I hope to live long enough to see strong AI become real.

Joined July 2014
Piero Savastano retweeted
“Affittare” o “possedere” un’intelligenza artificiale? Nella nuova puntata di "Intelligenze Artificiali – In mezzo a noi" parliamo di sovranità digitale e della differenza tra usare modelli nel cloud delle Big Tech o installarli direttamente sui propri sistemi. Al centro, il mondo Open Source: soluzioni che possono essere scaricate e fatte girare sui propri computer, particolarmente rilevanti per settori come sanità e finanza, dove la gestione di dati sensibili richiede controllo e riservatezza. Ne discutiamo con @_pieroit_ Savastano, fondatore di CheshireCat.ai @Stregatto_AI per fare il punto sulla maturità dei modelli aperti: non più alternative di “serie B”, ma strumenti concreti per ridurre i costi, mantenere il pieno controllo del software e costruire assistenti digitali personalizzati sui propri dati. 📺 Appuntamento sabato 14 febbraio, alle 17.00 su @MediasetTgcom24 . #IntelligenzaArtificiale #OpenSource #SovranitàDigitale #Cybersecurity
1
9
1,081
Piero Savastano retweeted
Che direzioni prenderà l'#IA generativa nel 2026? Di questo parleremo venerdì 16 alle 18 con @_pieroit_ (esperto di IA e creatore di @Stregatto_AI) e @GiuseppeMayer (AD di @TalentGardenit). Domande benvenute! Chiedi la notifica qui youtube.com/live/6SVXdIyQBK0
1
1
2
543
Piero Savastano retweeted
🔓Take Full Control of Your AI – No Third-Party Limits! 🤖 Watch: 🔗 youtu.be/Yo8UQsfpUP4 👈🏻 The future of AI is open, customizable, and in your control. Why rely on third-party services with hidden limitations when you can build, train, and deploy your own AI agents on your terms? ✍🏼 In this AI Heroes 2024 session, @_pieroit_ (Cheshire Cat AI) and @m4oc (@seeweblive) will guide you through creating a vertical AI, leveraging open models and the open-source @CheshireCatAI framework ("Stregatto") to build personalized AI agents that are truly yours. 🏗️ 💡 What you'll learn: - How to build AI agents that aren’t bound by external vendors - The power of open-source models for AI independence - How to deploy AI models efficiently using Seeweb’s cloud infrastructure 🎟️ Join AI Heroes 2025 on May 27! Secure your spot: aiheroes.it/2025/tickets/ #AIHeroes2025 #AI #OpenSourceAI #MachineLearning #CustomAI #ArtificialIntelligence #CheshireCatAI #AIInnovation
1
8
2,513
Live su #Hinton e #Hopfield pazzesca. Un @_pieroit_ incontenibile. Grazie davvero per qsti contenuti, valore altissimo. Mi auguro di vederti in live spesso.
1
1
1
1,720
È disponibile la Rassegna n. 17! Su app Android e iOS: rassegnasocial.com 1 - UNA PACE GIUSTA @LucaLovisolo 2 - SI SONO TRASFORMATI IN MOSTRI @BielorussaRM 3 - MATTARELLA: DISINFORMAZIONE RUSSA IN ITALIA @_Brick_Block_ 4 - AI: LA VISIONE DEL PAPA @TV2000it Segue 👇
1
1
2
1,676
"When the AI purrs: Cheshire Cat AI, Production-ready AI Assistant Framework" by @alesspallina #CheshireCatAI @PythonTorino — Journey to the Center of the AI #llm #aimemory #rag #ccatplugin #ai #python @_pieroit_
3
4
1,697
Piero Savastano retweeted
I just bought a 2024 Chevy Tahoe for $1.
396
4,948
791
91,373
20,332,584
Piero Savastano retweeted
Talking to good old friends is always an immense pleasure 🫂 This time, @_pieroit_ invited @LukawskiKacper to his podcast to speak about the role of vector embeddings in modern AI applications. Please check out the recordings! buff.ly/3uXNJo4
1
3
6
950
Piero Savastano retweeted
Replying to @acor3 @_pieroit_
Grazie per il suggerimento, davvero notevole 👏 Finalmente un progetto open source, italiano, nell’Intelligenza Artificiale generativa. – at Rome, Lazio
1
2
148
mitico @_pieroit_ ed il suo progetto dello Stregatto, non lo conoscevo e sembra super interessante. Grazie Marco e Cesare @PythonMilano #python #ai
Ep 45 Intervista Pythonista con Piero Savastano e lo Stregatto!
1
1
2
299
Piero Savastano retweeted
📹 Ho fatto una bella chiacchierata con @_pieroit_ data scientist e sviluppatore machine learning, sul presente e sul futuro dell’#intelligenzaartificiale generativa. Open vs Close, AGI, GPTs e del suo progetto Stregatto. youtu.be/rGKkIMHcFQI
2
3
362
Piero Savastano retweeted
The more powerful the models become, the more programs they store. And the benefits of finding the right program (by figuring out the right prompt) become even greater. Unless LLMs change substantially in the future, prompt engineering is not going anywhere. What's going to happen, however, is that it will likely get automated so the end user doesn't have to deal with it. Better program search.
13
29
3
374
79,882
Piero Savastano retweeted
With many 🧩 dropping recently, a more complete picture is emerging of LLMs not as a chatbot, but the kernel process of a new Operating System. E.g. today it orchestrates: - Input & Output across modalities (text, audio, vision) - Code interpreter, ability to write & run programs - Browser / internet access - Embeddings database for files and internal memory storage & retrieval A lot of computing concepts carry over. Currently we have single-threaded execution running at ~10Hz (tok/s) and enjoy looking at the assembly-level execution traces stream by. Concepts from computer security carry over, with attacks, defenses and emerging vulnerabilities. I also like the nearest neighbor analogy of "Operating System" because the industry is starting to shape up similar: Windows, OS X, and Linux <-> GPT, PaLM, Claude, and Llama/Mistral(?:)). An OS comes with default apps but has an app store. Most apps can be adapted to multiple platforms. TLDR looking at LLMs as chatbots is the same as looking at early computers as calculators. We're seeing an emergence of a whole new computing paradigm, and it is very early.
290
1,830
404
9,088
2,139,708
Gli si vuole sempre bene a Piero
1
2
145
Piero Savastano retweeted
I lost this livestream today. Didn't reach me . So want to spread it myself 😊 @_pieroit_
The free workshop “Efficient Fine-Tuning for Llama-7b on a Single GPU," is just around the corner! Drop any question or suggestion for the speakers in the comments and we’ll take them into consideration. 🖥️ See you in two hours! hubs.la/Q020xZrQ0
1
1
163
Piero Savastano retweeted
We are excited to share the Cheshire Cat - an LLM framework. Its unique approach results in vectorizing documents, conversations, and tools. We are incredibly proud that Cat chose Qdrant as their vector search backend! github.com/cheshire-cat-ai/c…
2
2
1
25
2,781
Piero Savastano retweeted
Appassionato, eclettico e preparato; @_pieroit_ è uno spettacolo sentirti parlare di AI. 💜
4
1
10
288
Piero Savastano retweeted
🧠 Su @startup_italia, in una chiacchierata con @_pieroit_, parliamo di modelli di linguaggio di grandi dimensioni (#LLM) e del bilancio tra la protezione dei dati degli utenti e le performance di questi sistemi. #AI #DeepLearning #ChatGPT #GPT4 startupitalia.eu/202155-2023…
1
2
3
248