Managing Model Collapse in AI Models

The chapter explores the risk of model collapse in AI models trained on AI-generated data and suggests grounding the prompt with trusted data to ensure accurate responses. They discuss evidence of collapse in the industry and emphasize the importance of human oversight in content generation by chat GPT or LLM models.

Play episode from 26:40

chevron_right

Transcript

chevron_right

Transcript

Episode notes

Discourse over responsible workplace use of generative artificial intelligence software like ChatGPT, is just getting started.

Synopsis: Every second Monday of the month, The Straits Times' Career Talk podcast helps you take charge of your own career and make your ambition pay off.

Generative artificial intelligence (AI) models can swiftly generate convincing text and images based on user prompts. Suddenly, it seems jobs thought immune from automation, like those involving creative expression and communication, could be forever changed with Gen AI.

In this episode, host Tay Hong Yi speaks with his guests on the limitations and responsible use of the technology at the workplace.

His guests are:

Mr Gavin Barfield, vice president and chief technology officer for solutions, Salesforce Asean

Assistant Professor Jerrold Soh from the Yong Pung How School of Law, Singapore Management University

Highlights (click/tap above):

2:35 How is generative AI being used at the workplace?

7:00 Why are some employers and workers anxious about generative AI?

11:28 Who is responsible if generative AI software makes a mistake?

12:47 Is it a good idea for companies to draw up rules of engagement with the technology?

15:22 What are some principles employers and employees can abide to help them use generative AI responsibly?

22:15 Can generative AI be overused?

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app

Home Top podcasts Popular guests Top books