Opções de idiomas

Idioma selecionado:

13 Min Read

Executive insights: Generative AI’s top obstacles and opportunities in the enterprise

março 21, 2024 / Joel Raper | Weston Morris

Short on time? Read the key takeaways:

  • The incredible potential drives much of organizations' interest in generative AI.
  • Employees of all ages are apprehensive about generative AI, citing fears of the technology taking people’s jobs. Some job titles are more disrupted than others, but this isn’t necessarily bad.
  • One of the biggest limitations to tapping generative AI’s potential is access to clean data – and keeping it secure also remains a concern.

Generative AI is getting plenty of buzz, but is it also being integrated into business operations? On the ground, organizations have plenty to say about the biggest motivations and top challenges with implementing generative AI.

During a recent conversation, Unisys Senior Vice President and General Manager of Digital Workplace Solutions (DWS) Joel Raper and Senior Director of DWS Global Strategy Weston Morris spoke about what Unisys is hearing from clients and industry peers. Read on for their insights and predictions for the future of generative AI in the digital workplace.

Why are enterprises exploring the implementation of generative AI?

Joel: So much money is being thrown at generative AI to answer that question: to determine the “why.” Organizations are being told to introduce AI because it will save them money. The potential is there for orders-of-magnitude changes in productivity.

Every single organization we talk to now has an edict from the top down to look into generative AI as executives say, “I read yesterday that generative AI is going to change the world. How will it change the world in our space? How much money can I save?” Figuring that out is where the challenge lies.

Weston: Think of it this way: Imagine if everyone in 1970 suddenly got a smartphone. People would say, “I’ve got a phone on the wall and there are payphones on every corner. Why do I need this?” They would have a hard time seeing the bigger picture and recognizing a smartphone’s use cases: browsing the internet, keeping track of to-do lists and reminders, chatting with friends and colleagues, and so much more. In a similar vein, if organizations don’t think about how generative AI will be applied to specific business problems, simply deploying AI to the masses will end up being an expensive science experiment with no return on investment.

When someone says, ‘I want…” with any new technology, I ask, “Why? What are you trying to accomplish with it?” With generative AI, let’s talk about the “why” first in terms of the business problem you’re trying to solve and how this technology will be used in your organization.

One perspective is that younger employees are more comfortable with digital workplace solutions and will accelerate AI deployment, while those with more tenure in the workforce are anxious about how AI will be used. What do you think of this perspective?

Joel: Many employees have apprehension about generative AI, regardless of age. It has more to do with job types; some job types are at a higher risk of being disrupted, if not displaced. This is especially true for creative job types with tasks such as content creation, which is an easy pick because you can use existing toolsets like ChatGPT or Microsoft Copilot to impact that business process today.

On the nerdy side, taking a code set written in Java and moving it to Go or creating a Python script is possible today with generative AI, even with noncommercial models, and you can get the answers quickly. In the enterprise space, you wouldn’t expose that code to the internet, so there is still work to be done there.

Some commentary around generative AI involves people’s anxiety about losing their jobs. Do you think those fears are justified?

Joel: On the contrary, AI has opened the opportunity for new jobs and different types of jobs because that human-in-the-loop factor is very much needed, as great as ChatGPT, Copilot and other generative AI technologies may be.

Recently, I heard of a company where many of their contracts were written by generative AI. You could take that to mean the lawyers’ jobs are at risk. But, that’s not what it’s saying at all. It's saying the busy work at the beginning stages of the contracts is now automated, so the lawyers are doing more meaningful work to add value to the organization. Will you need fewer lawyers? It’s quite possible. That's why people have both fear and excitement about generative AI.

Generative AI changes the type of role you have—it doesn’t necessarily eliminate it. The jobs of people in lower entry-level roles will change quite a bit, in the sense that their responsibilities will no longer involve creating from scratch. People will evolve to become manipulators of many different tasks that feed into a larger business outcome.

I'd argue that if you don't cut your teeth on the beginning work, however, you'll never develop the expertise you’ll need as a future lawyer to be able to review what will be written by generative AI. Half of anyone’s professional skill set is a result of the experience gained early in their career. We will have to take a step back at some point as a society and consider what that means if generative AI performs most of the early-stage work in a career path.

Weston: Is generative AI so awesome that it will completely remove the need for people? The answer is generally no. Instead, it will help people do things better and faster than they do manually today. When deployed effectively, Gen AI will streamline labor-intensive tasks.

Back when people made cloth by hand, someone came up with a loom, which lets you produce a large blanket in an hour instead of in three months. Like Joel said, the person’s role is changing in terms of what they’re doing. Now, they’re a loom manager. I need fewer people to make the blankets. So, one option is to reduce the size of my workforce. The other is to use the freed resources to make ponchos in addition to blankets.

What is needed for organizations to realize the potential of generative AI fully?

Joel: Understanding your data. We think we’re in a big data world right now. We’re not. We’re nowhere close to where need to be with big data in the enterprise for generative AI to be truly effective.

The limiting factor is 100% clean data – meaning consistent, correct and classified data – and in a quantity that’s relevant. It will take the industry a long time to get there. AI and data are closely connected, and you will have to spend time on data creation, organization, classification and validation to benefit from AI.  

Weston: Our customers are asking, “How should I use generative AI?” That’s not a question we saw before the pandemic. Their organizations have all this disparate knowledge stored in different locations, and it’s constantly being updated. Finding a solution to that data sprawl is hard, especially if changes are made to the tools that create the data or within the organizations that make use of the data. In theory, generative AI should be able to help solve that, which is what many of our AI pilot programs with existing clients address.

The bottom line is that out-of-the-box generative AI does not automatically and securely understand all your business data and processes. For example, these tools can’t inherently know the answer to company-specific questions like:

  • How do I submit a PO for a new vendor in Colombia?
  • What is the process for getting international travel approved?
  • What is our latest carbon reduction target for our European factories?

The answers must be fed to the AI first via secure, classified data. Much of the focus around generative AI has been on how friendly and human-like it is, but its real value is in giving you accurate answers and pulling data from various locations that have been siloed in the past. That functionality is what’s harder, but if you get it right, it will help people be more productive.

What does the future hold for generative AI?

Joel: We haven’t seen a massive implementation of generative AI yet. We’re going to see it continue to grow. I talk about Microsoft Copilot a lot because it is a tangible solution. As Copilot develops, we can better understand its value. You won’t experience the ROI of generative AI until you see its application to real business problems.

Weston: Generative AI excels in two important areas that will eventually make it ubiquitous. Every user interface has the potential to make use of an easy-to-use true natural language capability. More important, though, is the ability to access knowledge and information in a way that sorts the chaff from the wheat to get to what’s meaningful.

Consider how generative AI can propel your business objectives

Conversations like this one will continue across the globe in offices, on video calls and everywhere in between as executives evaluate how generative AI can benefit their organizations. Unisys can help you explore strategies for integrating generative AI into your workflows that will result in the most significant benefits for your business goals. Learn more about how generative AI is already shaping the digital workplace and how Unisys is guiding organizations on generative AI with Unisys Core AI.