Inside PricewaterhouseCoopers’s $1 billion generative AI upskilling campaign

What does it mean to invest $1 billion in generative AI?

PricewaterhouseCoopers (PwC), one of many consultancies pouring money into the buzzy technology, is making a pricey gamble based on the belief that generative AI will revolutionize the future of work and the global economy. PwC wants to position itself at the forefront of educating clients on best practices and use cases.

In order to spread generative AI knowledge and expertise, the professional services firm is enacting an upskilling campaign to familiarize its US rank-and-file with the tech.

“We think it’s really important that our people understand both the power of the tools and the responsibility they have to apply to the tools in the right way,” Joe Atkinson, PwC’s vice chair and chief products and technology officer, said at PwC’s tech showcase in May.

As the generative AI hype train barrels forward, some organizations increasingly view fluency in the tech as a vital skill, and AI ethicists implore companies to train their people to work alongside it. Unlike a variety of corporate giants that have banned use of certain generative AI products such as ChatGPT, PwC will attempt to acquaint 65,000 US employees with generative AI on various levels of technical proficiency over a three-year period, according to a company announcement.

Julia Lamm, workforce transformation partner at PwC, told HR Brew that the upskilling campaign will immerse employees in practical scenarios, allowing them to apply what they’ve learned. “If you don’t create a space for the people who did the learning to apply the learning soon thereafter, you might as well set your money on fire,” she said.

Responsibility. PwC is partnering with OpenAI and Microsoft, which are providing the consultancy access to their GPT-4/ChatGPT and Azure OpenAI services respectively.

The momentum behind generative AI is growing: Prominent tech giants have made commitments to invest in the area after months of layoffs, and leaders in the generative AI space have recently met with lawmakers and the president to discuss potential harms and the consequences of misuse.

Despite the uncertainty expressed by certain companies surrounding the rapid adoption of the technology, PwC sees early adoption as an advantage. When asked about the mounting unease voiced by prominent technologists regarding the blinding pace of development, Mohamed Kande, vice chair of US consulting solutions of PwC, said the company is prioritizing safety and responsible use of generative AI.

“A lot of the people saying slow down for AI, or to not use AI [are speaking] out of fear,” he argued at the showcase. “We are actually saying that AI will be used responsibly. And that is a role that we can play.”

Chief among PwC’s goals in the project is to tamp down on the growing climate of mistrust. In its statement announcing the program, the company posed a question: “How can we drive trust in the models, confirming they are fair, explainable and transparent?”

Learning AI. The upskilling program at PwC, hosted on a central hub called My+, also contains benefits information and leadership programs. According to Lamm, employees can access tutorials that vary in technical complexity, from primers on the basics of generative AI to tactical exercises. She said that “some people will get much more technical and will build generative AI models” based on the My+ program material.

PwC has experience orchestrating a massive upskilling program, and is drawing on the criteria of a previous campaign from 2017. In partnership with the software developer UiPath, PwC sought to upskill the entirety of its global workforce in robotic process automation, which involved skills such as “data visualization [and] data modeling,” according to Lamm. During that process, Lamm explained, “We had people who were in formal roles where they spent their time learning and then going back into the practice areas to immediately apply when they learned.”

The generative AI campaign will look similar, and measure progress through conversations between managers and their direct reports, with regular questions such as, “Have you had your teams create something or use responsible AI?”

With the upskilling campaign in its nascent stages, the organization doesn’t expect to be “minting PhDs in quantum AI,” Atkinson said. But the overall goal is to lead as technology evolves and potentially creates new roles, like prompt engineers, who train and refine text-generating AIs.

“In generative AI today, you have new professions that did not exist six months ago,” Kande said. “We will have to develop our own prompt engineers, taking some of our data engineers, data scientists, and make sure that we give them the training…to be more specialized when it comes to generative AI.”