🚨 NEWS: Addestrare un modello AI da zero per 1.500 dollari è possibile

Ecco i punti chiave in breve:
💡 Una startup ha aperto un varco nell'economia dell'intelligenza artificiale. Sapient Intelligence ha addestrato da zero un modello linguistico fondamentale da 1 miliardo di parametr...

🚀 LINK: https://meteoraweb.com/news/addestrare-un-modello-ai-da-zero-per-1500-dollari-e-possibile-hf8on

#hRMText #sapientIntelligence #addestramentoAI #foundationModel #aIABassoCosto

Addestrare un modello AI da zero per 1.500 dollari è possibile

Sapient Intelligence addestra un modello AI da 1 miliardo di parametri con 1.500 dollari. Performance competitive senza supercomputer. La democratizzazione del machine learning

Meteora Web
#Apple announced a new #FoundationModel|s #framework with image input support, custom skills, and server-side model execution. #Xcode enhancements include app localisation, simulated device interaction, and expanded App Intents support. https://www.macrumors.com/2026/06/08/apple-unveils-xcode-and-models-improvements/?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI
Apple Unveils Xcode and Foundation Models Framework Improvements

Apple today announced a new Foundation Models framework for developers alongside a set of Xcode enhancements aimed at agentic coding workflows. The...

MacRumors

First Law of the LLMs:

1) No LLM shall be configured to take the rights of the many’s interests to serve those of the few
2) ….

Can they be written from a training set to avoid the paradox of Asimov’s Robot laws?

#Asimov #Foundation #FoundationModels #FoundationModel #paradox #Models #EthicalLaws #Philosophy #Dialectic #LLM #Ai

Infomaniak secures its independence and its DNA for the long term • Infomaniak

Infomaniak transfers the majority of its voting rights to a Swiss public-benefit foundation. Your cloud remains sovereign, independent, and true to its values. Forever.

Infomaniak Network News

RT @Zai_org: GLM-5V-Turbo Tech Report: Auf dem Weg zu einem nativen Foundation-Modell für multimodale Agenten

mehr auf Arint.info

#AgentFramework #FoundationModel #GLM5V #MultimodalAI #ReinforcementLearning #TechReport #arint_info

https://x.com/Zai_org/status/2052426777654387168#m

Arint - SEO+KI (@[email protected])

<p>RT @Zai_org: GLM-5V-Turbo Tech Report: Auf dem Weg zu einem nativen Foundation-Modell für multimodale Agenten</p> <p><a href="https://arint.info/@Arint/116544045547793628">mehr</a> auf <a href="https://arint.info/">Arint.info</a></p> <p>#AgentFramework #FoundationModel #GLM5V #MultimodalAI #ReinforcementLearning #TechReport #arint_info</p> <p><a href="https://x.com/Zai_org/status/2052426777654387168#m">https://x.com/Zai_org/status/2052426777654387168#m</a></p>

Mastodon Glitch Edition

RE: https://mastodon.social/@thetransmitter/116476879384094089

Tl;dr: transformers didn’t learn the underlying principle.

“Keyon Vafa and his colleagues simulated tens of millions of planetary orbits from Newton’s laws and trained a transformer on the resulting sequences. The model predicted future positions with very high accuracy. But when they fine-tuned it to infer the underlying gravitational force vectors, the model produced nonsense. The implied laws of gravitation were different depending on which subset of data the researchers examined. The transformer had assembled a patchwork of heuristics, accurate for every solar system in the training set, but it hadn’t discovered the universal gravitational principle. Without that principle, the model could predict the movement of points of light in the sky but could never send a rocket to the moon.”

Also link to the preprint: https://arxiv.org/abs/2507.06952

#neuroscience #prediction #FoundationModel

Ich arbeite an einem kleinen Nebenprojekt für #macos: ich scanne oft Konvolute von Papier und muss die Dokumente wieder vereinzeln aus einem PDF. Ich nutze gerade ein #foundationmodel von Apple zur Erkennung der einzelnen Dokument. Hat jemand einen ähnlichen anwendungsfall und würde testen wollen? #mac #macos
#OpenAI launched #GPT54, a new #foundationmodel optimised for #professionalwork. The model comes in three versions: standard, reasoning (GPT-5.4 Thinking), and high-performance (GPT-5.4 Pro). It boasts improved benchmark results, token efficiency, and error reduction compared to its predecessor. https://techcrunch.com/2026/03/05/openai-launches-gpt-5-4-with-pro-and-thinking-versions/?eicker.news #tech #media #news
OpenAI launches GPT-5.4 with Pro and Thinking versions | TechCrunch

GPT-5.4 is billed as "our most capable and efficient frontier model for professional work."

TechCrunch

MBZUAI's Viện Mô Hình Nền Tảng (Institute of Foundation Models) ra mắt mô hình K2-V2, đạt hiệu năng vượt trội dù có mức độ mở ngang bằng Olmo 3 trong cuộc đua mô hình ngôn ngữ lớn (LLM).

#AI #MBZUAI #K2V2 #MôHìnhNgônNgữLớn #FoundationModel #MôHìnhFoudation #OpenSource #CôngNghệAI

https://www.reddit.com/r/singularity/comments/1ppwbgf/mbzuais_institute_of_foundation_models_enters_the/

Un modello di base, addestrato su 3M di giorni‑persona di dati Apple Watch, prevede condizioni mediche con accuratezza.

Si basa su JEPA, che apprende a inferire parti mancanti 📈⌚

#applewatch #medtech #foundationmodel

https://kiro.it/4rq7S