Artificial intelligence and generative AI have maintained a residency atop headlines this year, commanding public discourse and creating regulatory noise. Technology leaders have helmed AI strategies in their organizations while also using the tech in their daily work.
More than 9 in 10 IT leaders and other upper management respondents currently use AI to support their workloads, according to a September Freshworks survey of 2,000 IT professionals globally.
AI adoption among top tech leaders outpaces others in the organization, Freshworks found, as just one-third of individual contributors implement the technology in daily workflows, according to the report.
To see exactly how CIOs and technology leaders are using generative AI in their day-to-day, CIO Dive asked C-suite executives from leading adopters and vendors. Here’s how five tech leaders wield generative AI:
Vittorio Cretella, CIO at Procter and Gamble
P&G rolled out an internal generative AI model in September, which was built using OpenAI’s API and supports more than 35 use cases, according to CIO Vittorio Cretella.
But in his day-to-day, Cretella does not use generative AI “as much as you would think,” he said.
“I’m pretty old fashioned and I feel guilty if I use it for, like, a speech,” Cretella told CIO Dive. Cretella mostly uses generative AI tools to summarize long email threads and meeting scripts to extract the key points. “I find that extremely useful,” he said.
Sandeep Davé, chief digital and technology officer at CBRE
CBRE Group built an AI playground for employees to leverage tools in a secure environment. Engineers mainly use GitHub’s Copilot, while other employees innovate with tools in the playground.
Similar to Cretella, Sandeep Davé, chief digital and technology officer at CBRE, typically only uses generative AI to summarize and extract key points from presentations and emails.
When The National Institute of Standards and Technology released its cybersecurity and AI guidance, Davé leveraged generative AI models within the self-service playground to summarize the lengthy documentation. He also uses it to assist crafting an email on a complex topic.
It’s all about “efficiency gains,” he added.
Naveen Zutshi, CIO at Databricks
Databricks is working on building proprietary generative AI tools for its customers while also deploying the technology for internal use cases, the company’s CIO Naveen Zutshi told CIO Dive. For in-house tools, IT teams are using retrieval augmented generation to tune off-the-shelf models, he said.
“We have teams building models and teams building retrieval augmented generation on existing models,” Zutshi said.
The data cloud company has an in-house model deployed on Slack its website, as well as chatbots in production for customers and customer-support reps.
“We also have a chatbot which will help improve our knowledge base and one that we’ve build for job descriptions,” said Zutshi
To train a contract-summarization chatbot, IT partnered with legal to create the necessary prompts.
“Only a small number of companies will build their own foundational model,” Zutshi said. “In most cases, you don't need it. For most customers, you either need to use a retrieval augmented generation use case on top of a foundational model or you may use multiple models. There are a lot of good ones available now in the marketplace. You just want to make sure that you control your data and your IP.”
Prasad Ramakrishnan, SVP of IT and CIO at Freshworks
Freshworks has several generative AI-powered solutions available to customers while also leveraging the technology internally.
“I’m taking the use case-based path,” said Prasad Ramakrishnan, SVP of IT and CIO at Freshworks. “I’m trying to figure out [if there is] a problem before I go into the solution.”
One of the solutions Ramakrishnan is evaluating is Google’s Bard API. Ramakrishnan sees the tool’s potential in helping high-level executives and those in marketing to spin up copy for emails or presentations, but he isn’t looking to deploy it to the whole enterprise.
“If I can save some time, if the pricing is right and if the value is right, I’ll deploy it,” Ramakrishnan said.
Apratim Purakayastha, chief technology and product officer at Skillsoft
Since the rapid rise in interest in generative AI, Skillsoft has launched several courses and curriculum related to the technology. Internally, teams have participated in innovation sprints and hackathons to bring them up to speed, explore new areas and potentially learn new skills.
Technology leaders within the company use generative AI tools for a host of tasks.
“I am regularly using generative AI to prepare drafts of technical talks, summarize large technical documents and judge the complexity of a coding task at a high level by asking generative AI to produce code for the task, which gives me insight on engineering estimates,” Apratim Purakayastha, chief technology and product officer at Skillsoft, said in an email.