Data Skeptic

Debiasing GPT-3 Job Ads

Oct 10, 2022

48:57

Creator website

Highlights

AI Chapters

Episode notes

Fine Tuning on Low Bias Job Ads Is Good From Both

02:00

What Is Prompt Engineering?

02:35

Do You Think GPT-3 Could Help With Job Ads?

03:00

Introduction

2min

Is GPT-3 Training on Bias Data Biased?

3min

Using Sentiment Analysis to Classify Texts

3min

Is There a Bias in Large Language Models?

2min

Detecting a Bias in a Large Language Model?

2min

Astrato - The Next Generation Solution for Machine Learning

6min

How Much of a Gain Can Be Made for Society?

5min

Do You Think GPT-3 Could Help With Job Ads?

3min

Is GPT-3 a Good Model to Write a Job Ad?

3min

10.

GPT-3 Bias and Job Ads

2min

11.

How to Fine Tune GPT-3 and Dali?

3min

12.

What Is Prompt Engineering?

3min

13.

GPTP

2min

14.

Fine Tuning on Low Bias Job Ads Is Good From Both

2min

15.

How to Measure Realism in the Final Analysis?

2min

16.

Machine Learning

6min

17.

Data Skeptic Add-in Tech - What's Next?

2min

We hear about the impeccable achievements of GPT-3 models, but such large generative models come with their bias. On the show today, Conrad Borchers, a Ph.D. student in Human-Computer Interaction, joins us to discuss the bias in GPT-3 for job ads and how such large models can be de-biased. Listen to learn more!