The panel discussion in this episode is one half of the annual InfoQ Architecture and Design Trends Report. The other half is the written report .

One of the regular features of InfoQ are the trends reports, which each focus on a different aspect of software development. These reports provide the InfoQ readers with a high-level overview of the topics to pay attention to this year, and also help the editorial team focus on innovative technologies across all the content on InfoQ.

Read a transcript of this interview: https://bit.ly/43WChIH

Subscribe to the Software Architects’ Newsletter for your monthly guide to the essential news and experience from industry peers on emerging patterns and technologies:
https://www.infoq.com/software-architects-newsletter 

Upcoming Events:

InfoQ Dev Summit Boston (June 9-10, 2025)
Actionable insights on today’s critical dev priorities.
devsummit.infoq.com/conference/boston2025

InfoQ Dev Summit Munich (October 15-16, 2025)
Essential insights on critical software development priorities.
https://devsummit.infoq.com/conference/munich2025

QCon San Francisco 2025 (November 17-21, 2025)
Get practical inspiration and best practices on emerging software trends directly from senior software developers at early adopter companies.
https://qconsf.com/ 

QCon AI NYC 2025 (December 16-17, 2025)
https://ai.qconferences.com/

The InfoQ Podcasts: 
Weekly inspiration to drive innovation and build great teams from senior software leaders. Listen to all our podcasts and read interview transcripts:

- The InfoQ Podcast https://www.infoq.com/podcasts/ 
- Engineering Culture Podcast by InfoQ https://www.infoq.com/podcasts/#engineering_culture 
- Generally AI: https://www.infoq.com/generally-ai-podcast/ 

Follow InfoQ:
- Mastodon: https://techhub.social/@infoq 
- Twitter: twitter.com/InfoQ
- LinkedIn: www.linkedin.com/company/infoq
- Facebook: bit.ly/2jmlyG8
- Instagram: @infoqdotcom
- Youtube: www.youtube.com/infoq

Write for InfoQ: Learn and share the changes and innovations in professional software development.
- Join a community of experts.
- Increase your visibility.
- Grow your career.
https://www.infoq.com/write-for-infoq

Transcript

InfoQ Architecture and Design Trends in 2025

Prompt Engineering Guide

Large Language Models

Agents Companion

🎙️ Welcome to "AI with Shaily," hosted by Shailendra Kumar, your trusted source for the latest and most insightful updates on artificial intelligence! Today’s episode dives into a critical and complex topic that blends technology, ethics, and society: how large language models (LLMs) like GPT handle sensitive and challenging discourse such as hate speech. 🤖🧠

Shailendra highlights a recent, thought-provoking essay by Dr. Emily Barrow DeJeu from Carnegie Mellon University, which sheds light on the limitations of these powerful AI systems. While LLMs excel at generating human-like text, they often struggle with the subtle nuances of racialized hate speech. This includes difficulties in understanding cultural contexts, power dynamics, and emotional layers embedded in hateful or discriminatory language. 🧩⚠️

A key takeaway from Dr. DeJeu’s research—and echoed by other studies from Carnegie Mellon—is the necessity of human-AI collaboration. AI tools can assist in detecting and moderating content, but they cannot replace the critical human judgment required to accurately interpret sensitive topics. Hate speech isn’t just about spotting keywords; it’s about comprehending deep societal tensions and historical backgrounds that AI isn’t fully capable of grasping yet. 🤝👥

Why does this matter beyond the tech world? Because these conversations are unfolding against a backdrop of rising global challenges like populism and threats to liberal democracy. Misinterpreting or overlooking hate speech can worsen social divides. Using LLMs without careful human oversight risks amplifying biases and causing real-world harm. 🌍⚖️

Shailendra shares a personal story from his experience in AI customer service automation, where early models mistakenly flagged culturally nuanced complaints as offensive—false positives that frustrated users and damaged trust. This experience underscored the importance of context and empathy, qualities AI alone cannot replicate. 💡🙅‍♂️

For AI enthusiasts and developers, Shailendra offers a crucial tip: always include a strong human review process when deploying AI to analyze or moderate sensitive content. This approach not only enhances safety but also promotes more ethical and accurate outcomes. ✔️🛡️

To close, Shailendra poses a reflective question to the audience: As AI evolves, how can society balance leveraging AI’s efficiency with preserving the essential human insights that maintain our social fabric? He quotes computer scientist Fei-Fei Li: “The future of AI is not AI alone; it’s human-centered AI.” This powerful statement reminds us that the true strength of AI lies in partnership with humans, not replacement. 🤔💬

For more in-depth AI news and discussions, follow Shailendra Kumar on YouTube, Twitter, LinkedIn, and Medium. Don’t forget to subscribe to "AI with Shaily" for weekly thoughtful insights and share your thoughts in the comments—Shailendra loves engaging with his audience! 📱💬

Thanks for tuning in! Until next time, keep questioning, keep learning, and keep humanizing AI. This is Shailendra Kumar signing off from "AI with Shaily." 🙌✨

When AI Misreads Hate Speech: The Crucial Role of Human Judgment in Language Models

Middle East AI News Minute - an artificial intelligence news roundup by Carrington Malin - brought to you by Presight  - a leading global big data analytics company powered by AI. In this podcast: First Arabic language model added to Falcon series; Tarjama& nets $15 million Series A funding; G42 & WWT launch industry AI accelerator.

Sign-up for the weekly email news digest here:
https://www.middleeastainews.com

#MiddleEastAINews


First Arabic language model added to Falcon series

Welcome back to *AI with Shaily,* hosted by Shailendra Kumar, a lively and insightful show that brings you the latest and greatest in artificial intelligence with a personal touch! 🎙️🤖 Today’s episode dives deep into how Large Language Models (LLMs), like GPT, are revolutionizing traditional machine learning workflows in exciting new ways.

Shailendra uses a tasty cooking analogy 🍳 to explain the transformation: imagine you’ve always prepared every ingredient yourself, but now you have a smart kitchen assistant who not only helps prep but suggests new spices you never considered. That’s exactly what LLMs are doing for feature engineering — automating and enhancing the process of extracting meaningful features from complex data like text and images. This upgrade is like swapping a butter knife for a chef’s knife — sharper, more efficient, and precise. 🔪✨

The episode also highlights a hot topic buzzing on LinkedIn and Twitter: the merging of MLOps (machine learning operations) and LLMOps (operations focused on large language models). Teams are creating hybrid systems where traditional ML algorithms and LLMs coexist, deployed and monitored seamlessly. Frameworks like LangChain are gaining popularity, helping move projects from prototype to production at lightning speed. ⚙️🚀 Shailendra shares his excitement about this convergence, comparing it to his first experience automating an ML deployment — only now on a much grander scale!

Another fascinating point is how LLMs are not just analyzing data but actively augmenting the training process itself. They can generate synthetic data, provide feedback, and identify the most important samples to improve dataset quality — acting like a coach who trains and scouts the best plays on the field. 🏈📊 Plus, hybrid inference pipelines are emerging, where LLMs handle unstructured text with their natural language skills, while traditional ML methods process structured data. This combo is especially powerful in sensitive fields like healthcare and finance, where both numerical accuracy and subtle context matter deeply. 💉💰

Importantly, Shailendra emphasizes the growing trend of bringing humans back into the AI loop. This blend of AI power and human judgment ensures trustworthy, explainable outcomes. Imagine an LLM flagging a tricky prediction for a human expert’s review or generating simple explanations so users aren’t lost in technical jargon. It’s a perfect balance of machine efficiency and human insight — a crucial step toward responsible AI. 🤝🧠

Before signing off, Shailendra shares a valuable pro tip: invest time in prompt engineering tailored to your specific datasets. Fine-tuning the questions you ask LLMs can dramatically boost feature quality and overall model performance. It’s like the secret seasoning that turns a good dish into a wow-worthy meal! 🌶️✨

He leaves listeners with a thought-provoking question: How ready are you to blend these powerful language models into your existing ML setups? Are you prepared to let AI be your sous-chef in the data kitchen? 🍽️🤔

To inspire, Shailendra quotes AI pioneer Andrew Ng: “AI is the new electricity,” reminding us that this powerful force is now being wired directly into the data pipelines fueling our systems. ⚡🔌

Don’t miss out on more episodes of *AI with Shaily* available on YouTube, Twitter, LinkedIn, and Medium. Subscribe for your weekly dose of AI breakthroughs and share your thoughts to keep the conversation vibrant and exciting! 🌐📲

Till next time, keep experimenting, questioning, and AI-ing smartly! This is Shailendra Kumar signing off from *AI with Shaily.* 🎧🤓✨

LLMs as Your AI Sous-Chef: Revolutionizing Machine Learning Workflows with Hybrid Intelligence

Welcome to "AI with Shaily," hosted by Shailendra Kumar, where every week you get the freshest updates and insights on artificial intelligence 🤖✨. In this episode, Shaily dives deep into the transformative impact of Large Language Models (LLMs) on marketing and content creation, highlighting how these powerful AI tools are revolutionizing ad campaigns 🎯📈.

Imagine the hectic life of a marketing team under tight deadlines, juggling multiple campaigns and struggling to produce unique, engaging ad copy quickly. Shaily shares his own experience with the stress of last-minute edits and writer’s block 😰✍️. Enter LLMs—the tireless AI assistants that never make typos and can generate customized content in minutes, freeing marketers from creative bottlenecks 🚀📝.

LLMs empower marketers to produce large volumes of high-quality, personalized ads at lightning speed. This means teams can rapidly test different ad variations to find the messages that truly connect with their audience, resulting in better engagement and higher conversion rates 📊💡. Thanks to their deep understanding of context and language nuances, these AI models create copy that feels authentic and relevant, boosting customer trust and interest 💬❤️.

But the power of LLMs doesn’t end there. When paired with other AI tools like smart chatbots or content gap analyzers, the creative process becomes even more strategic and data-driven, offering insights that help tailor messaging perfectly to market demands 🤝🔍. For brands focused on consistency and compliance, AI platforms such as Acrolinx ensure all content remains on-brand, inclusive, and regulation-friendly, safeguarding reputation and coherence ✅🌍. Plus, LLMs can craft ads in multiple languages without losing the brand’s unique voice—ideal for global campaigns aiming to resonate locally 🌐🗣️.

Shaily’s bonus tip for marketers eager to embrace AI is to start small: use LLMs to generate alternative headlines or social media snippets. This low-risk approach helps teams experiment and build confidence in AI-assisted content creation step by step 🛠️📣.

He closes with an inspiring call to action: imagine what creative campaigns you could launch if you had an AI-powered writing assistant handling your content workload. The future of advertising is already here, and with LLMs, you have the tools to invent it yourself 🌟🚀.

Stay connected with Shailendra Kumar across YouTube, Twitter, LinkedIn, and Medium for more AI insights and stories. Don’t forget to subscribe to "AI with Shaily" for your weekly dose of AI breakthroughs, and join the conversation by sharing your thoughts in the comments 💬👍.

Until next time, keep innovating and questioning—the future is yours to create! 🌟🤖🎉

AI with Shaily: How Large Language Models Are Revolutionizing Ad Campaigns

We go over the summary judgment rulings in the recent lawsuits against Anthropic and Meta for their LLMs, and then we zoom out and consider whether AI is the technology that will finally unravel copyright in a way that requires a complete institutional rethinking.

S07E07 - Will AI Finally Kill Copyright?

This week’s guest is Nick Kolba, CEO and co-founder of Connectifi, who joins the podcast to discuss the Model Context Protocol, LLMs and agentic AI.  

0:00 – Tony and Shen talk about Astronomer’s ad campaign
5:30 – Nick Kolba joins the podcast 
8:00 – Step changes or big jumps in tech
18:00 – Excitement for the Model Context Protocol
21:00 – Parallels between FDC3 and MCP
24:30 – Concerns around security
30:00 – Work that still needs to be done
41:00 – Code generation for “senior engineers”


Episode 326: Connectifi’s Nick Kolba

The Middle East AI News Minute gives you a daily one-minute daily briefing on the top AI-related stories coming out of the Middle East. If you are a busy technology, business or government leader who want to get a quick update, then this is the podcast for you.

Unlock the secrets of data mastery with Shailendra, a 27-year veteran in AI, ML, and Data Science, driving AU$ 6 Billion in transformative value. As the author of 'Making Money out of Data,' their strategies redefine digital success. With expertise honed at IBM and Accenture, they drive innovation and data monetization. Explore how loyalty programs like Flybuys and Everyday Rewards revolutionize business—connect with Shaily today.

Wei-Shen Wong and Anthony Malakian discuss the hottest financial technology topics in the capital markets every week.

Software engineers, architects and team leads have found inspiration to drive change and innovation in their team by listening to the weekly InfoQ Podcast. They have received essential information that helped them validate their software development map. We have achieved that by interviewing some of the top CTOs, engineers and technology directors from companies like Uber, Netflix and more. Over 1,200,000 downloads in the last 3 years.

Hi I am Aygün. I glue AI with smart spaces.
I have completed my credits here and migrated to spotify.
https://open.spotify.com/show/5iNiriwBv8WnBQPGFh6B9F?si=AeQhTeZLTpygNwEW_tkQug

A podcast about geek culture..by lawyers. 

Can I patent a time machine? Who owns the copyright to the Klingon language? Can I sell my original story set in the Star Wars universe? We are here to discuss the many weird and interesting ways that geek culture collides with the law. @LGGpod

New tracks tagged #large language models