top of page

Another Crazy Day in AI: Why Prompts Aren’t One-Size-Fits-All

Another Crazy Day in AI: An Almost Daily Newsletter

Hello, AI Enthusiasts.


Almost Friday, but AI doesn’t sleep!


A new study from Wharton asks: Does adding “please” actually improve AI responses? The answer? Not as much as you’d think.


Meanwhile, Google is giving search an AI boost by integrating Gemini 2.0, making Overviews more advanced and interactive.


And in classrooms, UK schools are using AI in creative ways, from history lessons with virtual Charles Darwin to AI-powered writing assistants.


The weekend is near—AI is still going full speed.


Here's another crazy day in AI:

  • What works in AI prompts and what doesn’t

  • English schools challenge AI skepticism with innovation

  • Google enhances search with AI Overviews and new AI Mode

  • Some AI tools to try out


 

TODAY'S FEATURED ITEM: Politeness May Not Help Your AI Prompt


A robotic scientist in a classic white coat with 'AI Scientist' on its back stands beside a human scientist with 'Human Scientist' on their coat, looking towards the AI Scientist.

Image Credit: Wowza (created with Ideogram)


Does saying "please" to an AI actually make it perform better?


A new report from Wharton's Generative AI Lab challenges assumptions about prompt engineering. Led by Ethan Mollick and his team, the study examines whether certain prompting techniques—like adding polite language—consistently improve AI performance. The findings? It’s not that simple.


What they found along the way:

  • There’s no single standard for measuring AI performance—different benchmarks can lead to very different conclusions.

  • Techniques like politeness or structured formatting can sometimes improve responses but can also make them worse.

  • AI responses aren’t always consistent—small changes in wording can shift results in unpredictable ways.


Meincke, Lennart and Mollick, Ethan R. and Mollick, Lilach and Shapiro, Dan, Prompting Science Report 1: Prompt Engineering is Complicated and Contingent (March 04, 2025). Available at SSRN: https://ssrn.com/abstract=
Meincke, Lennart and Mollick, Ethan R. and Mollick, Lilach and Shapiro, Dan, Prompting Science Report 1: Prompt Engineering is Complicated and Contingent (March 04, 2025). Available at SSRN: https://ssrn.com/abstract=

The study highlights an important reality: AI doesn’t follow fixed rules when it comes to prompts. Many people assume that small tweaks—like making a request more polite—will always lead to better results. But as this research shows, that’s not always the case. What works in one context may fail in another, and AI’s behavior can be surprisingly inconsistent.


It also raises questions about how we evaluate AI performance. Since results can vary based on how success is measured, businesses and researchers need to be intentional about their benchmarks. A model that looks effective under one test might not hold up under another. Understanding these nuances will be key as AI continues to play a bigger role in decision-making, from education to business strategy.




Check out Mollick's LinkedIn post here.

Read and download the full paper here.

 

OTHER INTERESTING AI HIGHLIGHTS:


English Schools Challenge AI Skepticism With Innovation

/Richard Adams, Education Editor, The Guardian


Schools across England are using AI in creative ways to make learning more engaging, from chatting with a virtual Charles Darwin about evolution to turning student writing into AI-generated images. Teachers say AI has boosted student participation, particularly among those who are typically more reserved. While educators remain cautious about potential risks, they recognize AI as a tool that will shape students’ future careers. Many schools are also developing policies and student-led discussions to ensure AI is used responsibly and effectively in the classroom.



Read more here.

 

Google Enhances Search With AI Overviews and New AI Mode

/Robby Stein, VP of Product, Google Search


Google is expanding AI Overviews with Gemini 2.0, making search results faster, more accurate, and capable of handling complex questions. The new AI Mode, currently in limited testing, enhances reasoning and problem-solving capabilities, allowing users to ask detailed questions and receive AI-powered responses with follow-up interactions and deeper insights. Unlike traditional search, AI Mode combines real-time web data with Google's advanced AI, offering a more comprehensive and intuitive way to explore information. Google plans to refine the feature based on user feedback before a wider release.



Read more here.

AI Mode helps you search more effortlessly by combining the advanced capabilities of Gemini 2.0 with Google’s best-in-class information systems. Note: Results are for illustrative purposes and include future-looking features. Source: Google
AI Mode helps you search more effortlessly by combining the advanced capabilities of Gemini 2.0 with Google’s best-in-class information systems. Note: Results are for illustrative purposes and include future-looking features. Source: Google
 

SOME AI TOOLS TO TRY OUT:


  • Gong - Analyzes sales conversations to predict revenue and boost team performance.

  • Gentle Invoice - AI inbox assistant for categorizing and extracting key invoice details.

  • Gemini Deep Research - Auto-browses the web to generate in-depth reports in minutes.

 

That’s a wrap on today’s Almost Daily craziness.


Catch us almost every day—almost! 😉

 

EXCITING NEWS:

The Another Crazy Day in AI newsletter is on LinkedIn!!!



Wowza, Inc.

Leveraging AI for Enhanced Content: As part of our commitment to exploring new technologies, we used AI to help curate and refine our newsletters. This enriches our content and keeps us at the forefront of digital innovation, ensuring you stay informed with the latest trends and developments.





Comments


Subscribe to Another Crazy Day in AI​

Catch us almost every day—almost! 😉

Thanks for signing up!

Copyright Wowza, inc 2025
bottom of page