Search the archive

BUILDING 0 TO 1 PRODUCTS AND THE INSIDE STORY OF FIGMA SLIDES (56 MINUTE PODCAST)

Mihika Kapoor, product lead at Figma, recognized the need to combine FigJam's interactive feedback with Figma's visual fidelity, which she initially pitched at a PM offsite in 2022. Through prototyping, hackathons, and persistence, she gained internal support, eventually leading to Figma Slides' successful launch. The story highlights the importance of collaboration, iteration, and a clear user-focused vision in building 0 to 1 products.

LEARN HOW TO SAY NO IF YOU WANT TO THRIVE

Success in digital products demands focus, but others will often push for more. Help them self-reflect by asking how their requests align with goals and strategy, leading them to prioritize what truly matters. This approach preserves focus without harming relationships.

PRICING MASTERCLASS: EVERYTHING YOU NEED TO KNOW ABOUT SAAS PRICING (88 MINUTE PODCAST)

This podcast covers essential strategies for optimizing SaaS pricing, including setting pricing, designing a pricing page, managing feature paywalls, converting free users to paid, handling cancellations, and quick pricing wins, with key takeaways on avoiding common pricing mistakes, evolving pricing as a company grows, and learning from both B2B and B2C pricing models.

POSITIONING STACKS

Effective positioning involves addressing the core problem, target audience, solution uniqueness, competitive edge, and product category during development and communicating in reverse, from product category to core problem.

HOW TO CRAFT ONBOARDING SURVEYS USERS LOVE: 5 DO'S AND DON'TS

When designing user onboarding, start with simple questions, explain how these questions benefit users, show care by asking about users' goals early, personalize follow-up questions based on previous answers, and avoid onboarding gaps by ensuring team alignment.

STRIPE'S MONOREPO DEVELOPER ENVIRONMENT

The developer environment at Stripe has tools and infrastructure built to support a large Ruby monorepo. Code is run on cloud-based devboxes instead of local machines. The company's team team invested heavily in automated code synchronization and a unified command-line tool called "pay" to manage devbox interactions. This post discusses the importance of organizational scale and codebase characteristics in shaping the design and success of the developer tooling.

AN APPROACH TO OPTIMIZING TYPESCRIPT TYPE CHECKING PERFORMANCE

You can optimize TypeScript's type-checking performance by first diagnosing inefficiencies using tools like --extendedDiagnostics, --generateTrace, and benchmarking libraries. You should iteratively test hypotheses by grouping related operators and refining overload order.

PARSING PROTOBUF AT 2+GB/S: HOW I LEARNED TO LOVE TAIL CALLS IN C

This blog post discusses a novel technique for optimizing interpreter loops and protobuf parsing using tail calls in C. The author discovered that by utilizing the `[[clang::musttail]]` attribute, they got a huge performance boost, achieving protobuf parsing speeds exceeding 2 GB/s. The key benefit of tail calls is the elimination of function call overhead, resulting in more efficient control flow and reduced stack usage.

GOOD REFACTORING VS BAD REFACTORING

Good code refactoring is incremental and preserves functionality while avoiding any new unnecessary complexity. Bad refactoring practices include drastic style changes, unnecessary abstractions, inconsistency in code patterns, refactoring without fully understanding the existing code, and ignoring the business context.

TIPS FROM 8 MONTHS OF TANSTACK/ROUTER IN PRODUCTION

After using TanStack Router in production for eight months, this author believes in organizing code into vertical modules, using loaders for page-level data and suspense queries for component-specific data, and avoiding premature code generalization. He promotes a “go gardening approach” that involves abstracting code only when clear patterns show up.

TYPES -> SETS

You can think about TypeScript's type system as a functional programming language that operates over sets of values. By imagining types as sets of all possible values they can construct, it becomes easier to reason about features like intersections, unions, and type mappings.

tickets now

SRCBOOK

Srcbook allows developers to create TypeScript and JavaScript notebooks to prototype quickly. These notebooks can run code locally with a web interface, export to Markdown, and be shared easily.

PARALLEL DOM

The parallel-dom npm package offloads heavy DOM operations to parallel iframes for faster rendering. It supports both plain JavaScript and React.

PRAGTICAL

Pragtical is a lightweight, highly extensible cross-platform text editor that prioritizes performance, ease of use, and open-source principles.

ECMASCRIPT SAFE ASSIGNMENT OPERATOR PROPOSAL

This proposal introduces the ?= operator, which makes error handling easier by transforming function results into a [error, result] tuple. It also supports recursive handling of promises and objects implementing Symbol.result, making error checking streamlined across various JavaScript APIs.

LLMS KNOW MORE THAN WHAT THEY SAY

An approach called Latent Space Readout (LSR) can improve the accuracy of AI application evals. LSR uses interpretability research by projecting evaluation model activations onto meaningful linear directions, which makes it more sample efficient than fine-tuning. This means that much less human feedback is required for these evals to be more accurate.

THE GIGANTIC AND UNREGULATED POWER PLANTS IN THE CLOUD

There's a dangerous lack of regulation for cloud-based management platforms controlling millions of solar panels across Europe. These platforms, mostly run by companies outside of Europe, have the power to shut down vast amounts of power, potentially causing grid collapse. While individual solar panels pose little threat, their concentrated control creates a huge risk, especially due to recent hacking attempts.

ASK HN: HOW DO YOU WORK AS A TECH LEAD? (HACKER NEWS THREAD)

A Hacker News commenter asks how to effectively manage a team of junior engineers, and responses cover topics such as team interaction, communication with managers, meetings, documentation, time management, and team alignment.

CLIENT-SIDE FILTERING OF PRIVATE DATA IS A BAD IDEA

A security researcher found that an app leaked sensitive user data due to client-side filtering, including preferred genders, sexualities, and relationship situations, through its GraphQL API.

SOURCEGRAPH WENT DARK

Eric Fritz, a former Sourcegraph engineer, explains how he preserved links to Sourcegraph's codebase after a repository went private by creating a fork, scraping relevant commits, and rewriting all references to his fork.

TIL: 8 VERSIONS OF UUID AND WHEN TO USE THEM

This blog post explains the eight versions of UUIDs, how they are generated, and when to use each version based on their strengths and weaknesses.

NEW AIRPOD MODELS TO BE INCLUDED IN FALL 2024 HARDWARE LINEUP

Apple plans to update its non-Pro AirPods lineup as early as next month. There will be two versions of the fourth-gen AirPods: a lower-priced model with basic functionality and a higher-priced version with active noise cancellation. There will likely be design changes to further distinguish the two models. The new AirPods will likely have Spatial Audio features and Apple Watch charging using the case, which will have USB-C.

EPIC GAMES STORE AND FORTNITE ARRIVE ON EU IPHONES

Fortnite is now back on iOS, but only in the EU. Epic has launched its games store on iOS (EU only) and Android. It currently only has three games, but Epic is working on bringing developers into its store. Fortnite was pulled from the Apple App Store four years ago, resulting in a legal battle that loosened Apple's grip on its platform but kept the status quo for the most part.

DESIGNER BABIES ARE HERE — SO WHY AREN'T WE TALKING ABOUT IT?

While human prenatal genome editing, which involves editing the DNA of a fetus visible inside the womb, has not happened yet as far as we know, the societal implications of the technology are still vast. The technology offers a chance to address cellular disease processes early and prevent symptoms from ever appearing. However, there are many ethical issues with using the technology. This article presents a summary of the many issues surrounding human prenatal genome editing and how it may affect society in the future.

THE INVISIBLE PROBLEM WITH SENDING PEOPLE TO MARS

There are obvious challenges to getting people to Mars, but the biggest obstacle might be something that's totally invisible and often overlooked: space radiation. Scientists are just beginning to learn about the many ways that exposure to space radiation can impact human health. The effects of space radiation may also combine with other factors such as sleep deprivation and microgravity - there is not enough information yet about how these factors interact. Finding ways to mitigate these effects will determine whether people will ever be able to visit Mars safely.

Schedule a demo to learn more

GITHUB ROLLS OUT AI-POWERED FIXES FOR CODE VULNERABILITIES

GitHub's new Copilot Autofix is a future that analyzes vulnerabilities in code and offers code suggestions to help fix them. The feature has been enabled by default for customers on GitHub Enterprise Cloud and will be offered for free in pull requests in open source projects beginning in September. It works with dozens of classes of vulnerabilities. Developers can dismiss, edit, or comment on suggestions in pull requests.

HOW TO GET FROM HIGH SCHOOL MATH TO CUTTING-EDGE ML/AI: A DETAILED 4-STAGE ROADMAP WITH LINKS TO THE BEST LEARNING RESOURCES THAT I'M AWARE OF

Even if you did well in high school math, the math in cutting-edge machine learning papers will still look like hieroglyphics. This article covers foundational math, classical machine learning, deep learning, and cutting-edge machine learning, aiming to help readers get from high school math to understanding cutting-edge machine learning. It contains links to free resources for learning each topic. While there is a wide debate about how much math is needed for machine learning, the necessary math is classified at the undergraduate level.

ROBLOX IS ALREADY THE BIGGEST GAME IN THE WORLD. WHY CAN'T IT MAKE A PROFIT (AND HOW CAN IT)?

More than 80 million people log into Roblox on average every day. Despite its linear growth, which is expected to continue, Roblox is very unprofitable. Its impressive rate of revenue growth has been outpaced by its costs. The company loses 49% of its revenue to app stores and content creators before it even gets into infrastructure and other costs. Roblox is increasingly focused on using generative AI and other features to attract more players, especially older players with money to spend and the discretion to spend it.

SAM ALTMAN'S WORLDCOIN IS BATTLING WITH GOVERNMENTS OVER YOUR EYES

Sam Altman's Worldcoin project has come under assault by authorities all over the world. More than a dozen jurisdictions have either suspended the project's operations or looked into its data processing. Worldcoin verifies people by using their irises, which it claims are better at distinguishing humans than fingerprints or faces. While the company says its technology is completely private, authorities are concerned that it could be used to build a global biometric database with little oversight.

SOCIAL MEDIA FOR STARTUP FOUNDERS: A PRACTICAL GUIDE TO BUILDING AN ONLINE PRESENCE

Social media is always evolving, but there are a few best practices that apply, no matter what's happening in the meta - this post covers time-tested guidelines for taking a social presence from 0 to 1.

MICRO-LIBRARIES NEED TO DIE ALREADY

Micro-libraries should either be copy-pasted into your code base or not used at all.

WORLD'S FASTEST CHARGER CAN FULLY POWER UP YOUR SMARTPHONE IN UNDER 5 MINUTES

Chinese consumer electronics company Realme unveiled its 320-watt fast-charging technology at the Realme 828 Fanfest event in Shenzhen, China on August 14.

IS THIS THE SLOW DECLINE OF THE APPLE “CULT”?

Despite achieving massive success, Apple continues to make decisions that put it at odds with the community that used to tirelessly advocate for it.

THE NETWORK SCHOOL

The Network School, a project that brings people together to empower the dark talent of the world, is starting with a 90-day popup from September 23 to December 23 on an island near Singapore.

DASEL

Dasel is a tool that allows users to query and modify data structures using selector strings.

⌛️ Round closes soon. Become an investor today

VenHub Smart Stores use AI to operate fully autonomously 24/7 [4],

anticipate inventory, and predict customer needs. They can fit nearly anywhere and can be opened in just a week.

GOOGLE'S UPGRADED AI IMAGE GENERATOR

Google has released Imagen 3, an advanced AI text-to-image generator, through its AI Test Kitchen and the Vertex AI platform. Imagen 3 is designed to produce images with enhanced details and lighting while maintaining content guardrails to prevent the creation of certain restricted imagery. Despite limitations, users have found ways to depict characters and logos resembling copyrighted material.

RUNWAY'S GEN-3 ALPHA TURBO IS HERE AND CAN MAKE AI VIDEOS FASTER THAN YOU CAN TYPE

Runway ML's new Gen-3 Alpha Turbo is now available, offering 7x faster AI video generation at half the price of its predecessor across various subscription plans, including free trials. This speed advancement significantly reduces time lag, fostering more efficient workflows, particularly for time-sensitive industries. Runway continues to push for improvements, including enhanced control mechanisms, while navigating the ethical landscape surrounding AI training data practices.

DEEPSEEK PROVER

DeepSeek has produced an incredibly powerful math model that was trained via process supervision and dramatically outperforms larger models on various MATH benchmarks.

DIFUZCAM: REPLACING CAMERA LENS WITH A MASK AND A DIFFUSION MODEL

This is a delightful project that uses a diffusion model to reconstruct really low quality images from a cheap camera.

BAM DENSE TO MOE UPCYCLING

This method allows dense models' FFN and Attention layers to be upcycled into a Mixture of Experts (MoE) model for further training. This saves a dramatic amount of compute cost while maintaining downstream performance.