David Pilato 🇺🇦🇪🇺🇫🇷

TUI Studio : un outil visuel pour designer des interfaces pour le terminal, sans code tui.studio

Pretty intuitive animation explaining seasons, equinox, solstice, length of day and night. Today marks the spring equinox in the Northern Hemisphere. [🎞️ German Valencia Garcia]

🚨 Do you understand what this man just pulled off.. > a guy from North Carolina used AI to generate hundreds of thousands of songs.. uploaded them to Spotify, Apple Music, Amazon.. then botted billions of streams on his own tracks and walked away with $8 million > 660,000 fake streams per day.. spread across thousands of AI songs so nobody noticed.. $1.2 million a year.. for music no human ever actually listened to real artists are out here grinding for 0.003 cents per stream.. promoting on TikTok.. begging for playlist placements.. and this guy just had AI make the music AND the audience first-ever criminal streaming fraud case.. he's paying back $8 million.. but the playbook is out there now.. and AI just got better since he started the music industry spent 10 years fighting piracy.. now they have to fight songs that don't exist being listened to by people who don't exist.

My fiend told me "oh, there exists non AI content in this world ^^" Yes, it still does. How testing Elasticsearch became simpler. Enjoy! elastic.co/search-labs/bl…

I came across an interesting read on head-based probabilistic sampling in OTel SDKs. For Node.js it's the default. You just need to configure the sampling rate. In @elastic EDOT it's via sampling_rate or OTEL_TRACES_SAMPLER_ARG. OTel SDK code in image.

Your agent doesn't call the right tool? It's probably because your tool description is bad. Tool selection is a common failure mode for agents: The agent doesn't call your tool at all. The agent calls the web search tool instead of the database search tool. In most cases, improving the tool description helps with this immediately. A tool description doesn't have to be long. But it has to have these core ingredients: 1. Core purpose: A high-level summary of what the tool actually does 2. Trigger: When the tool should be called 3. Action: What does the tool do? For example, what specific data does it retrieve? 4. Limitations: Are there any constraints? For example, does it only return a limited number of results? 5. Relationships: How does it relate to other tools? Does it need to be called before or after a specific other tool? 6. Examples: Add examples of user queries and how to use the tool for them.

Your embeddings might be 32× larger than it needs to be! Binary quantization compresses each embedding dimension from a (32-bit) float to a single bit. Positive values become 1, negative become 0. That takes a 4,096-byte vector down to... (checks notes) 128 bytes!

🚨𝘽𝙍𝙀𝘼𝙆𝙄𝙉𝙂: European Commission President Ursula von der Leyen unveiled EU–INC, a new framework that lets you launch a company in 48 hours for under €100 Starting a company across the EU today = 27 legal systems, 60+ company structures 🤯 That might be about to change… The European Commission just introduced 𝗘𝗨 𝗜𝗻𝗰., a new optional corporate framework designed to make Europe actually function like one market. Here’s what stands out: → Set up a company in 48 hours → Cost: < €100 → Fully online, no minimum capital → One single framework across all EU countries → Easier share transfers & fundraising → EU-wide employee stock options (huge for talent) Especially the EU-wide stock option plans, taxed only when employees actually sell (instead of when granted) is huge. This makes it far easier for startups to attract and retain top talent, finally putting Europe closer to the US playbook. Source/More info: ec.europa.eu/commission/pre… In short: This is Europe trying to compete with the simplicity of a Delaware C-Corp 🇺🇸 And honestly… it’s long overdue. For years, European founders had 2 choices: 1. Stay local and deal with fragmentation 2. Move to the US to scale 𝗘𝗨 𝗜𝗻𝗰. is trying to remove that trade-off. If executed well, this could be one of the most important structural changes for European startups in decades. What do you think?

obligatory GTC keynote tweet when you make the top slides: building HNSW graphs with up to 12x the throughput and 7x faster merges on elasticsearch and NVIDIA cuVS (a GPU-accelerated library for vector search) powered by CAGRA (graph-based ANN algorithm built to run natively on GPUs) that still works with CPUs for search PS: yeah, this is a lot of acronyms

Official Agent Skills for Elastic are here!

Qui aurait pu prédire ?

Sound advice.

🤯BREAKING: Alibaba just proved that AI Coding isn't taking your job, it's just writing the legacy code that will keep you employed fixing it for the next decade. 🤣 Passing a coding test once is easy. Maintaining that code for 8 months without it exploding? Apparently, it’s nearly impossible for AI. Alibaba tested 18 AI agents on 100 real codebases over 233-day cycles. They didn't just look for "quick fixes"—they looked for long-term survival. The results were a bloodbath: 75% of models broke previously working code during maintenance. Only Claude Opus 4.5/4.6 maintained a >50% zero-regression rate. Every other model accumulated technical debt that compounded until the codebase collapsed. We’ve been using "snapshot" benchmarks like HumanEval that only ask "Does it work right now?" The new SWE-CI benchmark asks: "Does it still work after 8 months of evolution?" Most AI agents are "Quick-Fix Artists." They write brittle code that passes tests today but becomes a maintenance nightmare tomorrow. They aren't building software; they're building a house of cards. The narrative just got honest: Most models can write code. Almost none can maintain it.

Sending all the best vibes to @GeekSessionsFAO for a fantastic conference! 🇵🇹 Sad to miss out on the action in Faro this year, but I know it's going to be an incredible event. Have a blast, everyone! 🔗 geeksessionsconf.pt

@pikubogame Just in time as I'm about to finish the pack 03 😅

Something weird showed up at the bottom of my screen when I loaded PiKuBo to replay some old levels. I guess I oughta give the new puzzles a playtest to ensure they are up to standard. 🧐 (The lack of a crown down there is also giving me some OCD 😅) 🧩 New pack progress: 14/36

La bonne vieille dynamo, mais à l'envers 😉

@clever_cloud @LostInBrittany @ldussart @ERougeux @AurrelH95 Merci pour les informations sur la 9.3 d'Elasticsearch. Pour préciser, JinaAI ce n'est pas un partenariat mais bien appartenant à @elastic. Donc nous avons bien nos propres modèles et encore plus à venir 😉

[𝗠𝗔𝗖𝗜] Vous avez déjà rêvé d’un agent IA qui accompagne les devs juniors ? Justement, on en parle dans le #MACI 156 : youtu.be/6b32Epy1WWM Avec @LostInBrittany et nos experts (Miami) @ldussart, @ERougeux et @AurrelH95.

This is a must read article. It explains one by one all the steps needed to make a great search bar sending back relevant results. Very well done Saurabh! 🤩

Yeah 𝚐𝚛𝚎𝚙 is cool but have you tried 𝚓𝚒𝚗𝚊-𝚐𝚛𝚎𝚙? It's a semantic grep alternative with 4 modes: 1. Standalone mode: for direct semantic search 2. Pipe mode: for reranking grep outputs 3. Zero-shot classification mode: to label line by line 4. Code search mode: for semantic search with code embeddings Check it out: github.com/jina-ai/jina-g…

Are you a #Java developer? ☕️ Laura Trotta shares how to master hybrid search using @langchain4j and @elastic. Highly relevant search in just a few lines of code! Read more: elastic.co/search-labs/bl…