Google anticipa i tempi e aggiorna Gemini 2.5 Pro (I/O Edition). Il migliore modello nel coding secondo la WebDev Arena e Lmarena

https://gomoot.com/google-anticipa-i-tempi-e-aggiorna-gemini-2-5-pro-i-o-edition

Neo4j treibt mit GraphRAG, Vektor-Indizes & Agentic RAG die #KI-Entwicklung voran. Ob #LangChain, #LlamaIndex, #SpringAI oder #VertexAI – das neue Python-Paket und das Model Context Protocol (MCP) verknüpfen Graphdaten nahtlos mit #LLM-Anwendungen.
#Neo4j #GenAI #RAG #GraphQL #Cypher #VectorSearch #AgenticAI
https://www.bigdata-insider.de/leistungssprung-bei-graph-datenbanken-mit-ki-integration-cloud-skalierung-und-terabyte-graphen-a-2307ed20cfaf562a1a0094b712b5be95/

During my holidays break I experimented with a #Bluesky topic trend visualization with #D3js using a #VertexAI embedding model, DBSCAN clustering to gather common topics, #Gemini to summarize clusters of posts.

All implemented in #java using #langchain4j

Oggi (giovedì 12) alle 18.30 alle Officine Credem, presso il Tecnopolo Reggio Emilia, il GDG Cloud Modena organizzerà un evento per illustrare *Vertex AI* e la creazione di agenti con questa soluzione by Google.

Maggiori dettagli e iscrizione a questo link: https://gdg.community.dev/events/details/google-gdg-cloud-modena-presents-costruire-agenti-con-vertex-ai-buffet-e-networking/

Lots of new cool Gemini stuff in LangChain4j 0.35.0

#ai #gemini #google #java #lang4chain #vertexai

https://glaforge.dev/posts/2024/09/29/lots-of-new-cool-gemini-stuff-in-langchain4j/?utm_medium=erik.in&utm_source=mastodon

I updated my #LLM token visualization application with the latest #googlecloud #VertexAI preview embedding models
https://tokens-lpj6s2duga-ew.a.run.app/

neato visualization of #ai #LLMs

#vertexai #chatgpt

https://bbycroft.net/llm

Great set of example
generative AI powered use cases for data engineers using #bigquery and #vertexai
https://youtube.com/live/F2Wn5yIUO7s?feature=shared

Should Mastodon add a robots.txt option to the admin dashboard?

Web Sites Can Now Opt Out of #Google #Bard and Future #AI Models via robots.txt with the User agent token "Google-Extended"

User-agent: Google-Extended
Disallow: /

Overview of Google crawlers and fetchers (user agents) https://developers.google.com/search/docs/crawling-indexing/overview-google-crawlers

An update on web publisher controls https://blog.google/technology/ai/an-update-on-web-publisher-controls/

Web Sites Can Now Choose to Opt Out of Google Bard and Future AI Models https://tech.slashdot.org/story/23/10/01/046215/web-sites-can-now-choose-to-opt-out-of-google-bard-and-future-ai-models #MastoAdmin #VertexAI

Take me to #googlecloudnext , show me all the AI things, give me a long weekend and apparently I create a starter #AI slackbot.

https://github.com/jeffbryner/gcp-ai-slackbot
#vertexai

Building an LLM-based application? Here's a tech stack that might help you. #Redis #vertexai

https://redis.com/blog/building-llm-applications-with-redis-on-googles-vertex-ai-platform/

Pretty amazing how easy google #colabNotebook makes it to take #vertexai #bison and #palmapi for a spin
https://colab.research.google.com/github/GoogleCloudPlatform/generative-ai/blob/main/language/getting-started/intro_palm_api.ipynb

I am, admittedly, bad at adding alt text to my posts on #Pixelfed. As I was falling asleep last night, I had an idea -- I already have an #n8n workflow that takes my posts from a Google Sheet that's fed by a Google Form, letterboxes the images to fit on Instagram, and posts them to both Instagram and Pixelfed so why not just add steps to use an AI to caption them?

Woke up this morning, settled on using Google Vertex AI, built out the steps, and it worked brilliantly. Until I fed it an image of two men in shorts on the boardwalk which seems to just be waaaaayyyy too much flesh for Google.

First experimentation with #VertexAI today. Testing what a minimal sentiment labelling of #UNSC speeches can provide in valid outputs on unlabelled data.

The fun of the black box of machine learning combined with the fun of our “UN Security Council Debates" corpus, available for your own experiments and analysis here:
https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/KGVSYH