Over the last several months, OpenAI, and ChatGPT in particular, has shown what’s possible with a user interface built on top of a large language model that can answer questions and create code or pictures. While that alone is remarkable, we can also interact with and adjust the byproduct by having a conversation of sorts with the AI. It’s amazing really, but think about how transformative this could be by applying it to the enterprise applications you use on a daily basis.
What if you could build an interface on top of your existing applications, so that instead of pointing and clicking, you could simply ask the computer to do a task for you and it would do it, based on the applications’ underlying model or your company’s internal language model.
That would be a huge leap forward in computing. Before now, the biggest leap happened in 1984, when Apple introduced the graphical user interface that began a slow shift from the command line approach and eventually went mainstream in the early ’90s with the release of Windows 3.1 and later Windows 95.
We’ve had other UX attempts, such as voice interfaces like Siri and Alexa, and while they brought some changes to the consumer side of things, they’re still not exactly the same thing as a computer producing work for us. It’s just finding some answers and in some cases executing simple commands.
It certainly didn’t change the way we work, and that is the true measure of whether a new computing approach is truly transformational. If you could simply type an action like “Help me onboard a new employee” or “Generate a monthly P&L statement” instead of explicitly guiding the systems on what to do, that would be a fundamental leap forward in UX design.
That’s what generative AI has the potential to do, but like anything else, it’s going to require some creativity to design these new interfaces in an elegant way, so it doesn’t feel like it’s bolted on to your old point-and-click interface. It’s also probably going to require more-focused large language models.
It’s important to remember that even though this is part of an ever-evolving AI landscape, we are still in the beginning of this particular stage of the technology. Generative AI still has a long way to go, no matter what the social hype machine may suggest. That doesn’t mean there aren’t some exciting possibilities, but it’s still early.
Ten years ago people began to expect a consumer-like experience in their enterprise software, the way they were used to interacting with software at home and on their phones and tablets. And now, expectations might be shifting again: If people can’t interact with your software in ChatGPT-like fashion, they could end up demanding it.
A 2-year-old can do it
The sheer simplicity of ChatGPT is what has stood out to Jeetu Patel, executive vice president and general manager of security and collaboration at Cisco. “The ease of use with which OpenAI has actually gone out and deployed ChatGPT, it’s like a 2-year-old could use it, and I don’t think AI systems in the past have been that elegantly designed, where a 2-year-old can use it,” Patel told TechCrunch+.
He says it takes time and thoughtfulness to build something that simple. “And I think we will see more innovation around user experience and innovation around custom sets of data that can then apply to very, very specific use cases,” he said.
Patel sees a future world where generative AI is actually helping build more elegant UIs. “We are in such early days where the combination of multimodal and having language models instruct how UIs get constructed themselves, there’s so much opportunity,” he said.
He believes as this plays out every UI will eventually have a natural language component to it, where you might just ask a question to conduct certain actions and activities.
Assaf Baciu, co-founder and COO of Persado, which provides specialized generative AI solutions for marketing, believes that such a UX transformation is coming. “Generative AI in its essence allows you to interact with machines via natural language, and it will drastically transform how we interact with software,” Baciu said.
He certainly sees the potential to change the way we use enterprise software by letting us ask the software for an answer. “[For example, consider] the ability of Salesforce users to auto-generate sales tasks like composing emails, scheduling meetings and preparing for the next interaction. This will further evolve with capabilities such as ChatGPT plugins, which will allow a user to execute tasks by simply describing them as the model connects to specific “features.” For example, ‘create a report showing each quarter with the lead in stage two and to which salesperson they are assigned along with their total booking value.’”
The way we interact with the interface is just the starting point when it comes to the impact of AI on enterprise software, according to Omar Johnson, founder and CEO at professional services firm Opus United and former CMO at Beats. “It’s going to start to anticipate you. So I believe enterprise software will start to move in the world more like a Netflix where it learns [your way of working] and it sends you things,” Johnson said.
That could change the way we think about software in the future. “That is going to open up how you think about users. Typically your enterprise software users in one department are really deeply vertical. But if you have something that’s really easy to use, you actually expand horizontally across functions,” he said. That also means it’s not confined to one group like marketing or operations but instead can integrate across workflows moving around the company.
All of this comes with a big caveat: We don’t know where any of this is going quite yet. But it seems that we are in the midst of some tech inflection point where how we interact with computers and software is about to change in an elemental way, and that’s hard to ignore. And somehow, AI is going to be a big part of that.