Oriol González

Web developer. Human after all.

[Anthropic @AnthropicAI]

New Anthropic research: How university students use Claude.

We ran a privacy-preserving analysis of a million education-related conversations with Claude to produce our first Education Report.

[Addy Osmani @addyosmani]

Personal Software is moving us from "buy what you need" to "generate what you need"

Software development is undergoing a profound shift, but not in the way most predict. While headlines trumpet the "end of programming," the reality is both more interesting and more nuanced. We're witnessing the emergence of "personal software", where the boundary between users and creators becomes increasingly fluid. This is enabled by advances in large language models and AI-assisted development tools.

> One of the biggest AI trends I’m excited about is personal apps. Right now, anyone can create an app tailored to their specific needs, all without hiring a developer or designer. As a product design studio founder, should I be worried about this? Not at all. The more people build things, the more they’ll start to appreciate the craft behind great software - @theShaneLevine

This transformation isn't about replacing programmers. Instead, it's about expanding the universe of what's possible in software creation, particularly for highly personalized and niche applications that were previously impractical to build. The real story here isn't about AI replacing human developers - it's about how AI tools are enabling a new category of software that simply couldn't exist before.

Consider the current software development landscape. Traditional software companies focus on building applications that serve broad market needs, leaving countless niche use cases unaddressed. These gaps aren't just features waiting to be built - they're entire categories of software that couldn't exist in the traditional model because they're too specialized to be commercially viable.

The rise of AI-assisted development is changing this equation. When someone can describe their specific needs conversationally and receive working code in response, the economics of personal software development shift dramatically. This isn't about building the next Photoshop or Figma - it's about enabling individuals to create highly specialized tools for their unique workflows, hobbies, and problems.

> I'm completely non-technical (never coded). But I just built a Web app from my phone by chatting with an AI agent on Replit. Now I have a tracker for to answer life's most important question: "Has the dog been fed?" - @venturetwins, a16z

A crucial distinction emerges here between personal software and traditional application development. Personal software operates in a different domain with different constraints. It doesn't need to scale to millions of users, handle edge cases beyond its creator's needs, or maintain backward compatibility across versions.

> Apparently, you don’t need coding experience to build apps anymore. So I made a little workout/yoga app just for me - @LenardFloeren

When the time and expertise required to create software drops dramatically, the minimum viable market size for a software project approaches one. This isn't just a quantitative change - it's a qualitative transformation in what constitutes a "marketable" software product. This simplified context is precisely what makes AI-assisted development particularly effective in this space.

> The least you know of a specific domain the more impressed by AI outputs you are aka the competency illusion. - @gerardsans

The skepticism from experienced developers about AI replacing programmers is well-founded - building and maintaining large-scale software systems does require deep technical knowledge, architectural understanding, and experience that current AI systems simply don't possess. But this skepticism sometimes blinds us to the genuine transformation happening in personal software development.

Think of it this way: just as spreadsheets enabled non-programmers to perform complex calculations and data analysis, AI-assisted development tools are enabling non-programmers to create personal software solutions. The key difference is that while spreadsheets provided a constrained environment for specific types of problems, AI-assisted development opens up the possibility of creating full-fledged applications tailored to individual needs.

This shift has profound implications for the software industry. We're moving from a world of "buy what you need" to "generate what you need." The total addressable market (TAM) for software expands dramatically when individuals can create solutions for their specific needs rather than waiting for commercial software to address their use cases.

We're not just talking about capturing more of the existing software market - we're talking about expanding the very definition of what constitutes a software market. In the traditional model, software development was like industrial manufacturing: high fixed costs required large markets to justify production. The new model is more akin to craft production, where tools can be created for markets of any size, down to individual users.

This shift has interesting implications for how we measure the software market itself. Traditional metrics like license sales or subscription revenues may become less relevant as more software is created for personal use rather than commercial distribution. We might need new metrics that capture the value of personal software creation, perhaps measuring the time saved or problems solved rather than direct monetary transactions.

However, this transformation comes with important caveats. Personal software created through AI assistance may lack the robustness, security, and maintainability of professionally developed applications. It's crucial to understand these limitations and recognize that personal software isn't a replacement for professional software development - it's an entirely new category with its own strengths and weaknesses.

The emergence of personal software also raises interesting questions about the future of software development as a profession. Rather than diminishing the role of professional developers, it's likely to push the field toward higher levels of abstraction and complexity. As basic software creation becomes more accessible, professional developers will likely focus more on areas where deep technical expertise remains crucial: scalability, security, performance optimization, and complex systems integration.

The entire history of software engineering can be viewed as a steady march toward higher levels of abstraction. Just as the invention of compilers didn't eliminate the need to think programmatically but rather elevated our work to a higher level of abstraction, AI assistants represent the next step in this continual progression. The key difference is that instead of learning a formal programming language, users are learning how to express computational intent through natural language with sufficient precision to generate working code.

AI isn't eliminating programming; it's providing a new, more abstract interface to the same underlying process of instructing computers to perform specific tasks.

> More and more apps are being built on @lovable and @boltdotnew by people with zero coding experience. That’s good news for developers. Eventually, they’ll hit the limits of these tools—and turn to devs for real solutions. What was supposed to replace developers might create more opportunities. - @JulienLeg78

The counterintuitive reality of AI-assisted software development is that it's likely to create more opportunities for professional developers, not fewer. As users begin their journey with tools like no-code platforms and AI assistants, they inevitably encounter the boundaries of what these tools can accomplish. Much like how website builders ultimately expanded the market for professional web developers by making web presence accessible to small businesses, AI-assisted development tools may create a new generation of technologically literate clients who understand the value of custom software solutions.

Looking ahead, we can expect to see the rise of new platforms and communities centered around personal software development. These might include marketplaces for sharing AI-generated code, tools for customizing and extending personal applications, and new frameworks designed specifically for AI-assisted development.

The real revolution in software development isn't about AI replacing programmers - it's about AI enabling a new category of software creation that was previously impossible. This transformation will likely lead to a more diverse and vibrant software ecosystem, where commercial applications coexist with a long tail of highly specialized personal software solutions.

As we navigate this transition, it's important to maintain a balanced perspective. The rise of personal software doesn't herald the end of traditional software development any more than spreadsheets marked the end of financial software. Instead, it represents an expansion of what's possible in software creation, opening up new opportunities while preserving the essential role of professional software development.

The future of software development isn't either/or - it's both/and. Professional developers will continue to build the complex systems, while AI-assisted tools will enable individuals to create personal software solutions that were previously out of reach. This promises to make software more accessible, more personalized, and more deeply integrated into our lives than ever before.

[claire vo 🖤 @clairevo]

When trying to build 🔥 teams, leaders often forget the things that matter most.

They focus on: process, staffing, scope, plans, presentation, predictability.

They forget: hustle, grit, depth, finding 100x talent, love of the game, fun, rizz.

[Ethan Niser @ethanniser]

Introducing NextMaster

An open-source ecommerce @nextjs template inspired by McMaster-Carr, optimized for performance with over 1 million products.

A thread on how I built this with @rhyssullivan and @ksw_arman 🧵
(video is not sped up)

[Rob Eisenberg @EisenbergEffect]

Uh. No.

https://t.co/BzYDgT0agk

Learn the Web Platform. Then you can:

* Understand how all frameworks work.
* Learn any framework and master it quickly.
* Make better technical choices.
* Contribute to a healthier engineering maturity model.
* Build your own framework.
* Etc.

[Turborepo @turborepo]

Introducing Turborepo 2.0

◆ High-performance build system for JavaScript / TypeScript repos
◆ Incrementally adoptable – works any way you run scripts
◆ Dependency-aware task watcher for any tooling
◆ Works with monorepos (or without)

https://t.co/kc6IxOo9oR

[Oriol González @oriolgonzalez]

@odin_delrio @carlesnunez @bonpreuesclat @CodelyTV Si no es httpOnly, seria vulnerable por cross site scripting

[Oriol González @oriolgonzalez]

@odin_delrio Las utility types quizá? https://t.co/pfXsLKm544

[Bun @bunjavascript]

Introducing Bun v1.0 - a fast, all-in-one toolkit for running, building, and testing JavaScript and TypeScript

https://t.co/pjelxnxOSt

[Guillermo Rauch @rauchg]

There's two kinds of migrations: incremental, and failed ones.

https://t.co/vl9A82XFKS

[Félix López @flopezluis]

The sooner you can start talking about the impact of reducing KTLO in terms that the business understands, the more you will be able to incentivize it properly. U really need to be able to measure KTLO load in order to demonstrate the impact of reducing it
https://t.co/CPa5QgtHPN

[Jecelyn Yeen 🐟 @jecfish]

🤩 [1/4] Coming soon in @ChromeDevTools - You can overrides API response (Fetch/XHR) and more!

Imagine you can mock APIs in frontend without waiting for the backend changes!

Not only that...

[Alex Xu @alexxubyte]

The CAP theorem is one of the most famous terms in computer science, but I bet different developers have different understandings. Let’s examine what it is and why it can be confusing.

CAP theorem states that a distributed system can't provide more than two of these three guarantees simultaneously.

Consistency: consistency means all clients see the same data at the same time no matter which node they connect to.

Availability: availability means any client which requests data get a response even if some of the nodes are down.

Partition Tolerance: a partition indicates a communication break between two nodes. Partition tolerance means the system continues to operate despite network partitions.

The “2 of 3” formulation can be useful, but this simplification could be misleading.

- Picking a database is not easy. Justifying our choice purely based on the CAP theorem is not enough. For example, companies don't choose Cassandra for chat applications simply because it is an AP system. There is a list of good characteristics that make Cassandra a desirable option for storing chat messages. We need to dig deeper.

- “CAP prohibits only a tiny part of the design space: perfect availability and consistency in the presence of partitions, which are rare”. Quoted from the paper: CAP Twelve Years Later: How the “Rules” Have Changed.

- The theorem is about 100% availability and consistency. A more realistic discussion would be the trade-offs between latency and consistency when there is no network partition. See PACELC theorem for more details.

Is the CAP theorem really useful?
I think it is still useful as it opens our minds to a set of tradeoff discussions, but it is only part of the story. We need to dig deeper when picking the right database.

--
Subscribe to our weekly newsletter to get a Free System Design PDF (158 pages): https://t.co/4QcX8btXGL

[Vaughn Vernon @VaughnVernon]

People are suddenly complaining about #DDDesign Aggregates. In reality, Aggregate is not a complex pattern. I've been explaining simple designs for 12 yrs. Still, I don't understand why people insist on Aggregates when use cases prove they aren't needed. Basic Entity is fine.

[Ivan Akulov @iamakulov]

How I typically test performance (a thread for a client who asked about this):

[Colin | clerk.com @tweetsbycolin]

Joining in on the RSC visualization fun!

I used separate planes to differentiate composition with imports from composition with the App Router's layouts and pages.

[Seb ⚛️ ThisWeekInReact.com @sebastienlorber]

To be clear:

- I know sql` template literals protect against SQL injection by using parametrized templates / prepared statements => that's not my point

- My point is: mixing client/server code using colocation and module extraction is as safe as using separated back/front

[Guillermo Rauch @rauchg]

Why does amazon.⁠com feel so snappy?

Zoom out, behold: the power of streaming.
IMO, the best new feature of Next.js (https://t.co/a57bSKJVuM)

Above the fold, instant. Below? Streamed over time.
Even better: Next supports non-linear streaming 🤯 (eg: async stock *and* reviews)

[Playwright @playwrightweb]

Playwright 1.32 is out!

📣 Introducing UI Mode (preview)

UI Mode = Watch mode + Time-travel debugging

Watch the overview: https://t.co/K7H0tMj9LH

[Tech at DocMorris @docmorris_tech]

📢New Post in our blog 📢 We tell you how we do the tests at PromoFarma by DocMorris. Take a look! 😉
https://t.co/8jfEqdCYNy