#Mastodon updates its terms to prohibit #AI #modeltraining
“We explicitly prohibit the scraping of user data for unauthorized purposes, e.g. archival or large language model (#LLM) training. We want to make it clear that training LLMs on the data of Mastodon users on our instances is not permitted,” Mastodon said in an email sent to users.
https://techcrunch.com/2025/06/17/mastodon-updates-its-terms-to-prohibit-ai-model-training/

Mastodon, #socialmedia where YOU are not the product!

ICYMI: Copyright debate intensifies over AI training data use https://ppc.land/copyright-debate-intensifies-over-ai-training-data-use/ #Copyright #AITechnology #ModelTraining #FairUse #IntellectualProperty

ICYMI: Copyright debate intensifies over AI training data use: Analysis of Andreessen Horowitz's position on AI model training and fair use in response to US Copyright Office inquiry https://ppc.land/copyright-debate-intensifies-over-ai-training-data-use/ #Copyright #AITechnology #ModelTraining #FairUse #IntellectualProperty

A few tips for optimizing Pytorch model training time from a Yandex ML engineer.

https://alexdremov.me/simple-ways-to-speedup-your-pytorch-model-training/

Sources reveal Apple & Baidu are collaborating to add AI features to Chinese iPhones, but they've clashed over model training data. The models are also facing issues with prompts. #Apple #Baidu #AI #iPhone #TechNews #ArtificialIntelligence #China #ModelTraining #AIChallenges

HIRING: Senior Software Engineer, Quantized Training / US, WA, Seattle
USD 180K+

https://aijobs.net/J539746/

HIRING: AIML - Senior Software Engineer, ML Systems and Evaluation Engineering / Cupertino, California, United States
USD 143K+

https://aijobs.net/J557370/

Google introduces #Gemma, a lightweight open AI model. It was designed using the same technology behind @gemini AI models from Google.
The company has made Gemma available in two different sizes — Gemma 2B and Gemma 7B. Both of these products are equipped with pre-trained and instruction-tuned variants and are sufficiently lightweight to be run directly on a developer's laptop or desktop. Gemma can be used to create AI-powered applications.

https://www.engadget.com/google-introduces-a-lightweight-open-ai-model-called-gemma-130053289.html

Just Posted: This LinkedIn article explores how Broadcom's Jericho-3AI and Tomahawk networking fabrics can accelerate model training and overcome networking challenges in AI initiatives, leading to faster completion of training and cost savings. #AIFabric #AIModels #AINetworking #ModelTraining #NetworkIO #NFD32
https://gestaltit.com/tech-field-day/sulagna/accelerated-model-training-with-broadcom-jericho-3ai-and-tomahawk-family/
#AIFabric #AIModels #AINetworking #ModelTraining #NetworkIO #NFD32

So lemme get this straight: If I know that someone is sending to or from a Gmail address, I'll soon be able to ask their AI system what that person is discussing?

#ModelTraining
#NotThoseKindsOfModels

#gpt4 #thematrix #redpill #simulations #llms #ai #openai #modeltraining CONVERSATION WITH GPT4 (Patent Kinetics)
https://www.patentkinetics.com/2023/05/conversation-with-gpt4.html

Brett on how “better data beats math” WASNT the full-story #datadaytexas #ddtx2023 #modeltraining #datascience #datacentricai

Next up: Bigger Data beats Better Math, No Question by
Brent Schneeman

This session explores the trade-offs between ML model architectures, the quantity of training data, and the quality of training data.