Artificial intelligence may not pose a threat to humanity, but it could soon dramatically change the way we work. For 2024, the most significant area of technology disruption may be within the enterprise.
Every corporation has a treasure trove of internal documents: emails, memos, and presentations. Custom AI models will enable workers to unlock knowledge that is rarely accessed.
On Nvidia
Of
On the last earnings call, CEO Jensen Huang explained how AI will change how employees work with data. Traditionally, information retrieval requires specific technical queries on a specific storage device. This process often fails because the computer system does not understand the request. But now, with the latest AI technology, employees can easily generate answers to their needs using common conversational language. “Generative AI is the biggest TAM. [total addressable market] software and hardware expansion that we’ve seen in decades,” Huang said.
Sandra Rivera, general manager of Intel’s data center and AI group, says a technology pilot with Boston Consulting Group demonstrated the potential of using AI in the workplace. Both companies trained an AI model on BCG’s internal data. For the first test, they asked for a presentation that several workers couldn’t find after eight hours of searching.
“We plugged in the model and explained the PowerPoint. In less than two seconds, he got it,” she says. “It’s a clear example of productivity. You can get what you’re looking for instantly with your native language.”
Advertisement – Scroll down to continue.
Nvidia and Microsoft have published separate academic papers on their ability to improve employee productivity, in addition to documenting large language AI models.
Nvidia has published an example of ChipNeMo, a custom big-language AI model trained on the company’s technical data. The model has helped engineers write code and search for complex information, empowering them to design chips faster.
Last month, Microsoft released data that shows users of its Copilot for Microsoft 365, its productivity AI assistant that uses large language models, can complete tasks, including writing and summarizing — 29 percent faster than those not using the service.
Advertisement – Scroll down to continue.
Examples show that companies can accelerate software development, marketing efficiency, and customer service by using AI models with their proprietary data.
This means that certain types of jobs may eventually be eliminated. The increasing use of AI models that can automate human resources and customer service requests will reduce the need for phone-based employees in call centers.
A tidal wave of creative AI tools for image and video creation from Adobe startups like Pica will reduce demand for outsourced artists.
Advertisement – Scroll down to continue.
But overall, AI has one big positive impact: It will allow employees to complete tedious, mundane tasks more quickly, leaving them more time for important, high-level work.
Write to Tae Kim at tae.kim@barrons.com.