don't follow me, I'm lost too
Milan
Joined March 2009
- Tweets2.5K
- Following785
- Followers285
- Likes1.2K
Consigli di lettura di luglio! usthemyours.com/consigli-di-… ma li seguite i consigli di @rxstnz ? #letture #editoria #estate #2026
34
andrea d0k retweeted
🚨 DeepSeek just did something wild.
They built an OCR system that compresses long text into vision tokens literally turning paragraphs into pixels.
Their model, DeepSeek-OCR, achieves 97% decoding precision at 10× compression and still manages 60% accuracy even at 20×. That means one image can represent entire documents using a fraction of the tokens an LLM would need.
Even crazier? It beats GOT-OCR2.0 and MinerU2.0 while using up to 60× fewer tokens and can process 200K+ pages/day on a single A100.
This could solve one of AI’s biggest problems: long-context inefficiency.
Instead of paying more for longer sequences, models might soon see text instead of reading it.
The future of context compression might not be textual at all.
It might be optical 👁️
github. com/deepseek-ai/DeepSeek-OCR
andrea d0k retweeted
Cercherò di essere il più accademico possibile, ma dire che un intero sistema critico per il Paese si ferma “per un chiodo” NON È una scusante ma semmai un’AGGRAVANTE, e proprio per i decisori, non certo per l’operaio che ha piantato il chiodo.
sift.com/pricing @GetSift if you have problem with pricing just ask for help, here I am to help you in the meanwhile consider to remove the page from the google index ... take this as a #seo and ##brand tip for free
1
60
Ok che Google ha il piacere di uccidere le sue creature, ma questo davvero è uno sgambetto ad internet che si poteva evitare, com'era? don't be evil, be messy?
Google's goo․gl links will stop working in August 2025 9to5google.com/2024/07/18/go… by @technacity
1
78
Ogni tanto mi capita di organizzare cose belle come “Agave”: la #mostra di Federico Guarino a #Milano usthemyours.com/agave-essenz… se non sapete cosa fare questa #domenica passate da @usthemyours
100
andrea d0k retweeted
Scientists at the University of Maryland have developed an "invisibility cloak" for AI using adversarial patterns on a sweater, making the wearer nearly undetectable to standard object detection methods.
teapawt.org/omg_interesting/status…
andrea d0k retweeted
🤯 The level of sophistication of the XZ attack is very impressive! I tried to make sense of the analysis in a single page (which was quite complicated)!
I hope it helps to make sense of the information out there. Please treat the information "as is" while the analysis progresses! 🧐 #infosec #xz
msn.com/it-it/notizie/tecnol… quando un'idea stupida, viene implementata male e gestita peggio
26
andrea d0k retweeted
CloudFlare mi ha informato che il mio progetto @PhishingArmy è stato oscurato dalla piattaforma anti pirateria #PiracyShield!
#PhishingArmy è un progetto che mensilmente protegge 500mila utenti dal #phishing!
Per interessi privati, #calcio, mettiamo a repentaglio gli utenti! 🤦♂️
Consigli di lettura usthemyours.com/consigli-di-… via @usthemyours tra i libri recensiti da pablito molto interessante la "Critica della democrazia occidentale" di @davidgraeber edito da @ed_eleuthera
1
154
andrea d0k retweeted
1\ My thoughts on the Chinese APT contractor leak 🇨🇳
Specifically, I want to talk about the leaked
- iOS Spyware
- Physical implantable devices
- Email surveillance system
Let's consider detection and how these would be installed.
#threatintel
someone just leaked a bunch of internal Chinese government documents on GitHub
github.com/I-S00N/I-S00N/
