Unveiling the Future: Key Takeaways from GAI World – A Dive into the Rise of Generative AI

The lowest point on the Dunning-Kruger scale is where a person has acquired only a small amount of knowledge on a subject and consequently feels the least confident in their understanding of it. It was at precisely this point that I stepped into Generative AI World last week, the inaugural conference from analyst firm GAI Insights. I was there to represent FleishmanHillard as a principal sponsor [disclaimer], but also to learn about how this fast developing technology is affecting virtually every client we have, as well as the effect GenAI is going to potentially have on the entire communications industry. And I was in rich company, as the conference brought together a unique collection of the most senior leaders from world-class institutions such as Harvard University, Mayo Clinic, PwC, Ensemble Health Partners, Microsoft, Coffee Labs, Tomorrow.io, Jerry, Mass General and more, to share real-life lessons on project and technology selection, ROI calculation, team organization structure, data and IP approach and lessons learned.

Here’s what I took away from two days immersed in all things GenAI.

Common GenAI Terms

There is a lot of new language (mostly acronyms) associated with GenAI. Here are the terms I heard used most often and what they mean:

AI (Artificial Intelligence): Computers that can think and learn like humans.

GenAI (Generative Artificial Intelligence): AI that produces media (e.g., text, video, images, audio)

GAN (Generative Adversarial Network): A type of computer program where two parts compete to create realistic-looking things, like images or text.

LLM (Large Language Model): A smart computer that understands and writes human-like text.

NLP (Natural Language Processing): Teaching computers to understand and talk like people.

WGAN (Wasserstein GAN): A special type of program for making realistic images.

VAE (Variational Autoencoder): A program that learns how things work and can make new things that look like the ones it learned from.

RL (Reinforcement Learning): Teaching computers to make good decisions by trying different things and getting rewards.

SD (Synthetic Data): Refers to artificially generated data that mimics real-world data but is created by computer programs or algorithms rather than being collected from actual observations or measurements.

MU (Machine Unlearning): Teaching a computer to forget something it learned before, often used to remove biases or outdated information from AI systems.

Predicting the Impact of GenAI

So, what does this mean for the industry and business leaders of today who are approaching GenAI with a mixture of excitement and fear? Here’s what stood out:

Cutting Through the Noise: Everyone feels like they now must have an AI story.

Hitting the New LLM Gold Rush: There is currently a rush for companies and organizations to build their own LLMs, most with little understanding of the risks associated or their own ability to scale.

Democratizing GenAI Integration with Synthetic Data: Synthetic data may help level the playing field for some by providing researchers and developers without access to large data sets with the ability to create diverse and privacy-preserving training datasets in LLMs. It can also improve the model’s performance and mitigate concerns related to privacy and data scarcity, as it avoids using real, potentially sensitive or limited data directly.

Selling GenAI into the C-Suite: There is a level of accessibility with GenAI that did not exist with Web 2.0 and the Cloud, so selling it into the C-Suite is already proving easier. “The FOMO is very real with GenAI,” said one Health Data & AI advisor.

Predicting Industry Regulation: There is unlikely to be sweeping regulatory legislation in the U.S. for the foreseeable future. The EU is passing laws this year that do not go into effect until at least 2025 and the U.S. is significantly behind in its own prioritization of AI regulation.

Recognizing Security and Compliance Risks: What is the potential for generating misleading or harmful information and the risk of infringing on copyright or privacy when generating content based on existing data? Several high-profile companies have seen recent cases of confidential and proprietary information being leaked due to employee GenAI misuse.

Impacting Global Labor Markets: GenAI may lead to increased productivity but also job displacement, shifting labor markets towards AI-related roles, potentially exacerbating economic inequality, and impacting global competitiveness.

Overall, one thing came through loud and clear – GenAI should NOT be considered by industry, organization and business leaders as a plug-n-play addition to their tool stack, it must be set at the strategic level. While it offers the promise of automation and efficiency, its strategic integration allows leaders to align it with broader business objectives, such as innovation, stakeholder engagement and long-term growth.

GenAI and the Communications Industry

There is no question that GenAI has the potential to significantly impact the communications industry. It can streamline and automate tasks like media monitoring, data analysis and content creation. This could enhance the efficiency of PR professionals, allowing us to focus on more strategic aspects of our work, like building relationships and crafting compelling narratives. GenAI can also help identify trends and sentiment in real-time, enabling quicker responses to crises and opportunities. However, there are concerns about AI-generated content’s authenticity and ethical implications. It may be challenging to maintain transparency and trust when using AI for PR, and there’s the risk of misinformation or biased messaging. Striking a balance between harnessing AI’s potential and upholding ethical standards will be a key challenge for the industry as it adapts to this evolving technology.

I leave you with perhaps the most compelling quote from the conference. It came from Harvard Business School professor, Shikhar Ghosh, shedding light on “The View from the C-Suite and the Boardroom” regarding AI’s impact on businesses. He said “AI should be likened to termites, not tornadoes. Its influence will not be a sudden disruptive force like a tornado but rather a gradual transformation affecting business models like termites weakening the structure of a house.” While this might sound a bit apocalyptic, it should be noted that termites, like all good technology disruptors, are actually agents of growth and renewal.

This communication is offered as general background and insight as of the date of publication, but is not intended to be and should not be taken as legal advice. Each organization should confer with its own legal counsel and its own business and strategic advisors for guidance that is specific to and considers the organization’s status, structure, needs and strategies.