Tropidurus (@tropidurusbr) 's Twitter Profile
Tropidurus

@tropidurusbr

Mea citrea, mea citrea arbor

ID: 121916177

calendar_today11-03-2010 01:11:16

65 Tweet

11 Followers

165 Following

LukeberryPi (@lukeberrypi) 's Twitter Profile Photo

se você já ouviu falar de Claude Code mas não sabe como funciona, assiste esse vídeo VOU TE MOSTRAR O QUE O CLAUDE CODE REALMENTE FAZ (codando uma feature de verdade) ninguém fez isso por mim e por isso demorei pra entender por favor RETWEET para ajudar mais pessoas

Rohan Paul (@rohanpaul_ai) 's Twitter Profile Photo

Rude prompts to LLMs consistently lead to better results than polite ones 🤯 The authors found that very polite and polite tones reduced accuracy, while neutral, rude, and very rude tones improved it. Statistical tests confirmed that the differences were significant, not

Rude prompts to LLMs consistently lead to better results than polite ones 🤯 

The authors found that very polite and polite tones reduced accuracy, while neutral, rude, and very rude tones improved it.

Statistical tests confirmed that the differences were significant, not
Giovanni Bassi 🇧🇷🦋 (@giovannibassi) 's Twitter Profile Photo

A bolha da IA tá tão vergonhosa que a OpenAI tá prometendo construir data centers mas nos lugares prometidos não tem (e não terá por anos) energia elétrica suficiente. Que piada!

Louis Gleeson (@aigleeson) 's Twitter Profile Photo

OMG… you can now fine-tune your own AI model and run it entirely on your device. No servers. No API keys. No data leaks. Gemma 3 270M just made local LLMs real and it only takes about an hour to build your own. Here’s how:

Alex Prompter (@alex_prompter) 's Twitter Profile Photo

Holy shit...Google just built an AI that learns from its own mistakes in real time. New paper dropped on ReasoningBank. The idea is pretty simple but nobody's done it this way before. Instead of just saving chat history or raw logs, it pulls out the actual reasoning patterns,

Holy shit...Google just built an AI that learns from its own mistakes in real time.

New paper dropped on ReasoningBank. The idea is pretty simple but nobody's done it this way before. Instead of just saving chat history or raw logs, it pulls out the actual reasoning patterns,
Santiago (@svpino) 's Twitter Profile Photo

Building software is only 10% about writing code, and 90% about thinking what code to write, why you should write it, and how to do it well. You can delegate the code-writing part to an LLM, but so far, they have shown they can’t handle the thinking part. It’s called

Santiago (@svpino) 's Twitter Profile Photo

Software development has never been about writing code. This was obvious to many before, but now AI has made it crystal clear to everyone. Building software is 99% thinking, 1% writing code, and 1% fighting off-by-1 errors.

Brave (@brave) 's Twitter Profile Photo

The security vulnerability we found in Perplexity’s Comet browser this summer is not an isolated issue. Indirect prompt injections are a systemic problem facing Comet and other AI-powered browsers. Today we’re publishing details on more security vulnerabilities we uncovered.

Tropidurus (@tropidurusbr) 's Twitter Profile Photo

The RAG Obituary: Killed by Agents, Buried by Context Windows Why Retrieval-Augmented Generation Won't Survive the Context Revolution and the End of Chunking, Embeddings, and Rerankers as We Know Them. open.substack.com/pub/nicolasbus…

problema cibernetico (@cyberproblem1) 's Twitter Profile Photo

o mais engraçado é que o pior caso do quick sort para esta lista seria um O(n²) que terminaria instantaneamente enquanto essa solução é um O(n) que no pior caso poderia durar até o fim do universo

Bruno Orlandi🔸 (@brorlandi) 's Twitter Profile Photo

CEO do Nubank: acaba com o home office. CEO do iFood: "Ninguém provou que o presencial é melhor" Mas uma coisa é certa: quem trabalha remoto pede mais iFood que quem trabalha presencial.

Könings (@edwardkonings) 's Twitter Profile Photo

Actively managed funds in Brazil are so inefficient (especially considering fees) that they are now lobbying to create an information asymmetry between agents (them) and the principals (the investors). I couldn't think of a better testament to their incompetence.