Profile avatar
davilima6.bsky.social
That has such people in't.
23 posts 46 followers 142 following
Prolific Poster
Conversation Starter

To improve INP on two pages (estimated save: 200K$/year), with less than 3 months available, tons of data, a dedicated team, and complete freedom, we improved Preply Next.js application's INP! medium.com/preply-engin... 👇

saiu na Folha @folha.com

git clone htt''ps://github.com/python-pillow/Pillow cd Pillow files-to-prompt -c . -e .py -e c -e h | \ llm -m gemini-2.0-pro-exp-02-05 \ -s 'Explain how Pillow interacts with the Python GIL - include example code snippets as part of your explanation' gist.github.com/simonw/5d8e9...

On the heels of multiple airplane disasters, including one less than a month ago that killed 67 people, and after years of warnings about our understaffed air traffic control system, Trump has begun firing hundreds of FAA air traffic control workers. apnews.com/article/doge...

Have you ever stared at a #performance Flame Graph in DevTools in desperation? Those graphs 📊 look very cool and important, but making sense of them is always a struggle. The new article tries to help with that ☺️ You'll learn 👇🏼 www.developerway.com/posts/client...

Klarna was the company that went all-on replacing customer support with an AI bot and went on to brag about the cost savings. Now they are reversing course. Easy to see more companies blindly replacing quality customer support with a worse AI implementation will follow...

react: "rethink best practices" me: haha fuck yeah!! yes!! react: *rethinks best practices* me: well this fucking sucks what the fuck

New BBC research: 51% of all AI answers to questions about the news were judged to have significant issues of some form 19% of AI answers which cited BBC content introduced factual errors – incorrect factual statements, numbers and dates And more… www.bbc.com/mediacentre/...

#Contentful modeling patterns: - www.xtivia.com/blog/content... - www.contentful.com/help/topics-... - www.contentful.com/help/content...

Inspired by the AI hostility on this platform, I decided to share how I use AI. “I treat AI as I would a collaborator or a pretty good intern but I remain responsible for the final outcome.” lucumr.pocoo.org/2025/1/30/ho...

My talk from @reactdayberlin.gitnation.org has been published. If you want to know how I learned some of the RSC internals and what does RSC really mean, after all, you can find it here: gitnation.com/contents/mee...

never gonna Alt+F4 never gonna Ctrl+W never gonna :q! and desert you never gonna Ctrl+X never gonna Esc never gonna Ctrl+D and hurt you

saiu na Folha @folha.com

Mais um dia normal na Venezuela.

DeepSeek-R1 is now available in GitHub Models (Public Preview) https://github.blog/changelog/2025-01-29-deepseek-r1-is-now-available-in-github-models-public-preview

Meta's Llama Framework Flaw Exposes AI Systems to Remote Code Execution Risks

Facebook BigPipe: the origin of React Server Components

Generative AI: how the data-center Power gold rush will play out (via @simonwillison.net)

Interviewer: Where do you see yourself in 5 years? Me: Listening. Yes I'd say listening is my biggest weakness.

How would you like to get my new course for free? Well, almost for free. I want to bring my course to other languages. I have done the first step and created subtitles for the course lessons, but I need help with checking the subtitle quality by a native speaker.

"Compatibility means deliberately repeating other people's mistakes." – David Wheeler Bom dia!

Just dropped: Zack DeRose has an article on TypeScript project references - check out what they are, how they help, and how Nx can help you automate the maintenance of references! Blog Post: nx.dev/blog/typesc... YouTube Video: youtu.be/SDE3cIq28s8

Detailed notes about Anthropic's new Citations API, including why it's interesting if you're building RAG applications and a little bit of code to explore what it can do simonwillison.net/2025/Jan/24/...

Here are my notes on OpenAI's new ChatGPT Operator browser "agent", including initial thoughts on their approach to mitigating prompt injection risks simonwillison.net/2025/Jan/23/...

Here's everything I've figured out about the new OpenAI o1 family of models so far https://simonwillison.net/2024/Sep/12/openai-o1/

Essa é uma lapada segura e essencial. Ñ basta dizer que não tolera esse tipo de comportamento, é preciso agir. Palmas p os 2

Today I learned about this area called "MarTech" It refers to "software marketing teams use so they do not have to bug engineers to build stuff like send customized emails, track customers, optimize ads etc" It's a MASSIVE category, apparently. Thousands of companies and big $$

CafÊ com: - Bolo de fubå - Broa de milho - Inhame cozido - Batata doce - Mandioca - Paçoca O maluco que come feijão com molho de tomate no cafÊ da manhã vem dizer que não combina.

Anyone else think Severance has the same austere and platonic visual aesthetic as Apple devices? This also seems to be true if a lot of other Apple TV content. Inn silo it’s disguised. In Foundation it’s clearer. I’m going to call it geometric noir

Morreu o Jules Feiffer. Aqui um cartum de 1980

Tempos Estranhos🤔

O TIKTOK VOLTOU! A "rede das dancinhas" voltou nos EUA porÊm com algumas observaçþes interessantes: comentårios contra Elon Musk, Meta, Mark Zuckerberg estão sendo censurados. Comentårios pró-palestina tambÊm. Essa Ê a liberdade de expressão das Big Techs.

Most times, talking about roles and responsibilities when teamwork is hard is the wrong move. When you’re outcome-driven, people can set aside their egos and collaborate. When it’s turf-driven, the best ideas can’t surface because people have to stay in their lanes.

Builder.io, AI and the future of headless CMS cmscritic.com/builderio-ne...

My notes on the new OpenAI ChatGPT scheduled tasks feature, including a copy of the new tool instructions from a leaked system prompt simonwillison.net/2025/Jan/15/...

How #Cursor applies suggestions faster (parallel speculations): fireworks.ai/blog/cursor

Why is it that Brazil is one of the few countries around that remembers it is, in fact, a state with state capacity

the missing o1 prompting guide it’s not a chat model, it’s more like email. You have to provide a lot of context and merely tell it *what* you want, not *how* it should be done www.latent.space/p/o1-skill-i...

LLM Scraper with @nuxt.com & @cloudflare.social • CF AI using llama-3.1-70b-instruct • CF Browser for web scraping • CF KV for caching results for 10 minutes All powered by @threepointone.bsky.social's excellent workers AI packages. github.com/atinux/llm-...