Are We Starting to Sound Like ChatGPT? | Lauren Mushro, Responsible AI Lead

Have you noticed how easy it is to recognize an email written with ChatGPT? Lauren Mushro is researching something bigger: whether AI is starting to change the words we use, and what happens to local dialects when language models tend to favor the most predictable language.

It's called "linguistic flattening," and it raises an interesting question: Are we starting to learn language from AI instead of just teaching AI our language?

Lauren is the Responsible AI Human-Centered Design Lead at BMO and an AI ethics professor at CUNY. We talk about how AI is affecting language, why critical thinking matters more as we rely on these tools, and how smaller AI models could help preserve the expressions that larger models overlook.

In this conversation

• Why ChatGPT and other language models can make writing sound increasingly similar
• What linguistic flattening could mean for slang, regional dialects, and languages with less training data
• The difference between using AI to learn and losing skills by relying on it too much
• Why AI companies scan physical books for model training, and the questions this raises about copyright
• The dead internet theory and how AI-generated content is changing what we trust online
• Why Lauren is building a small language model to better recognize Caribbean Spanglish

About Lauren

Lauren Mushro, PhD, is the Responsible AI Human-Centered Design Lead at Bank of Montreal (BMO) and an Adjunct Associate Professor in AI Ethics at the City University of New York. She holds a PhD in Modern Languages from Johns Hopkins University and researches how AI systems affect language, culture, and the way we communicate.

Connect with Lauren

Website: https://laurenmushro.com/

LinkedIn: https://www.linkedin.com/in/laurenmushro

Mentioned in this episode

Center for AI Safety: https://safe.ai/

Machine Gods with Casey Newton and Kevin Roose: https://machinegods.fm/

Shae O. Omonijo, Critical Thinking in the Age of AI: https://www.instagram.com/iamshaeo/

Partner

Paid partnership: This episode includes paid promotional content from Omnisend.

Omnisend helps ecommerce brands manage email and SMS marketing. By connecting Omnisend to ChatGPT or Claude, you can ask questions about campaign performance and get a clearer picture of what needs your attention.

Learn more about Omnisend: https://your.omnisend.com/HOWAI

Timestamps

00:00 Introduction: How AI Is Changing the Way We Think

04:11 Responsible AI and the Impact on Language

07:21 AI in Education and the Critical Thinking Crisis

12:49 Why AI Companies Are Scanning Physical Books

17:59 Linguistic Flattening: Is AI Changing Our Language?

22:04 What Should We Stop Outsourcing to AI?

25:04 The Dead Internet Theory and AI-Generated Content

32:19 Final Takeaways and Building AI for Spanglish

Listener offer

Save 5% on Leela Quantum Tech with code BROOKE: https://leelaq.com/?ref=Brooke

Affiliate disclosure: I may earn a commission if you purchase through this link, at no additional cost to you.

The AI Movement With Brooke is a weekly podcastexploring how AI is changing the way we build, create and make decisions.

Follow on LinkedIn: Brooke Gramer

Website: TheAIMovementPodcast.com

Disclaimer: This podcast is for informational and educational purposes only and does not constitute professional advice. Guest opinions are their own. Content is provided without guarantees of accuracy or results. Always verify information before acting. To the fullest extent permitted by law, Brooke Gramer, The AI Movement, and EmpowerFlow Strategies LLC disclaim liability arising from reliance on this content.