By Ian Liddicoat, Chief Technology Officer and Head of Data Science, Adludio
The world has moved on significantly from the 1950s – when Alan Turing gave us the Turing Test and John McCarthy first coined the term “artificial intelligence” (AI). Advancements over the past 70-odd years have led to AI-based technologies now forming a key part of several aspects of our everyday lives, but thanks to the chatbot craze, it is almost impossible to ignore.
These recent developments, by the likes of OpenAI’s ChatGPT and Google’s BARD, have left marketers excited at the prospect of generative AI in particular which, according to Gartner, will be behind 10% of all data created by 2025.
However, while the potential uses of generative AI are many, the technology remains fairly nascent. Moreover, the hype around it has arguably taken the shine away from the other, more developed, enterprise applications of AI-based technologies.
Clearing the hurdles
The reality is that this excitement is underpinned by a great deal of shortsightedness to AI-technologies, which stems from a lack of understanding.
For example, many marketers still fear the prospect of utilizing AI as a core part of their marketing strategy due to concerns over whether increased automation will lead to job losses. Others may view AI as being inherently “black box” and feel that, to fully understand the capabilities these technologies deliver, there needs to be expertise placed in-house or outsourced to a larger platform – both of which cause commercial concerns. There are also questions around potential biases within AI, highlighted by recent examples of ChatGPT’s output tending to lean in a certain direction politically.
These misperceptions are partly explained by a lack of skilled data scientists and engineers available with a technical understanding of AI. This is especially apparent at the board level, where the requisite expertise simply isn’t appearing on a routine basis.
On the one hand, there’s a lot more that can be done to allay the concerns of marketers and highlight the successful and easy application of AI technologies. Some open-source technologies have been helping here to remove some of the educational barriers. The increasing pace of regulation around AI from governments globally will also promote safe practices for AI application. However, there is still a degree of responsibility on technology vendors in explaining AI in a more user-friendly manner.
The many varieties of AI
It can be argued that, above all else, that the biggest misperception around AI, and the one which has hindered its uptake the most, is the view of AI as a catch-all term. However, just like any other piece of technology, it takes many forms and has many applications.
For instance, machine learning algorithms have long been used in digital advertising bidding platforms to cope with the volume of data being processed, and the speed at which this processing has to be done at. Meanwhile, computer vision, which derives information from visual items, is being utilized to enhance how campaigns are presented to consumers visually. Similar, deep learning algorithms are increasingly being leveraged for sophisticated creative optimization objectives.
As these techniques develop, they’ll become even more beneficial for marketers, to apply to various data and business outcomes and into their complex decision processes. This is particularly the case within creativity, where technological advancements mean there is huge scope for using AI to innovate. For example, as generative AI technologies become more sophisticated, through convergence with other AI-based techniques, its potential uses will be numerous. Combined with computer vision, it could help digital marketers create custom content in real-time, deterministically designed to maximize attention for each consumer audience.
Unlocking AI’s true creative potential
The combination of different AI technologies will give marketers the opportunity to truly increase the effectiveness of their campaigns and drive consumer attention. These technologies can be used to analyze various ad components, and bring all of that data into the bidding process, enabling marketers to optimize their ad creative and placement in near real-time. Being able to maximize engagement in close to real time is where the real value of AI for marketers lies.
Harnessing AI also helps marketers to future proof their strategies for the pace of digital privacy regulation, because the technology can ensure effective advertising in a privacy-conscious manner. The modeling focuses on aggregate behavior, rather than individual users, so doesn’t require any personally identifiable information.
There are plenty of real use cases for AI technology outside of the mainstream, and often gimmicky, surface-level applications of the technology. Just because something is the loudest in the market doesn’t mean it is the only method out there. The first step for marketers is to look beyond the hype and the myths of these embryonic technologies, and realize that other AI methods can unlock better online experiences for consumers and drive real business outcomes.
Once the fears are overcome, through education and a lessening of the skills shortage, then AI will have a significant role to play in the digital marketing process for marketers, and its true potential can be achieved.
