Enhancing productivity through generative AI

Business travelers know the situation well. You have an upcoming meeting in a city you have never visited, and need to book travel. In today's paradigm, you would leverage one or more travel websites and search engines to book transportation, reservations at reputable restaurants and hotels, and then enter times, locations, and activity details in your calendar.

Thankfully, now an AI travel assistant (ATA) can help you accomplish those tasks—including entering events in your calendar. Even better, your ATA can interpret your natural spoken language to create your itinerary.

Example: "ATA, build a trip itinerary for me arriving on Tuesday in Jacksonville, Florida, and leaving on Thursday night. We'll need dinner reservations for three on Tuesday evening and eight on Wednesday evening. Tuesday can be casual, and Wednesday needs to be a professional setting. Also, provide a few choices for hotels that have a 24-hour fitness center and dry-cleaning service on-site."

Your ATA will source online content, including consumer reviews, as well as your personal travel history, to complete the bookings for your three-day itinerary and provide recommendations for additional activities during your trip.

The technology that will soon be enabling these types of comprehensive customer experiences and natural language interactions is a branch of the artificial intelligence family tree called generative AI. It’s the same technology currently revolutionizing the delivery of products and services for many businesses and driving technologies like ChatGPT, Bing search and Google’s Bard. 

What is generative AI?

Many organizations are working to understand generative AI and how to use it. Ben Bilsland, a partner and senior analyst focused on media and technology at RSM UK, described the technology in a recent article:

“Generative AI—broadly—is the use of AI to create new content. However, it is just one form of artificial intelligence that sits alongside a range of other fields, including fuzzy logic, predictive AI, deep learning, machine learning and robotics. The nature of AI means that some of these fields overlap. And whilst AI is typically believed to be a product of scientists starting in the 1950s, we are still at the very starting stages of its scope and potential."

Expanding on Bilsland's description, the emergence of generative AI can be traced to the early 2010s, when researchers began exploring the use of deep neural networks to generate new content.

One of the earliest examples of generative AI is Google's 2015 DeepDream project, in which software used such networks to generate images from existing photos. The resulting images were dreamlike and often included surreal and unexpected features. Amper is another company that tapped generative AI early on, launching software in 2017 that creates custom music tracks for users based on their preferences and desired mood.

Perhaps the most famous example of generative AI is ChatGPT. Developed by OpenAI, ChatGPT leverages an LLM (currently GPT-4) to generate humanlike responses on a wide range of topics. The ATA scenario above is just one example of how this technology will enable new and advanced capabilities in every industry.

The buzz behind generative AI

The team at RSM US LLP's Acceleration Center for Innovation Research Lab believes that generative AI uniquely differs from other technology cycles. A close look at the cycles the team has documented since the year 2000 holds hidden truths:

1. Infrastructure requirements compound edge use cases.
2. Innovations are intertwined and build upon each other.
3. Technology applications are initially dismissed by the majority of consumers and embraced by only a small cohort of technical experts.
4. Startups leading the initial burst of innovation are surpassed by later-stage startups specializing in the same field.

1. OpenAI's technology stack is the edge case of advanced infrastructure innovations, and its application programming interface is the infrastructure requirement that compounds use cases.
2. GPT-3 and GPT-4 are the combinations of not just one, two, or three domains of innovation but instead consume them all as training parameters.
3. ChatGPT scaled from no global monthly active users in November 2022 to 100 million by January 2023. In reaching that number, it outpaced TikTok by seven months, Instagram by 28, Pinterest by 39, Spotify by 53, Telegram by 59, Uber by 68 and Google Translate by 76.
4. OpenAI has received a total of just over $4 billion in funding across six rounds, the largest on Jan. 13, 2023, by Microsoft for $2 billion. Microsoft has since integrated OpenAI's GPT-4 into its Bing search browser.

The ACI Research Lab leverages a scoring model to determine the phase of maturity of new or emerging technology. The team assessed generative AI in November 2022 and again in February 2023. The change in only three months was significant, with the technology advancing from a purely experimental and research phase to the moving of new products and services into the marketplace. Highlights of the team’s findings included:

• $2.67 billion of funding for Generative AI in 2022 (up 67% from 2021) across 115 deals. 
  • More than two-thirds of generative AI companies have not yet raised a Series A round.
  • Generative AI text captured $825 million of 2021−22 deal flow, with generative AI visual media capturing $822 million.
  • Since 2021, Insight Partners has backed seven generative AI companies, beating Sequoia and Kleiner Perkins, which each backed four.
• A 198% increase in the number of references to generative AI in scholarly publications year over year
• Internal interviews on the ways RSM plans to operationalize the utilization of generative AI

Along with earning the dramatic increase in ACI’s maturity score, generative AI has expanded rapidly in the venture capital and startup landscape. As of March 15, $3.15 billion had already been raised across 21 companies in 2023. The deal sizes are increasing, and so are the use cases.

As generative AI matures, funding is expected to not only increase but continue to support the development of highly specialized modules that serve niche business areas. Generative AI will spearhead the newest iteration of workplace efficiency, prompting the next question: How can companies capitalize on this trend?

Where to begin

Generative AI is so exciting because of the opportunities it will bring across all industries and areas of business. Asked which industries should respond to the recent advances in generative AI, George Casey, advanced analytics leader at RSM US, responded: “Everyone. Everybody is going to find value and opportunities using this technology.” He added that this technology is particularly relevant in the health care, industrials and consumer products industries.

Generative AI is rapidly evolving and will continue to do so as models grow and become more sophisticated. With more advanced models and tools built upon them announced on a weekly basis, middle market business leaders need to think about how—not if—AI will disrupt their industries and how they will utilize the technology, both internally and in their products and services, to stay competitive.

In addition to taking the steps recommended above, business leaders should evaluate the risk and compliance factors involved in adopting AI tools and processes. Companies should begin updating or drafting new policies to account for the new risks introduced by AI, balancing experimentation and advancement with cybersecurity and compliance with legal standards.

RSM provides a range of services to guide your strategy during this critical period of transformation, as well as insights from our blogs:

• Data analytics blog
• Cybersecurity War Room blog