Opinion: Generative AI in 2024

Disclaimer: The following article is an opinion piece. This article represents the personal views and perspectives of the author and does not necessarily reflect the opinions or positions of Fonterra. Readers are encouraged to verify any claim or statement made below for themselves and Fonterra is not responsible for any decisions taken in reliance on the same.

Emergence and Mainstream Discovery of Generative AI

The world changed in November 2022 with the release of ChatGPT to the public and in turn unleashing the potential of Artificial Intelligence (AI) to the general public. While AI has been around as a branch of work since the 1950’s it has been largely invisible to our daily personal or professional lives. However, the potential to ask questions in simple written or spoken language and receive detailed well-crafted responses, generate pictures or short videos on a plethora of topics is truly a revolution in how we interact with technology and data in our work and personal lives.

Globally the technology industry is investing billions of dollars in research and capacity, people and data centres, to advance and support the explosion of AI driven potential. This is spurring a focus on all aspects of AI from traditional machine learning, which Fonterra for example uses to identify potential defects in NZMP milk powder bags, to modern Generative AI, the brains behind ChatGPT and Fonterra’s internal version for staff, in supporting a transformation across the way we work; with this comes incredible opportunities.

2023 was the year of discovery for the majority of users and corporations, with early deployments of Generative AI in many companies, focussed on experimentation, building knowledge and understanding the risks, benefits and how to best enable the workforce with this technology. 2024 has seen a shift away from proof of concepts and experimentations to a strong focus on scale and value realisation.

To empower our staff with the benefits of Generative AI, we have taken a multi-faceted approach leveraging industry leading “bought” options such as Microsoft 365 Copilot, targeted at our corporate functions; specific GenAI powered solutions, for our marketing and sales staff; and, we have built our own safe, internal Co-op GPT for all our corporate workforce. Fonterra has seen strong adoption of these tools amongst teams and with them a growing awareness of the potential and productivity gains, however our real opportunity is to leverage our information and data, in a considered and safe way to truly realise value.

Generative AI for Media and Marketing

Marketing, media, and advertising, in particular, are areas of massive opportunity for AI, with mature commercial offerings accelerating copy and content generation across text, image and increasingly audio and video generation, to Chatbots and insights across social and engagement channels.

Across 2023 and early 2024 we have seen the emergence of commercial AI solutions to support specific tasks within the marketing and media domains, such as content / print generation or image creation, with integrated workflows and publishing. An example opportunity here, is leveraging Generative AI to accelerate the marketing process, producing a long form video, before using AI enabled tooling, alongside company branding and simple prompts to repackage the video for target markets (e.g., Indonesia, Vietnam, and Laos) and platforms (Facebook, YouTube, TikTok) matching the appropriate scenes, maintaining branding and overlaying messages in the target language automatically.

The underlying technology and models are evolving rapidly, but a high skillset barrier remains for many organisations looking to build end-to-end solutions across a process; mainstream platform organisations such as Adobe and Salesforce are rapidly developing strong commercial offerings incorporating the latest technology into their platforms, simplifying adoption with familiar user experiences. I expect late 2024 will see that continued acceleration of progress across all types of media, with a much richer and more integrated experience, smarter models with a better ability to reason and perform complex tasks, further enabling the opportunity from bringing together insights from data, e.g., CRM (Customer Relationship Management) and social information. These improvements, along with the ability within applications to recommend and automate actions with AI, creates an amazing potential to automate, personalise and publish rich multi-media content / campaigns. An example could be the automated segmentation and profiling of customer groups, mining key interest and sentiment information for input to personalised campaigns, with rapid prototyping using GenAI for text, image and video options.

Considerations around Generative AI

Generative AI while creating substantial benefits, has a number of key considerations around risk, use and governance. For a majority of users, the focus is around cautious curiosity; a recognition that public services may be using your data for training and the answers require some careful thought before acting on them.

Unfortunately, there are also some far more challenging and less recognisable risks around Generative AI models, in particular the amplification of bias, inclusion of copyrighted and personal information in training data and hallucination, where answers may seem correct but are entirely made up. The commercial use of these products requires a strong consideration of risk and often a “human in the loop” to review the outputs and quality check content before publication.

For businesses, another key concern with the emergence of Generative AI has been the impact on people and roles. An often-quoted phrase is “AI will not replace you, but the person using AI will” which really highlights some of the challenges and opportunities we see in industries transformed by, particularly, Generative AI. It’s important that we recognise the risks, educate ourselves and our workforce around the technology. We need to be deliberate and collectively find the right way to incorporate it into our practices and improvements.

Increasingly we are seeing regulation and guidance come into place, that provides an improved set of expectations for companies around the use of information with AI and the use of AI as whole. Such regulation and guidance should provide increasing certainty as we embark on this evolution of value creation through AI.