How Should Communications Leaders Approach Generative AI in the Workplace?
The approach to adopting generative AI and ChatGPT in the workplace has already given way to division amongst business leaders. Many leaders, including Elon Musk and Steve Wozniak, have called on AI labs to temporarily pause the training of powerful AI systems in order to allow for potential risks to be fully researched in the wake of GPT-4, a multimodal large language model even more powerful than its predecessor GPT-3. However, Andy Jassy’s annual shareholder letter was particularly open about Amazon’s investment in large language models (LLMs) and generative AI, with Jassy taking the stance that this technology shift is already present and active in the company’s work. Palantir CEO Alex Karp has adopted a similar stance to Jassy, stating recently that not only is this technology here and happening, but that we can’t slow it down, and further that “Every large organization in the world will soon require a system with these capabilities”. According to Bloomberg news, some agencies are already opting to use AI instead of third-party copywriters and designers. With leaders strongly voicing their opinions at each end of the spectrum, and adoption of this technology becoming a fast reality – which is the right approach?
How can leaders handle the inevitable transition to generative AI while the technology is still in the process of being developed? When is the right time to bring it into the workplace? And how will this transition impact the communications industry in particular?
Leaders must consider what’s next for ChatGPT and how we will shape the future with its use. Ruder Finn’s TechLab recently hosted a closed-door salon session in which senior communications leaders from a span of industries including financial services, CPG, supply chain management, and retail industries, got together to discuss the state of generative AI tools such as ChatGPT and their potential impact on the future. Although our discussion was fruitful and informative, what shone through was the hesitancy and conservatism towards adopting communications tools like ChatGPT into the workplace in some of the world’s largest companies.
It is, however, becoming more and more clear that the impact of generative AI will be significant. Goldman Sachs reported that up to 300 million jobs could impacted by automation worldwide, with administrative and legal staff the most at risk. In a recently published research paper from OpenAI and the University of Pennsylvania, it’s projected that 80% of the US workforce could have at least 10% of their work tasks affected by the introduction of LLMs such as ChatGPT. Up to 19% of workers may find at least 50% of their duties impacted by general-purpose technologies (GPTs). To make a long story short, generative AI is already making waves in the labor market.
In less than two years, generative AI in the form of text-based (ex. ChatGPT) and image-based (ex. Dall-E) tools, will likely be a mainstream technology. This technology has an incredible capacity for content creation, and much like in any other technology hype cycle, is being accepted early on by leaders for fear of being ‘left behind’. However, business leaders are hesitant to become early adopters for fear of relinquishing a necessary human touch on communications. Given that generative AI pulls from training data, the outcomes are dependent on the data being put in, which is subjected to limited quality control. On the other hand, there is also the opportunity for sensitive data to be accidentally leaked due to user error. For example, Samsung recently experienced confidentiality issues, just weeks after lifting an employee ban on using the tool. Another barrier to adoption was the lack of existing guidelines for using the tools for content creation in the communications space. Lack of regulation, coupled with the fear of job automation, has created barriers to entry for both employees and leaders to adopt the tool in the workplace.
Despite these barriers to adoption, the opportunities that generative AI present for the communications industry are very exciting. ChatGPT has the capability to create baseline communications that are almost indistinguishable from human generated content, which can then be supplemented with a human touch. While communicators may specialize in generating content for people, as more jobs become automated, tools like ChatGPT can emerge as better communicators to machines and generative systems, offering a different lens on how we communicate for the future.
The key to success in including ChatGPT in the communications industry will be dependent on three areas:
Collaboration: between the tech and human communicators. According to PRDaily, brands like OkCupid and Amazon have already found success through the collaboration method, from generating survey questions to expediting the product description process, respectively. Those utilizing ChatGPT in the workplace already stress the need to vet generated content and the use of the technology as a tool rather than a substitute for workers.
Better decisions: machines can summarize content and provide answers, but human will still make the decisions. The input to our decisions will rest on access to large data pools. It will be critical to make sure we have the largest datasets, that we avoid siloed data, and that we are aware of inadvertent biases baked into imperfect data. As the FT pointed out “algorithms and machine learning can help use make better-informed decisions, but better-informed decisions aren’t the same as “better decisions”. We will still have choices and judgments to make, and assessments that align with our culture and values.
Communicating across the generational divide: As communicators, it’s important to consider the potential impact of ChatGPT on workers, for whom the impact of automation on their jobs is top of mind. There are also vast differences across generations with younger cohorts keen to adopt new technologies while older employees are more challenged with some of the new tools. The communications function will play an essential role in encouraging "threatened" members of any organizations to embrace change, as opposed to resisting it. Internal communications and training are key to minimizing both the fear of redundancy and easing the transition to this next stage of tech. Leaders need to help employees embrace this change, rather than feel intimidated by it. It’s essential for leaders to show that they are listening as their employees’ express doubts and resistance to change. A non-empathetic stance will not benefit leaders whose employees are already concerned about change.
The sentiment among the business leaders who attended our salon was that of being “cautious, but curious” towards adopting generative AI. The leaders we talked to were particularly receptive to the idea of more regulation in the space. Helping your current workforce understand and embrace this next iteration of tech along with the guidelines that emerge along the way will ease this transition considerably, and help in the transition to what’s next.
