iiinova — where your digital self keeps evolving

Alice Alexandra @alicealexandra.com · 19/02/2026

Running multiple AI agents gets ridiculous fast. - You have 100 windows open - Your computer fans are crying - Agents keep bonking into each other in the code - You can't ever shut the laptop, which makes it tough to just... have a life Here's how to manage the chaos: builder.io/blog/ai-age...

A graph shows a decline in developer sanity as the number of AI agents increases, with a call to "stop the decline" alongside.

000 0

Alice Alexandra @alicealexandra.com · 10/09/2025

Anthropic agreed to pay $1.5B to settle claims it trained AI on millions of pirated books. But the judge hasn’t approved it yet, questioning if $3,000 per book is the right precedent. What do you think? Does this move us forward on fair compensation or just leave questions?

AI-generated scene of a futuristic server room with glowing blue humanoid figures made of data holding open books, while more books float above a floor of streaming code; two books are stamped “PIRATED EDITION,” and the word “Anthropic” glows on the back wall.

110 0

Alice Alexandra @alicealexandra.com · 12/08/2025

Cars have EPA window stickers and EU CO2 caps. AI has... "trust us, bro, it's fine." We need enforceable ‘AI Energy & Emissions’ labels on every major model.

A doctored screenshot of a ChatGPT conversation. A user asks, "How much water do you use per query?" ChatGPT responds that it uses a very small amount, "roughly 0.32 milliliters." However, a line of red text has been edited in below this, humorously contradicting the official answer by stating, "This response used 100mL of water."

030 0

Alice Alexandra @alicealexandra.com · 05/08/2025

Some tips for code reviews without the tears: www.builder.io/blog/code-r...

A two-panel meme using stills of the main character from the TV show Squid Game. The top panel shows the character smiling gleefully with the caption, "watching ai generate code." The bottom panel shows the same character with a serious, stressed expression, captioned, "reviewing ai generated code."

010 0

Alice Alexandra @alicealexandra.com · 17/07/2025

you gotta scrape OpenAI back sometimes

100 0

Alice Alexandra @alicealexandra.com · 14/07/2025

Keeping Figma in sync with Storybook (and your entire design system) just got way easier.

110 0

Alice Alexandra @alicealexandra.com · 03/07/2025

Hot take: Design handoffs won't work until designers can ship code. Current MCP clients force you into dev tools or basic chat. PMs spec in docs, designers mock in Figma, and devs rebuild everything. What if one MCP client let the whole team code? 👇 builder.io/blog/mcp-cl...

A diagram illustrating the MCP architecture, with the title "Pick the right one" pointing to the client. The 'MCP client' is shown connecting to various AI models (like ChatGPT), while it communicates with an 'MCP server' that connects to different backend services (like Figma, Notion, and GitHub).

000 0

Alice Alexandra @alicealexandra.com · 27/06/2025

Debugging dark mode issues in an old, messy SvelteKit codebase is usually nightmare fuel. This morning, I connected my repo and put Builder Fusion on the case. It agentically iterated with the code and full DOM context, and it figured out the problem with a single prompt.

110 0

Alice Alexandra @alicealexandra.com · 27/06/2025

new favorite word just dropped

Dictionary entry for "bafflegab," a noun that means "incomprehensible or pretentious language, especially bureaucratic jargon."

000 0

Alice Alexandra @alicealexandra.com · 10/06/2025

What if you could mock up prototypes that stakeholders can actually click through? You don't have to just pretend it works. Use a real GitHub PR workflow and mock a site all in less time than it takes to wireframe. Here’s how. www.builder.io/blog/mock-u...

A visual comparison of two development workflows. The top timeline moves from “Idea” through “Wireframe,” “Prototype,” and “Code” to “PR,” ending with a red X. The bottom skips intermediate steps, jumping from “Idea” directly to “PR,” with a dotted line labeled “Time saved” and a green checkmark.

000 0

Alice Alexandra @alicealexandra.com · 02/06/2025

Figma's new MCP server gives AI access to structured design data instead of screenshots. But is the resulting code pixel-perfect? I've put it to the test is my latest post. www.builder.io/blog/figma-...

Image illustrating Figma Model Context Protocol (MCP) translating a product card design (left, showing visual layers and dimensions) into a React code component structure (right, with placeholders like ProductImage, ProductName).

000 0

Alice Alexandra @alicealexandra.com · 15/05/2025

My problem with vibe coding isn't the AI; it's the vibe. Many AI tools disrespect your team's hard work and talent. What we need is AI for grown-ups. Here's my thoughts on what that means. builder.io/blog/ai-for...

Visual comparison highlighting a preferred method for AI development. The "Vibe coding" approach (disapproved with a red X) shows a direct, simplistic request to "Create an ecommerce app". In contrast, "AI for grown-ups" (approved with a green checkmark) shows a more structured approach involving design (Figma), code, and data considerations feeding into the AI prompt.

000 0

Alice Alexandra @alicealexandra.com · 05/05/2025

You can teach any agent to fish, but wouldn't you rather it know who to call to get fish on demand? This is what Google's new A2A protocol promises: your agent gets a list of contacts for when the questions get too tough. www.builder.io/blog/a2a-pro...

010 0

Alice Alexandra @alicealexandra.com · 05/05/2025

AI agents are like Twitter devs: great solo, terrible at collaboration. MCP gave agents standardized tool access to information across the internet, but the story isn't done. A generalist agent with tons of tools still isn't as useful as an orchestrated network of specialists.

121 0

Alice Alexandra @alicealexandra.com · 28/04/2025

If you're in the midst of yet another 5,000-word manifesto-as-AI-prompt, it might be time to take a beat. There's a nicer route...

Diagram showing the composition of a full prompt sent to an LLM, consisting of a large "chonky system prompt" ($0.25 per request) and a small "user query" ($0.001 per request), highlighting the cost difference.

121 0

Alice Alexandra @alicealexandra.com · 25/04/2025

What's worse is that most of the AI tools we use today are "stateless." Every interaction is a completely blank canvas for the AI. Even though we may feel like we've been building a working relationship with the machine, it has no idea who we are or what we do.

Sad Pablo Escobar: "Me realizing the AI doesn't care about the last hour we spent together."

100 0

Alice Alexandra @alicealexandra.com · 23/04/2025

not the likes and reposts echoing the name 🙈 just needed to drop 11 days ago, and the whole world would have collapsed into coincidence. but anyway, congrats on the launch!! gonna ruin the numbers now.

4 reposts and 12 likes on the post when I encountered it.

010 0

Alice Alexandra @alicealexandra.com · 20/04/2025

So, this is Citizen Sleeper's power: Every single action you spend is a choice _not_ to spend it somewhere else. What will you prioritize? Just like real life, there aren't often wrong answers. Only interesting consequences for you and the people you've met.

Three of the early characters you meet in Citizen Sleeper.

110 0

Alice Alexandra @alicealexandra.com · 20/04/2025

Drives aren't linear quests—they're just a list of goals you could be progressing. But drives are often bound by time. If you don't complete one in x amount of days, the opportunity goes away forever. You can't complete them all, and you don't always know when they'll fade.

A list of drives in Citizen Sleeper, such as "Disable your tracker" and "Improve the Overlook."

100 0

Alice Alexandra @alicealexandra.com · 20/04/2025

This minimalism—I only have x amount of things I can do today—seems like it would lead to choice paralysis. But the game overlays all your choices with meta-narratives called "drives," which is also how your character levels up to do more interesting things.

100 0

Alice Alexandra @alicealexandra.com · 20/04/2025

How? Every "day" in the game, you get 1-5 actions to go spend absolutely anywhere in the world. Each of your actions is rolled as a dice, so it has varying amounts of power each day. The game progresses only by you spending actions.

A location within the world map of Citizen Sleeper, with "actions" shown up top as dice that can be spent in options at the location.

100 0

Alice Alexandra @alicealexandra.com · 20/04/2025

That said, most games have a weird sense of time. If there's a monster pillaging a village, you can go on an 18-hour sidequest, and the village will be just fine. What Citizen Sleeper does is make you choose what to prioritize, because it won't be there forever.

An early location in Citizen Sleeper, with the clock on Dragos' debt running down each day.

100 0

Alice Alexandra @alicealexandra.com · 20/04/2025

Most games ask you to: - Make choices in dialogue, often between a finite number of "good" and "less good" options - Somewhat linearly go on "quests" based on those choices

Citizen Sleeper also has choices in its dialog. In this screenshot, Ankhita is talking to the player, and the player can respond with, "Sorry to hear it" or "Anything I can do to help?"

100 0

Alice Alexandra @alicealexandra.com · 20/04/2025

Finally getting around to playing Citizen Sleeper, and I'm obsessed with the design of choice throughout the narrative. 🧵

The banner for Citizen Sleeper—a space station city cast in a red haze, while a tiny passenger ship flies by and one lone figure watches it.

120 0

Alice Alexandra @alicealexandra.com · 17/04/2025

kinda cute how Gemini keeps complimenting ESLint in its lil' agentic back-and-forth. sensing a Cursor office romance blooming. 🌼

Gemini responding to the Cursor linter, saying, "Good catch by the linter."

110 0

Alice Alexandra @alicealexandra.com · 16/04/2025

2. A cloneable template repo, with lots of inline explainers: github.com/3mdistal/cs...

Architecture diagram of an AI system. An AI client interacts with an MCP server core, which registers Prompts, Resources, and Tools modules. Resources connect to Persistent Storage, and Tools connect to an OpenRouter API.

100 0

Alice Alexandra @alicealexandra.com · 16/04/2025

1. A step-by-step guide on how to build your first MCP server (without vibe coding): www.builder.io/blog/mcp-se...

Gru makes a plan meme: 1. "Vibe code with Cursor." 2. "Don't read the code." 3. "Expose keys in prod." Wait.

100 0

Alice Alexandra @alicealexandra.com · 16/04/2025

I've gotten frustrated with the constant context switching between all the AI editors. So, I figured it was time to build my own MCP server to share memories between them. (But I didn't want to hoard the knowledge—here's some goodies for ya'all. 🧵)

"Own your AI workflow" pointing to an MCP server that connect major AI providers like Claude, OpenAI, Gemini, Cursor, and Windsurf.

110 0

Alice Alexandra @alicealexandra.com · 11/04/2025

Before you pack up your things to go lie down in a meadow and wait to be replaced by AI, you might want to hear this. Web devs: AI's not coming for your job. You're actively building its role in our world.

Central diamond graphic saying, "You are here," positioned over a background collage of faded agentic architecture examples, with arrows pointing to surrounding blank boxes.

111 0

Alice Alexandra @alicealexandra.com · 31/03/2025

honestly the most aggressive critique i've ever seen of my choice of passwords 🥲

A password input that's labeled, "Strength: Useless" in red text.

000 0

Alice Alexandra @alicealexandra.com · 14/03/2025

the tl;dr of why DevRel exists

Programmer: "Just read the documentation." The documentation: [A diagram of a 3x1 Lego piece directed to fit on a 4x1 platform with no leftover space.]

000 0

Alice Alexandra @alicealexandra.com · 16/11/2024

it’s giving gender dysphoria

A weather report saying 9, but feels like 2.

020 0