Logo
    Search

    Podcast Summary

    • Shopify's Busy Week Before Black Friday and Cyber MondayShopify powers a significant portion of holiday shopping and uses AI to optimize performance during the high-traffic season. Intel's Advent of Gen AI hackathon offers a chance to learn about generative AI using Intel hardware and the latest LLMs.

      The week leading up to Black Friday and Cyber Monday at Shopify is a busy time with teams preparing for the biggest shopping season of the year. Shopify powers a significant portion of holiday shopping, and the live globe, a 3D visualization of live orders happening around the globe, is an annual feature that showcases this activity. During this period, Shopify also focuses on AI integration. Russ Maschmeyer, Shopify's project lead for Spatial Commerce and Magic Labs, shared insights into how Shopify has evolved over the years, supporting both small and large brands. He also discussed the importance of AI in optimizing Shopify's performance during the high-traffic holiday shopping season. Additionally, Intel is hosting a hackathon, Advent of Gen AI, in December 2023, offering participants access to Intel hardware and the latest open-access LLMs via Prediction Guard. This event is an excellent opportunity for individuals to learn about generative AI and showcase their skills.

    • AI Transforming Ecommerce: Tools for MerchantsAI streamlines tasks for merchants, enabling them to save time and focus on creative aspects of their business, from generating product descriptions to optimizing imagery and website layout.

      AI is significantly transforming the ecommerce landscape by accelerating tasks for merchants and enhancing the overall shopping experience. Shopify, an early adopter of AI, has seen the most impact in tools for merchants, including generating product descriptions, optimizing product imagery, and streamlining website layout. These AI-powered solutions enable merchants to save time and focus on creative aspects of their business. The impact of AI extends beyond product development, as it also influences web content development and advertising. Merchants are increasingly handing off repetitive tasks to AI, freeing up time for more strategic work. The result is a more efficient and effective ecommerce ecosystem.

    • Streamlining e-commerce with AI toolsShopify's AI-powered tools, called Shopify Magic, help merchants save time on tasks like writing product descriptions and creating email subject lines, allowing them to focus on growing their businesses.

      Shopify is utilizing AI technology to help merchants save time and effort on undesirable tasks, such as writing product descriptions or creating email subject lines. By providing easy-to-use tools for generating high-quality content, Shopify aims to give merchants back the time they would have spent on these tasks, allowing them to focus on other aspects of their business. This AI-enabled suite of tools, called Shopify Magic, covers various areas including email marketing, blog content, and product descriptions, as well as image generation. For new merchants, these tools are integrated seamlessly into the Shopify admin, making the process user-friendly and efficient. The impact is immediate, with tools like auto-generated product descriptions helping merchants get started without feeling overwhelmed. Overall, Shopify's focus on AI technology is aimed at streamlining the e-commerce experience for merchants, enabling them to save time and focus on growing their businesses.

    • Shopify's user-friendly features for new merchantsShopify simplifies e-commerce for beginners with AI-generated product descriptions and intuitive UI

      Shopify is making it easier for new merchants to build and customize their online stores with user-friendly features, such as drag-and-drop image and 3D model uploads, and AI-generated product descriptions. These tools help merchants create appealing product detail pages without needing extensive copywriting skills or spending hours researching best practices. Shopify's seamless integration of AI into their platform is an exciting step towards empowering merchants to create engaging webstorefronts, while also raising questions about the role of AI in our lives. The Trace Route Podcast's new season explores these themes, delving into the human and technological aspects of AI and its impact on our world. By combining AI with intuitive UI, Shopify is paving the way for a more accessible and efficient e-commerce experience for new merchants.

    • Revolutionizing product photography in ecommerce with AIAI tool generates high-quality, consistent product images for ecommerce merchants, enabling quick adaptation to changing tastes, seasons, and commerce trends

      The motivation behind the project is to revolutionize product photography in ecommerce by using AI to generate high-quality, consistent images that merchants can use across various platforms. Product photography plays a crucial role in online commerce as it helps merchants showcase their products in an appealing way and drive sales. With the emergence of open source image models like Stable Diffusion, there's an opportunity for merchants to be more agile and cost-efficient in creating product images. The ultimate goal is to create a tool that can recreate a product in high fidelity in any scenario, allowing merchants to adapt to changing tastes, seasons, and commerce trends with ease. However, early attempts at using AI to generate product images have faced challenges, such as a disjointed appearance between the product and its background, due to the lack of communication between the pixels of the product and the pixels of the environment. The team is working on solving this issue to create more realistic and seamless product images. The potential impact of this technology on commerce is massive, as it could enable merchants to quickly generate a wide range of product images for various use cases, making their businesses more adaptable and competitive.

    • Shopify uses open-source generative models to enhance product imagesShopify's team leverages open-source Stable Diffusion XL for visually grounded product images, contributing to the community and accelerating development.

      Shopify's team identified an opportunity to bring more "magic" to merchants in the form of visually grounded product images, even before achieving perfect personalization. They tackled this issue by exploring open-source generative models, specifically focusing on Stable Diffusion XL and using Comfy UI as a tool. The team, which values open source and contributes to the community, was able to move faster and collaborate effectively using these resources. While this is an early field, Shopify aims to bring the best technology to merchants, regardless of whether it's open source or not. They've had a positive experience with open source image generation models, which have helped them address grounding issues and accelerate their development process.

    • Exploring AI technologies with a dedicated teamHaving a dedicated team to experiment with AI technologies can lead to time savings and innovative solutions for businesses.

      For businesses looking to adopt new AI technologies, having a dedicated team to explore and experiment with these capabilities can lead to significant time savings and innovative solutions. The example given is Shopify's experience with Comfy UI and its use in generating product descriptions for merchants. By identifying the technology's strengths and the maximum value it could bring to merchants, Shopify was able to quickly develop and launch this feature. The Magic Labs team played a crucial role in this process by focusing on areas where AI technologies could make the most impact and accelerate merchant activities. As the AI landscape continues to evolve, having a team dedicated to staying informed and exploring new possibilities can help businesses stay competitive and provide the best possible platform for their customers.

    • Exploring new technologies for commerceMagic Labs uses a three-week cycle to evaluate new tech for commerce and merchants, prototyping quickly to determine feasibility and potential applications. They identified a challenge for merchants in grounding their product images using AI and used Shopify Fusion to effectively transform existing media into brand-representative images.

      The team at Magic Labs continually explores new technologies in three-week cycles to find potential impacts on commerce and merchants. They approach this process with curiosity and flexibility, prototyping quickly to determine feasibility and potential applications. For merchants starting out with limited resources, grounding their product images using AI can be a significant challenge. The team identified this problem as merchants having decent images but not ones that fully represent their brand. They considered a range of solutions, from improving prompts to retraining models, and landed on using Shopify Fusion to address the issue effectively. This approach allows merchants to transform their existing media into media that better represents their brand, unlocking the potential of their products in the competitive marketplace.

    • Using AI to create enhanced product backgroundsMerchants can use AI to create new, realistic backgrounds for their product photos, allowing them to unlock value from their existing image media and create visually appealing product images.

      Merchants can enhance their existing product photos by using AI technology to create new, elevated backgrounds, while keeping the product pixels and details intact. However, creating realistic shadows, ground reflections, and maintaining the correct camera angles proved to be initial challenges. To overcome these obstacles, a multivariate approach was taken, focusing on providing clear product and grounding descriptions in prompts. By starting with a declaration of the foreground object and its placement, the AI model can generate more accurate and realistic backgrounds. This approach allows merchants to unlock value from their existing image media and create visually appealing product images, even without a professional photography setup.

    • Using masked depth maps for more accurate Stable Diffusion imagesProviding a masked depth map with gradient info as context can lead to more accurate and high-fidelity generated images using Stable Diffusion models.

      The use of a masked depth map with a little extra depth information as context for Stable Diffusion models can lead to more accurate and high-fidelity generated images. This approach works by delivering a masked depth map of the original product to the model, which includes a gradient of depth info as you leave the product pixels. This additional context allows the model to understand the scene around the product, including shadows, angles, and camera position, resulting in a grounded and realistic image. This hack is an example of creative problem-solving and using existing models in new and powerful ways. It didn't require retraining the whole model but instead leveraged the capabilities of Stable Diffusion and Control Net to achieve impressive results. The idea likely came from a team of brilliant individuals who had a deep understanding of these models and the freedom to experiment and explore their potential. This environment of collaboration, creativity, and continuous learning is essential for pushing the boundaries of what's possible with AI models. If you're interested in trying this approach, start by understanding the input requirements of your chosen model and experimenting with different depth maps and contextual information. Remember, the key is to provide enough context for the model to understand the scene around the product, allowing it to generate more accurate and high-fidelity images.

    • The Future of Technology in CommerceAI will revolutionize online shopping with personalized, interactive experiences and virtual product visualization

      The intersection of technology and commerce has always driven innovation and cultural shifts. The speaker's experience of rapid iteration with AI machines and open-source tools in the early days of web development unlocked new possibilities for building complex machines. Now, they're excited about the future of technology, particularly AI, in commerce. They envision a future where online shopping experiences are more personalized and interactive, allowing customers to visualize products on themselves and receive recommendations based on their past purchases. This merging of technology and commerce continues to evolve and shape our culture.

    • Revolutionizing Shopping with AIAI technology is transforming shopping, providing personalized experiences for merchants and customers, and bringing one-on-one interactions to a wider audience, especially during major shopping events

      AI technology is revolutionizing the shopping experience for merchants and customers alike, enabling personalized and customized interactions through AI-driven solutions. Russ Frushtick from Shopify discussed the exciting potential of this technology, which will bring one-on-one shopping experiences to a wider audience. This development is particularly noteworthy as we approach major shopping events like Black Friday and Cyber Monday. The work being done by Russ and his team at Shopify is highly anticipated, and listeners are encouraged to subscribe to Practical AI and share the podcast with others to stay informed about these advancements. Fastly and Fly continue to partner with the show to bring listeners all the latest tech news. Brakemaster Cylinder provides the best beats for the show, and we'll be back next time to discuss more practical applications of AI.

    Recent Episodes from Practical AI: Machine Learning, Data Science

    Apple Intelligence & Advanced RAG

    Apple Intelligence & Advanced RAG
    Daniel & Chris engage in an impromptu discussion of the state of AI in the enterprise. Then they dive into the recent Apple Intelligence announcement to explore its implications. Finally, Daniel leads a deep dive into a new topic - Advanced RAG - covering everything you need to know to be practical & productive.

    The perplexities of information retrieval

    The perplexities of information retrieval
    Daniel & Chris sit down with Denis Yarats, Co-founder & CTO at Perplexity, to discuss Perplexity’s sophisticated AI-driven answer engine. Denis outlines some of the deficiencies in search engines, and how Perplexity’s approach to information retrieval improves on traditional search engine systems, with a focus on accuracy and validation of the information provided.

    Using edge models to find sensitive data

    Using edge models to find sensitive data
    We’ve all heard about breaches of privacy and leaks of private health information (PHI). For healthcare providers and those storing this data, knowing where all the sensitive data is stored is non-trivial. Ramin, from Tausight, joins us to discuss how they have deploy edge AI models to help company search through billions of records for PHI.

    Rise of the AI PC & local LLMs

    Rise of the AI PC & local LLMs
    We’ve seen a rise in interest recently and a number of major announcements related to local LLMs and AI PCs. NVIDIA, Apple, and Intel are getting into this along with models like the Phi family from Microsoft. In this episode, we dig into local AI tooling, frameworks, and optimizations to help you navigate this AI niche, and we talk about how this might impact AI adoption in the longer term.

    AI in the U.S. Congress

    AI in the U.S. Congress
    At the age of 72, U.S. Representative Don Beyer of Virginia enrolled at GMU to pursue a Master’s degree in C.S. with a concentration in Machine Learning. Rep. Beyer is Vice Chair of the bipartisan Artificial Intelligence Caucus & Vice Chair of the NDC’s AI Working Group. He is the author of the AI Foundation Model Transparency Act & a lead cosponsor of the CREATE AI Act, the Federal Artificial Intelligence Risk Management Act & the Artificial Intelligence Environmental Impacts Act. We hope you tune into this inspiring, nonpartisan conversation with Rep. Beyer about his decision to dive into the deep end of the AI pool & his leadership in bringing that expertise to Capitol Hill.

    Full-stack approach for effective AI agents

    Full-stack approach for effective AI agents
    There’s a lot of hype about AI agents right now, but developing robust agents isn’t yet a reality in general. Imbue is leading the way towards more robust agents by taking a full-stack approach; from hardware innovations through to user interface. In this episode, Josh, Imbue’s CTO, tell us more about their approach and some of what they have learned along the way.

    Private, open source chat UIs

    Private, open source chat UIs
    We recently gathered some Practical AI listeners for a live webinar with Danny from LibreChat to discuss the future of private, open source chat UIs. During the discussion we hear about the motivations behind LibreChat, why enterprise users are hosting their own chat UIs, and how Danny (and the LibreChat community) is creating amazing features (like RAG and plugins).

    Mamba & Jamba

    Mamba & Jamba
    First there was Mamba… now there is Jamba from AI21. This is a model that combines the best non-transformer goodness of Mamba with good ‘ol attention layers. This results in a highly performant and efficient model that AI21 has open sourced! We hear all about it (along with a variety of other LLM things) from AI21’s co-founder Yoav.

    Related Episodes

    When data leakage turns into a flood of trouble

    When data leakage turns into a flood of trouble
    Rajiv Shah teaches Daniel and Chris about data leakage, and its major impact upon machine learning models. It’s the kind of topic that we don’t often think about, but which can ruin our results. Raj discusses how to use activation maps and image embedding to find leakage, so that leaking information in our test set does not find its way into our training set.

    Stable Diffusion (Practical AI #193)

    Stable Diffusion (Practical AI #193)
    The new stable diffusion model is everywhere! Of course you can use this model to quickly and easily create amazing, dream-like images to post on twitter, reddit, discord, etc., but this technology is also poised to be used in very pragmatic ways across industry. In this episode, Chris and Daniel take a deep dive into all things stable diffusion. They discuss the motivations for the work, the model architecture, and the differences between this model and other related releases (e.g., DALL·E 2). (Image from stability.ai)

    AlphaFold is revolutionizing biology

    AlphaFold is revolutionizing biology
    AlphaFold is an AI system developed by DeepMind that predicts a protein’s 3D structure from its amino acid sequence. It regularly achieves accuracy competitive with experiment, and is accelerating research in nearly every field of biology. Daniel and Chris delve into protein folding, and explore the implications of this revolutionary and hugely impactful application of AI.

    Zero-shot multitask learning (Practical AI #158)

    Zero-shot multitask learning (Practical AI #158)
    In this Fully-Connected episode, Daniel and Chris ponder whether in-person AI conferences are on the verge of making a post-pandemic comeback. Then on to BigScience from Hugging Face, a year-long research workshop on large multilingual models and datasets. Specifically they dive into the T0, a series of natural language processing (NLP) AI models specifically trained for researching zero-shot multitask learning. Daniel provides a brief tour of the possible with the T0 family. They finish up with a couple of new learning resources.