In this week’s eclectic roundup, we delve into the art of showing appreciation at work, uncover the secrets behind HubSpot’s $30B success story, and sniff out the latest research on the science of smell. Also on the agenda: exploring Pinterest’s innovative Text-to-SQL project, highlighting the dark world of HTML email security, dissecting credit card rewards programs, and much more. Whether you’re a tech aficionado, a future-focused professional, or simply curious, there’s something here for everyone. Join me as we navigate through fascinating insights and discoveries from various corners of the digital world!

• Simple Ways to Show Appreciation at Work: Appreciation, like trust, is relationship-based — each interaction we have with someone either strengthens or weakens that invisible connection. The more we feel appreciated, the stronger those bonds become, and the more tension they can withstand when something challenges it. Consequently, knowing how to build and maintain relationships where people feel appreciated is a foundational skill — one that’s important to learn from the very early stages of your career. Appreciate people’s presence: People need to know that others care that they are there. Letting someone know their presence alone is having an impact on you or the organization can make a big difference. Notice when people are missing at work and reach out to make sure they are doing fine. Even if it is people’s job to be there, thank them for coming. Appreciate their ideas and contribution. Managers and individual contributors alike need to show colleagues that their input is welcome and celebrated. Honor people’s ideas and expertise by talking them up in senior leader meetings. When people share input or feedback on a project, whether you like their idea or not, actively listen. Appreciate people’s lives outside of work. Our jobs should help us live our best lives, not take them over. For people to feel appreciated, they need to know that we understand their passions, priorities, and responsibilities outside of work. They need leaders who respect boundaries and role model self-care, and coworkers who nurture a culture of support. Appreciate people’s need for growth and development. While many people think development is about promotions and attending trainings, it is also about being around people who challenge us — managers that take time to understand and support our career goals, and coworkers who help us learn and grow. If you don’t already know the career goals and aspirations of your direct reports, take time to understand them. Ask your coworkers for feedback and tips on how you can improve and invite them to do the same with you.
• Zigging vs. zagging: How HubSpot built a $30B company | Dharmesh Shah (co-founder/CTO): Dharmesh Shah is the co-founder and CTO of HubSpot (currently valued at $30 billion) and one of the most fascinating founders I’ve ever met. Dharmesh is the keeper of HubSpot’s Culture Code, built ChatSpot (an AI chatbot built on top of HubSpot CRM) and a game called WordPlay (which grew to 16 million users), and also founded and writes for OnStartups, a top-ranking startup blog and community with more than 1M members. He’s also invested in 100+ startups including OpenAI, AngelList, Coinbase, and Dropbox.
• The science of smell is fragrant with submolecules: A chemical that we smell may be a composite of multiple smell-making pieces.
• How we built Text-to-SQL at Pinterest: Adam Obeng | Data Scientist, Data Platform Science; J.C. Zhong | Tech Lead, Analytics Platform; Charlie Gu | Sr. Manager, Engineering Writing queries to solve analytical problems is the core task for…
• Kobold letters: Anyone who has had to deal with HTML emails on a technical level has probably reached the point where they wanted to quit their job or just set fire to all the mail clients due to their inconsistent implementations. But HTML emails are not just a source of frustration, they can also be a serious security risk.
• Anatomy of a credit card rewards program: Credit card rewards are mostly funded out of interchange, a fee paid by businesses to accept cards.
• JetMoE:
• The Design Philosophy of Great Tables:
• The hardest part of building software is not coding, it’s requirements:
• Building LLMs for Code Repair: Introduction At Replit, we are rethinking the developer experience with AI as a first-class citizen of the development environment. Towards this vision, we are tightly integrating AI tools with our IDE. Currently, LLMs specialized for programming are trained with a mixture of source code and relevant natural languages, such as GitHub issues and StackExchange posts. These models are not trained to interact directly with the development environment and, therefore, have limited ability to understand events or use tools within Replit. We believe that by training models native to Replit, we can create more powerful AI tools for developers. A simple example of a Replit-native model takes a session event as input and returns a well-defined response. We set out to identify a scenario where we could develop a model that could also become a useful tool for our current developers and settled on code repair. Developers spend a significant fraction of their time fixing bugs in software. In 2018, when Microsoft released “A Common Protocol for Languages,” Replit began supporting the Language Server Protocol. Since then, the LSP has helped millions using Replit to find errors in their code. This puts LSP diagnostics among our most common events, with hundreds of millions per day. However, while the LSP identifies errors, it can only provide fixes in limited cases. In fact, only 10% of LSP diagnostic messages in Python projects on Replit have associated fixes. Given the abundance of training data, repairing code errors using LSP diagnostics is therefore the ideal setting to build our first Replit-native AI model. Methodology Data
• A little guide to building Large Language Models in 2024: A little guide through all you need to know to train a good performance large language model in 2024.
  This is an introduction talk with link to references for further reading.
• MVC vs MVP Architecture: What’s the Difference?: Are you a Product Manager looking to navigate the complex landscape of software development with confidence and creativity? The choice of architectural patterns — Model-View-Controller (MVC) and…
• Biases in Expected Goals Models Confound Finishing Ability: The assessment of finishing skill in soccer using xG remains contentious due to players’ difficulty in consistently outperforming their cumulative xG. We attempt to pin down the limitations and nuances and borrow techniques from AI fairness to address them.
• Where Graph Theory Meets The Road: The Algorithms Behind Route Planning: Back in the hazy olden days of the pre-2000s, navigating between two locations generally required someone to whip out a paper map and painstakingly figure out the most optimal route between those d…
• Death is a Feature: When Parisians got tired of cemeteries during the French Revolution, they conscripted priests to relocate bones of more than six million deceased forebears to empty limestone quarries below the cit…
• Just for Fun: A Five-Card Poker Library Using C#: Chances are if you’ve had many coding interviews you’ve been presented with a poker problem. Here’s a great take from Dr. James McCaffrey of Microsoft Research.
• Where Did My Traffic Go? Hint: It Wasn’t an Algorithm Update: