What's AI Podcast by Louis-François Bouchard

Deploy Your AI Agents 8x faster with LangWatch. Get a demo: https://langwatch.ai/?utm_source=louis-yt ► Master the most in-demand skill for building AI-powered solutions—from scratch: https://academy.towardsai.net/courses/python-for-genai?ref=1f9b29 ► Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Join Our AI Discord: https://discord.gg/learnaitogether

Have you ever wanted to take a language model and make it answer the way you want without needing a mountain of data? Well, OpenAI’s got something for us: Reinforcement Fine-Tuning, or RFT, and it changes how we customize AI models. Instead of retraining it with feeding examples of what we want and hoping it learns in the classical way, we actually teach it by rewarding correct answers and penalizing wrong ones, just like training a dog — but, you know, with fewer treats and more math. Let’s break down reinforcement fine-tuning compared to supervised fine-tuning! Both essentially have their use that we can discuss in one line: Supervised fine-tuning teaches new things the model does not know yet, like a new language, which is powerful for small and less “intelligent” models. While reinforcement fine-tuning orients the current model to what we really want it to say. It basically “aligns” the model to our needs, but we need an already powerful model. This is why reasoning models are a perfect fit. I’ve already covered fine-tuning on the channel if you are interested in that. Today, let’s get into how RFT actually works!

ChatGPT is completely changing how we learn programming. Instead of getting bogged down by coding theory, even beginners can jump right into building projects from day one. Quite the difference compared to university! With tools as simple as ChatGPT, you can experiment with building real applications right from the start quite easily without understanding much. This hands-on approach lets you learn by doing, offering instant feedback and a way to explore coding in a practical, exciting way. But there's a good and a wrong way to approach this. Relying solely on copy-pasting code won’t make you a programmer. When ChatGPT gives you a code snippet—say, a script that processes data or handles user login—use it as a starting point. TAKE THE TIME to UNDERSTAND why the code works, experiment with modifications, and see how changes affect the outcome. True mastery comes from engaging with the code, troubleshooting errors, and making it your own. If you can't explain anything, even if your app runs, it won't make you a better programmer or get you a good job. It will also have the downside of making a precarious app. You'll one day end up with too much code to follow what's happening, and ChatGPT will be stuck in an endless debugging loop. Yes, do embrace the power of AI to kickstart your projects, but just keep in mind that real growth (and value) happens when you do things and learn the logic behind every line. We've built a whole course about that principle to learn Python: https://academy.towardsai.net/courses/python-for-genai?ref=1f9b29

Here's an overview of the impact of LLMs on human work, which is complex and varied across different job categories...

Software engineers vs. ML engineers vs. prompt engineers vs. LLM developers... all explained The rise of LLMs isn’t just about technology; it’s also about people. To unlock their full potential, we need a workforce with new skills and roles. This includes LLM Developers, who bridge the gap between software development, machine learning engineering, and prompt engineering. Let’s compare these roles... Master, Use and Build with LLMs in this Programming Language Agnostic Course: https://academy.towardsai.net/courses/8-hour-genai-primer?ref=1f9b29 Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 Our ebook: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29 Episode 2/6 of the "From Beginner to Advanced LLM Developer" course by Towards AI (linked above). This course is specifically designed as a 1 day bootcamp for Software Professionals (language agnostic). It is an efficient introduction to the Generative AI field. We teach the core LLM skills and techniques together with practical tips. This will prepare you to either use LLMs via natural language or to explore documentation for LLM model platforms and frameworks in the programming language of your choice and start developing your own customised LLM projects.

What most people call agents aren’t agents. I’ve never really liked the term “agent”, until I saw this recent article by Anthropic, where I totally agree and now see how we can call something an agent. The vast majority is simply an API call to a language model. It’s this. A few lines of code and a prompt. This cannot act independently, make decisions or do anything. It simply replies to your users. Still, we call them agents. But this isn’t what we need. We need real agents, but what is a real agent? That what we dive in into this episode... Links; Anthropic’s blog on agents: https://www.anthropic.com/research/building-effective-agents Anthropic’s computer use: https://www.anthropic.com/news/3-5-models-and-computer-use Hamul Husain’s log on Devin: https://www.answer.ai/posts/2025-01-08-devin.html

In the early days of LLMs, context windows, which is what we send them as text, were small, often capped at just 4,000 tokens (or 3,000 words), making it impossible to load all relevant context. This limitation gave rise to approaches like Retrieval-Augmented Generation (RAG) in 2023, which dynamically fetches the necessary context. As LLMs evolved to support much larger context windows—up to 100k or even millions of tokens—new approaches like caching, or CAG, began to emerge, offering a true alternative to RAG... ►Full article and references: https://www.louisbouchard.ai/cag-vs-rag/ ►Build Your First Scalable Product with LLMs: https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev?ref=1f9b29 ►Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 ►Our ebook: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29 ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Join Our AI Discord: https://discord.gg/learnaitogether

I think the first though about LLMs and generative AI, is often, “Cool tech buzzwords, but do I really need to know this?” YES. Here’s why diving into LLMs is practically essential... 🚀 1. They transform how we work Think about all the repetitive, boring tasks in your day. You can (almost) automate them, building tools that make you 10x more productive. That’s what LLMs can do. If you can't, someone else can. If it's too complex, it will be possible soon. 🧠 2. Reaching their full potential isn’t automatic LLMs don’t come with a magic "win button," even if ChatGPT by itself is fantastic. To use them effectively, you’ve got to understand what they’re good at, what they’re not, and how to make them work for you by adding features. ⚠️ 3. Misuse = trouble LLMs can mess up big time without the right skills—wrong answers, misinformation, or just plain inefficiency. Learning how to avoid these pitfalls is critical. ✍️ 4. Prompts are everything Crafting clear, precise instructions is half the battle. A well-thought-out prompt can turn mediocre results into game-changing insights. It's just the basics of good, clear and concise communication. 🎯 5. Knowing when to use them is key Not every problem needs AI, but knowing where LLMs can deliver the biggest impact? That’s a game-changer. The right tool at the right time = massive efficiency gains. 🔒 6. Privacy matters more than ever LLMs can accidentally expose sensitive information if you’re not careful. Learning to use them responsibly isn’t optional—it’s a must. (Unless you want to be the person who accidentally leaks proprietary data) ⏳ 7. Don’t get left behind Those who embrace and learn these tools early are already gaining a competitive edge. The ones who resist? Well... let’s say the AI train is moving fast, and you don’t want to be stuck at the station. I know LLMs can feel intimidating at first, but the rewards are worth it. Whether you’re a developer, a business leader, or just someone curious about the future, learning how to use these tools is a skill that’ll pay off in ways you can’t even imagine yet.

When we talk about building powerful machine learning solutions, like large language models or retrieval-augmented generation, one key element that often flies under the radar is how to connect all the data and models and deploy them in a real product. This is where APIs come in. In this one, we’re diving into the world of APIs — what they are, why you might need one, and what deployment options are available. Build Your First Scalable Product with LLMs: https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev?ref=1f9b29 Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 Our ebook: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29

Build Your First Scalable Product with LLMs: https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev?ref=1f9b29 Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 Our ebook: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29 Video 8/10 of the "From Beginner to Advanced LLM Developer" course by Towards AI (linked above). The most practical and in-depth LLM Developer course out there (~90 lessons) for software developers, machine learning engineers, data scientists, aspiring founders or AI/Computer Science students. We’ve gathered everything we worked on building products and AI systems and put them into one super practical industry-focused course. Right now, this means working with Python, OpenAI, Llama 3, Gemini, Perplexity, LlamaIndex, Gradio, and many other amazing tools (we are unaffiliated and will introduce all the best LLM tool options). It also means learning many new non-technical skills and habits unique to the world of LLMs. Learn more for free... Twitter: https://x.com/Whats_AI Substack (newsletter): https://louisbouchard.substack.com/

In this one, I discuss the dilemma between using retrieval-based generation and the newer "long context models". Long context models, like the Gemini suite of models, allow us to send up to millions of tokens (thousands of text pages), whereas retrieval (RAG)-based systems allow us to search through as much (if not more) content and retrieve only the necessary bits to send the LLM for improved answers. Both have advantages and disadvantages. This short episode will help you better understand when to use each. Build Your First Scalable Product with LLMs: https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev?ref=1f9b29 Master LLMs and Get Industry-ready Now: https://academy.towardsai.net/?ref=1f9b29 Our ebook: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29

► Get your copy of "Building LLMs for Production": https://amzn.to/4bqYU9b ►The e-book version: https://academy.towardsai.net/courses/buildingllmsforproduction?ref=1f9b29 ► Our new course "From Beginners to Advanced LLM Developer": https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev?ref=1f9b29 ►Full article and references: https://www.louisbouchard.ai/openai-o1/ ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Join Our AI Discord: https://discord.gg/learnaitogether Extra Ressources: OpenAI release blog: https://openai.com/index/introducing-openai-o1-preview/ OpenAI release blog 2: https://openai.com/index/learning-to-reason-with-llms/ OpenAI system card: https://openai.com/index/openai-o1-system-card/ Nathan Lambert’s great article on it: https://www.interconnects.ai/p/openai-strawberry-and-inference-scaling-laws David Shapiro fun livestream testing it: https://youtu.be/AO7mXa8BUWk How to start in AI/ML - A Complete Guide: ►https://www.louisbouchard.ai/learnai/ #gpt4o #o1 #openai

In this episode, Luis Serrano and I dive into the transformative impact of AI on education, forecasting a radical shift in how future generations learn and think. ► Luis' website: https://serrano.academy/ ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether How to start in AI/ML - A Complete Guide: ►https://www.louisbouchard.ai/learnai/ Chapters: 00:00 Coming up in the conversation 00:01:50 Sharing journey: Why Luis became an educator 00:06:03 Can someone develop skills to become a better educator, and what are they? 00:08:07 Deciding the depth of explanation 00:10:57 AI’s impact on education 00:22:35 How does an explanation without graphic aid look? 00:27:15 Luis is explaining embedding in an intuitive way? 00:31:05 Is AI hard to explain because of newness or complexity? 00:34:01 Necessity of understanding the basics of AI 00:36:57 Why do people not want to learn about how AI works? 00:39:15 Importance of good story telling and explanation 00:42:01 Strategy to explain tough topics 00:48:12 Strategy to introduce complex words in explanation 00:55:14 Evolution in AI Education Approaches 01:02:03 Is it possible to bring good value through shorts or reels? 01:04:46 Rise of Podcast and reels

Register to GTC (attend in person, or free online): https://nvda.ws/3XQRtkl Interested in end-to-end PM job hunting and up-skilling program by Dr. Nancy Li’s PM Accelerator? Register this free masterclass about product portfolio and stay until the end to learn more about the program (Use the code LOUIS500 for 500$ off on her program!): https://www.drnancyli.com/a/2147615411/2HzsofFw Introducing Dr. Nancy Li, a versatile entrepreneur, Director of Products, YouTuber, and a Forbes-featured professional with 8 years of experience in driving cutting-edge technology products. Dr. Li currently serves as the CEO of PM Accelerator, the fastest-growing Product Management Professional Development Company in the industry, known for its engaging alumni network, and top-rated program, and she has a remarkable record of helping over 1000 aspiring product managers secure high-paying roles at tech giants and unicorn startups. Her journey, from being the youngest engineering Ph.D. to Director of Product in just four years, is a testament to her extraordinary career. Having personally launched award-winning AI products and mentored many into high-paying AI PM roles, Dr. Nancy offers a rare blend of expertise and experience. From her day-to-day interactions with AI engineers to the challenges of training AI models, she provides a comprehensive look into the dynamic world of AI product management. References we discussed in the episode: PM Accelerator by Dr. Nancy Li: https://www.drnancyli.com The ONLY 4 Ways to Become an AI Product Manager with No Experience: https://youtu.be/aQTuPUIkrxk?si=JJMih2qzC6iP2a8_ A Day in The Life of An AI Product Manager: https://youtu.be/waVyVcUzfeg?si=YOqUao6HCSHQ9MWG ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether 00:00:00 Coming up in the conversation 00:02:46 Nancy introduces herself 00:04:02 The reason Nancy couldn't drop her PhD 00:07:35 These are the people PhD is for 00:09:40 Secret revealed: How Nancy completed her PhD in 3.5 years! 00:14:07 Tips that helped Nancy peer with people from MIT 00:23:25 Are companies still prioritizing titles over practical skills? 00:26:21 Have PM skill requirements changed in recent years? 00:29:20 Crazy story: This is why she will never go to university to teach! 00:35:53 Online education vs offline education 00:41:29 Shifting from Material to AI: How she Landed a Job! 00:44:32 Staying up-to-date with technology and deciding when to implement which 00:46:41 Secret recipe to make successful AI products 00:51:19 Day to day life of a PM 00:55:28 Louis shares about his start-up Towards AI 00:58:21 Nancy shares information about her PM accelerator program

In this episode, I talk with Avery Smith, a data analytics expert and educator who gives practical strategies for breaking into the data analytics field, leveraging AI for learning and career development. Avery shares his journey into data and teaching, and insights on helping others transition into data careers through his Data Analytics Accelerator program, emphasizing the importance of practical projects and how he leverages AI in enhancing learning and job preparation processes (and he shares tips to help you do that too!). References: ►Avery Smith: https://www.linkedin.com/in/averyjsmith/ ►Data Career Jumpstart: https://www.datacareerjumpstart.com/ ►Podcast: https://podcasters.spotify.com/pod/show/datacareerpodcast ►AveryGPT: https://www.datacareerjumpstart.com/averygpt ►AI Interview Simulator: https://www.datacareerjumpstart.com/interviewsimulator ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether How to start in AI/ML - A Complete Guide: ►https://www.louisbouchard.ai/learnai/ Timestamps: 00:00 Coming up in the conversation 01:45 Avery shares about his background 03:00 Making people land data job in 90 days! 07:02 Theory vs Practical knowledge 08:34 Importance of Explainability in Models 10:28 The Future of Traditional and Online Education 12:00 Networking while studying remotely 14:09 Maintaining consistency in value in LinkedIn posts. 16:20 Is greater studies still relevant in the era of ChatGPT? 17:45 Becoming freelancing ready in data analytics 20:53 Keeping course content up to date 23:56 This is how Avery utilizes AI 29:16 Discussion on AI Avatars 38:01 Does Avery provide lessons on how to better use ChatGPT? 40:08 Avery shares his learning resources 43:12 Book recommendations 44:52 Is the field of data field too saturated to join right now? 46:58 Discussion on the current reality of freelancing

In this episode I had the opportunity to talk with Tina Huang, founder of the Lonely Octopus platform, a highly successful YouTube channel and experienced freelancer in the AI space. Tina shares her invaluable insights on leveraging AI in education, the nuances of freelancing in the tech industry, and strategies for enhancing personal productivity. The episode is for anyone looking to navigate the landscape of technology (especially AI), offering practical tips to work in the field or just leverage AI better. ►Check out Tina's channel @TinaHuang1 ►Lonely Octopus: https://www.lonelyoctopus.com/ ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether Timestamps: 00:00 Coming up in the conversation 02:00 How did Tina get into AI and YouTube? 03:17 Tina's goal and mission 04:09 Tina’s niche 06:40 Higher education in the AI and data science space 10:36 Tips for beginners to become freelancing-ready 17:24 What will be more important in the future, LLMs or coding languages? 22:30 Tips for those who want to change field while balancing their current job 25:16 Using YouTube to force ownself to learn 27:17 How to make commitments and what kind of commitments should you have? 33:05 Louis shares about the AI market he believes has the most potential 37:44 Tina discussed where she wants to contribute more 39:09 Tine shares the benefits that her YouTube venture has brought 40:40 How can one use content to create leverage in freelancing? 43:05 Is audience conversion from shorts to long-form content really an issue? 46:46 Freelancing vs corporate employment vs entrepreneurship 50:33 What skills should one develop to secure freelance opportunities in the field of AI? 54:00 Tina shares about her upcoming plans

In this episode, I received Mariam Brian, CEO of Holo Art, to talk about the transformative role of AI in the art world. She discusses how artificial intelligence is reshaping artistic creation and expression and addresses the ethical implications of this technological evolution. This conversation, accessible to anyone, offers a fantastic perspective on the intersection of art and AI, highlighting the potential for a new era of creativity and collaboration between humans and machines! ►Mariam's LinkedIn: https://www.linkedin.com/in/mariamhashemi/►Holo Art: https://holo-art.io/about-us ► Holo Art announcement: https://medium.com/@mariambrian/patented-ai-process-for-executives-organizations-looking-to-level-up-e465c1c35a07 ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether Timestamps: 00:00:00 Coming up in the conversation 00:01:32 Mariam shares about his background 00:02:15 The Intersection of AI and Philosophy 00:05:39 The Impact of AI on Art and Artists 00:08:36 The Future of AI and Art 00:09:13 The Role of AI in Business and Ethics 00:10:55 AI might the Pandora box of lot of problems! 00:14:39 Simultaneous rise of Podcast & Shorts and their impact on the lives of billions 00:23:42 The Creativity of AI and its Impact on Artists 00:28:53 Can AI generated art hurt creativity of artist? 00:33:22 To be an artist, ethics becomes a way of life 00:35:45 Mariam's Personal Use of AI in Art 00:40:30 AI's Potential in Human-Machine Co-Creation 00:41:27 Understanding Ourselves and AI's Perception of Us 00:46:32 W.I.E.R.D Science 00:50:02 While using AI model do you try to control it or let it surprise you? 00:54:38 Public Perception of AI-Generated Art 01:01:44 The Risks and Opportunities for Artists Using AI 01:10:53 Mariam's message for listeners

A new episode with Jerome Pasquero, a Machine Learning Director at Sama, a leading company for data annotation solutions, where we dive into the role of data in AI's evolution. We explore the nuances of data annotation, the ethical implications of data in AI, and how data is shaping the future of technology. Don't miss Jerome Pasquero's insights on the intersection of data and AI! ►Jerome Pasquero: https://www.linkedin.com/in/jeromepasquero/ ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether Timestamps: 00:00:00 Coming up in the conversation 00:01:34 Jerome shares about his background 00:04:07 How did Jerome get into the data field? 00:05:23 AI back in the days of 2000s 00:07:20 Back then, what piqued Jerome's interest the most in AI? 00:08:40 Using AI to try to mimic human comprehension 00:12:47 Present challenges and the prospective outlook of computer vision 00:14:54 Using Humans vs. ML Models to Annotate Data 00:17:46 Jerome's perspective on Constitutional AI or RLAIF 00:24:52 Impact of LLM and AI on the Job market 00:26:27 Is the AI revolution bigger than previous tech revolutions? 00:28:35 Will there be something more interesting than AGI? 00:31:15 Dealing with complex annotation tasks and different perspectives 00:33:33 Dealing with biases 00:36:18 Using a single annotator vs. multiple annotators on the same data 00:37:49 Synthetically generated data 00:40:47 Scaling quality assurance for large datasets 00:42:46 When is machine learning better at annotation than human annotators? 00:45:34 Reduction of Humans-in-the-loop due to the constant evolution of AI 00:46:42 Data Requirements for Training Autonomous Vehicles 00:51:43 Sensors for transferring human driving skills to Autonomous cars 00:53:20 Why don’t we build only autonomous subway system? 00:55:26 Use of AI in the vision industry and example of vision technology used in our daily life 01:00:17 The potential of haptics and its link with AI

►Think Autonomous: https://www.thinkautonomous.ai/ ►Jeremy’s linkedin: https://www.linkedin.com/in/jeremycohen2626/ ►Newsletter: https://www.thinkautonomous.ai/private-emails-home/ ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether How to start in AI/ML - A Complete Guide: ►https://www.louisbouchard.ai/learnai/ Become a member of the YouTube community, support my work and get a cool Discord role : https://www.youtube.com/channel/UCUzGQrN-lyyc0BWTYoJM_Sg/join Chapters: 0:00 Hey! Tap the Thumbs Up button and Subscribe. You'll learn a lot of cool stuff, I promise. 00:01:31 Jeremy shares about his background 00:03:55 The future of Self-driving cars is not that straightforward! 00:07:49 If there are numerous challenges, why are companies still developing autonomous cars? 00:08:46 The future might involve more self-driving buses and trucks instead of cars 00:09:24 Are AI Start-ups dead? 00:14:25 Can 'wannabe' AI start-ups harm the actual AI economy and market? 00:16:50 Should AI research prioritize progress over control and a deep understanding of algorithms? 00:21:14 Can AI replace experts? 00:24:04 If we make AI hyper-personalized or make it impersonate someone, can it replace experts? 00:25:18 If AI cannot replace experts, should we be worried about our jobs? 00:26:18 How to find out if your job can be taken by AI or not? 00:33:06 Potential of AI in creative expression, entertainment, and journalism 00:38:11 Will AI make us dumb? 00:40:16 Can hallucination be fixed or not? 00:46:29 Is it possible to build a biasless AI model? 00:52:24 Transparency is going to be a big thing AI economy 00:55:59 AI can make your content boring! 01:02:39 Your mom might not use AI unless this happens! 01:09:08 Is AI democratizing opportunities or is it still only benefitting the rich?

Follow the podcast for more interesting conversations with experts in the AI space! For more: ►Twitter: https://twitter.com/Whats_AI ►My Newsletter (My AI updates and news clearly explained): https://louisbouchard.substack.com/