Accenture’s Chief AI Officer Lan Guan on Why This Is a Defining Moment

A year ago this week, artificial intelligence vaulted into public consciousness with the release of OpenAI’s instantly viral consumer-facing, conversational interface. AI became one of the biggest stories of 2023. And businesses everywhere (including OpenAI itself, as reflected in its leadership pandemonium earlier this month) are struggling to navigate the potential benefits and risks of the extraordinary technology.

Few have a better window into how all this is unfolding than Lan Guan, Chief AI Officer of Accenture, the consulting juggernaut with more than 700,000 employees, most of the globe’s biggest companies as clients and its own recently announced $3 billion investment in AI. She’s also a member of Stanford’s Institute for Human-Centered AI, a prominent center focused on ensuring AI benefits humanity–and a very early adopter, having built a robot to teach kids English in rural China when she was just 16.

I spoke to Guan at the annual workplace summit hosted by TIME’s partner, Charter, in New York. An edited and condensed version of that conversation follows:

It’s been a wild couple of weeks for OpenAI and the industry. What does it mean for companies and the uptake of this technology?

It’s important to decouple the technology from these developments—because what we can say definitively, is that the value proposition of generative AI is not in question. Since ChatGPT’s launch, we’ve seen this particular space go from largely a few sets of democratized solutions, to being embedded across the entire value chain. All major platforms used by companies have, or are beginning to have, GAI capabilities. Across our own client base, we’ve seen things happening in matter of months versus years. The momentum is strong and will continue.

How are you using generative AI personally?

I do research a lot during my free time! I used to go to Google Scholar to find an archive paper for something that I’m working on. I have found that over the last couple months I have been relying on ChatGPT a lot. At work, I was actually preparing a course I’m teaching within Accenture on generative AI. So I asked it, what are the most popular machine learning algorithms before Generative AI?

Did you like the answer?

I was very satisfied with the answer. So I passed my test! At the office, we use a lot of generative AI; for example, to summarize meeting notes. We also build a lot of in-house applications. Over the last 10 months, we have built 300 internal applications using generative AI. Think about customer proposals; easily, dozens of pages. Now, we can have AI do the first draft. We have AI learning what we have developed before, and how we responded before, to write the first draft.

What are you seeing with your customers? What is that level of adaptation relative to what you’re describing inside Accenture?

A lot of it is about knowledge management. Think about the modern enterprise workplace as a lot of knowledge. This is actually underutilized—all kinds of documents, all kinds of images, all kinds of recordings. Like the recording that we’re doing here [of this interview]—this can be analyzed by AI years later to find what you and I talked about, and what were the trends at that point in time. So that’s just one example—using the immense power of generative AI to actually parse a lot of knowledge from unstructured documents within every enterprise to actually make use of that. For example, that’s something that we have seen very commonly used by insurers globally to help claim agents. They can use generative AI to answer whether a particular claim is meeting the policy requirements in this local market. 

A lot of us woke up at the end of November last year with the ChatGPT release, tried it ourselves and realized, you know, we’re in this new world. How much of what you’re describing is a post-ChatGPT phenomenon? Could that insurance example have been done two years ago? Or is that a phenomenon of recent months?

The ability existed before but was not easily accessible by people in the workplace. ChatGPT and this new class of AI completely changed the game. The example of insurance agents is regular now because of the ease of the use of the technology. AI is actually making sense to the general public and becoming more pervasive within modern enterprise.

There’s enormous agreement in the C-suites that we’ve got to move quickly here, but we also know that many CEOs and other leaders feel stuck. How do we get unstuck? What does that look like?

We need to figure out a way to articulate the value of investment into generative AI so that the business case is much more compelling. There is a lot of uncertainty and fear. But while generative is a new class of AI, it’s building upon decades of investment. This is not a black box. I think that kind of conversation needs to happen more to instill confidence into the C-suite conversation. We’re getting there.

The formal title of this session is “Hype vs. Reality.” Are we in a hype phase? Or is this really a grab-it-or-die kind of moment for businesses?

I don’t think this is hype. This is the defining moment across many industries. We have a lot of work to make sure this technology is democratized and not limited to a small group.

Your title is Chief AI Officer. Should that be a more widespread role?

I believe every organization needs to have a Chief AI Officer or someone in that capacity to define the overall AI strategy within the organization.