Enterprise IT moves forward — cautiously — with generative AI

- Advertisement -

Jeter has similar concerns. While his team used ChatGPT to identify code fixes and deploy them to a website in under 30 minutes – “without ChatGPT it would have taken much longer” – and he thinks it helps draft terms and conditions in contracts. While useful to prepare, it is not fully proven. “We will not disclose any productive AI to outside members,” he says. “Truestone will not be bleeding edge in this space.”

- Advertisement -

Gary Jeter, EVP & CIO, Truestone Financial Credit Union

Truestone Financial Credit Union

- Advertisement -

When Truestone eventually begins using the technology for the benefit of its members, he says, it will monitor interactions through human and automated review to protect its members and the brand.

Today, the key to successful deployment is still a human in the loop to review generated content for accuracy and compliance, says UCSD’s Kellen. “Making sure that the machine makes the right decisions becomes an important litigation point,” he says. “It will be a long time before organizations [use it] For anything that is high-risk, such as a medical diagnosis. But generative AI does a fine job of generating something like review summaries, provided a human is overseeing them. “It slows us down a bit, but it’s the right thing to do,” he says. Eventually, he says, “we’ll find automated ways to make sure the quality is good. But right now, you should have a review process in place to make sure the content you produce is accurate.”

- Advertisement -

vince kellen

Vince Kellen, CIO, UCSD


Another well-documented risk, in addition to accuracy, is the potential for bias in models introduced from the data used to train them. This is especially problematic when generative AI is using content from the internet, as ChatGPT does, but it may be less of an issue when the model is trained against your own private corporate data that you can potentially bias. Can review for, says Kellen. “The higher you go into the enterprise, where the class of data is more constrained and more mundane, the more productive AI shines,” he says.

The thing to understand about large language models, says Cenkl, is that these machines are somewhat intelligent. “They don’t understand, but they’re very good at computing,” he says.

Change in job responsibilities, roles

“Technology has made things better, but it has also created a lot of extra work for us,” says Mohamed. However, he believes that Generative AI is different. “It’s exciting because it’s going to take away some of the stuff we don’t like to do and make us more intelligent,” he says. “It will enhance humans.”

But Curran points out that there is no expectation that generative AI will completely replace the role in the short term. “This can reduce the number of people needed to perform a role, such as in content development, product information management or software development,” he says. “But there will always be a need for a human in the loop.” And Mohamed says that even though technology can write and summarize, human intelligence will always be needed to control what is generated to ensure quality and to improve it.

steps to get started

Kellen says now is the time to get generative AI technology up to speed and start experimenting. “CIOs need to put their minds inside this puzzle before they are duped by vendors who are embedding the technology into their enterprise software offerings,” he says. “If you spend the next year procrastinating, you’ll be behind the curve.”

It’s important to get educated and go deeper than the public discussion on ChatGPT, says Curran, to understand that this technology is much more complex than a single application. Then start considering use cases where generative AI can improve the efficiency or quality of existing processes. Finally, ask what kind of capabilities you’ll need and whether you should get it from a vendor or build it yourself.

From there it is a matter of testing the technology and considering possible use cases. “A lot of your systems, whether they use structured or unstructured data, will have at least some component of natural language and conversational interface,” Senkal says. “Think about the data you have and what parts of it can be augmented by these technologies,” and then demonstrate the potential. For example, Jeter says he drafted the terms and conditions and sent it to his compliance department to show them how they could use it.

Generative AI models are large, and training them from scratch is expensive, so the best way to get started is to use one of the cloud services, Kuran says. CarMax, for example, uses Microsoft’s Azure OpenAI service with GPT 3.5. “The data we load is our own – it is not shared with others,” says Mohamed. “We can have massive amounts of data and process it very quickly to run our models. If you have a small team or business problem that could take advantage of Generative AI techniques, try this “

Source link

- Advertisement -

Recent Articles

Related Stories