Has generative AI killed machine learning?

Features
By
published

As the sector embraces new AI models, what role can machine learning continue to play?

The words ’Has generative AI killed machine learning?’ overlaid on a blurred, desaturated render of ribbons billowing out from a central, square-shaped hole to represent machine learning and AI. Decorative: the words ‘generative AI’ and ‘machine learning’ are in yellow, while other words are in white. The ITPro podcast logo is in the bottom right corner.
(Image credit: Future/Unsplash - Google DeepMind)

Machine learning (ML) has been a field of research for more than 50 years, and as a subset of artificial intelligence has also been the focus of great innovation in that time. 

Many businesses use machine learning throughout their stack and some will have relied upon ML frameworks without realizing it. With the rise in popularity of newer forms of AI developments such as generative AI, however, some have questioned the extent to which traditional approaches such as machine learning algorithms still have a place in the tech industry. 

In this episode Jane speaks to Sascha Heyer, senior machine learning engineer at DoiT, to explore whether ML still has a role to play in a world that is more interested in conversational AI.

Highlights

“The future is always hard to predict, no one was expecting ChatGPT and GPT in general and I don't know what's happening next year. But I think we will have some more traditional machine learning projects for quite some time, especially because you can combine those traditional machine learning approaches with natural language model approaches.”

“Without large language models engineers had to consider various models, from simple linear regression to more complex ensemble models or deep learning models. And they had to adjust them, according to their use cases, which required a lot of manual steps.”

“If you're just going a couple of months back before ChatGPT and then after ChatGPT it’s a completely different world on how you solve challenges with machine learning. But that doesn't mean that traditional machine learning is obsolete. It still excels when a data set is small, use cases easy, interpretability is crucial, or when the use case is too individual to be solved with large language models.”

Footnotes

Subscribe

Rory Bathgate
Rory Bathgate
Features and Multimedia Editor

Rory Bathgate is Features and Multimedia Editor at ITPro, overseeing all in-depth content and case studies. He can also be found co-hosting the ITPro Podcast with Jane McCallion, swapping a keyboard for a microphone to discuss the latest learnings with thought leaders from across the tech sector.

In his free time, Rory enjoys photography, video editing, and good science fiction. After graduating from the University of Kent with a BA in English and American Literature, Rory undertook an MA in Eighteenth-Century Studies at King’s College London. He joined ITPro in 2022 as a graduate, following four years in student journalism. You can contact Rory at rory.bathgate@futurenet.com or on LinkedIn.