content to avoid model collapse. In other words, when AI feeds off of other AI, mistakes are replicated, applied and repeated until the system starts talking gibberish.
So, these systems do need real people to produce original content to work – at the moment, at least. Research has found that AI based on a small data set made up of authentic human content is more reliable than systems that use far larger data sets of synthetic content.
While the scale of disruption we face is without parallel, so are the opportunities. Applications in medicine, finance and manufacturing will be game-changing, but so too will advances in how marketing will work.
85% productivity savings is also another way of saying we can achieve 85% more. At Stream, I met many new companies that were already exploiting the tech. They underline there will be jobs in this new world for those who start preparing now.