Apple's Leap into On-Device AI with OpenELM Language Models


Jim Miller

Apple Unveils OpenELM: Compact AI Language Models for On-Device Use

Welcome back to AI Hungry, where we dive into the latest breakthroughs shaping the future of artificial intelligence. In this update, we explore Siemens' launch of a generative AI assistant for industrial engineering, aiming to revolutionize automation projects. We'll also delve into Apple's unveiling of OpenELM, a suite of compact AI language models tailored for on-device use, signaling a leap forward in mobile AI technology. These stories highlight the continuous integration of AI into various sectors, promising advancements that could redefine efficiency and user experience.


Siemens Introduces Generative AI Assistant for Industrial Engineering

Siemens has launched a new generative AI assistant, Siemens Industrial Copilot, designed to enhance industrial engineering. Revealed at Hannover Messe in Germany, the product integrates with Siemens' automation portal to assist engineers by generating code and reducing repetitive tasks. This innovation aims to cut development times, minimize errors, and boost productivity in automation projects.

The Siemens Industrial Copilot, set to be available in the Xcelerator marketplace, is a collaboration with Microsoft, utilizing Azure OpenAI Service. It represents Siemens' commitment to advancing AI in industrial processes, with a vision to create a series of AI copilots that will revolutionize operations across various industries, including automotive and healthcare.

Main course

Apple Unveils OpenELM: Compact AI Language Models for On-Device Use

Apple has introduced OpenELM, a collection of small AI language models designed to run on smartphones, marking a significant step towards on-device AI capabilities. These models range from 270 million to 3 billion parameters, indicating a variety of complexities and potential uses. OpenELM models are available in two types: pretrained and instruction-tuned, the latter being more suitable for AI assistants and chatbots. The models are trained on a substantial dataset and utilize a layer-wise scaling strategy for efficient parameter allocation, which has led to improved performance.

In addition to the models, Apple has released CoreNet, a library for training OpenELM, and reproducible training recipes, which is a move towards greater transparency in AI research. While these models are not yet part of Apple's consumer offerings, there are hints that the upcoming iOS 18 update may incorporate new AI features that leverage on-device processing, enhancing user privacy and potentially boosting Siri's capabilities.


๐Ÿง  Google Cloud's John Abel Discusses the Impact of Generative AI on Business Intelligence. Google Cloud's John Abel highlights how generative AI is transforming business intelligence, making data analytics more accessible and driving innovation across industries by enabling dynamic, data-driven decision-making. (Link)

๐Ÿš€ NVIDIA Acquires AI Workload Optimization Firm Run:ai for Enhanced Capabilities. NVIDIA is set to acquire Run:ai, a company specializing in AI workload management, to boost its AI technology and diversify its offerings. The deal, estimated at $700 million, will integrate Run:ai's efficiencies into NVIDIA's AI platforms. (Link)

๐Ÿ‘“ Meta Introduces AI-Powered Ray-Ban Smart Glasses with Voice Commands. Meta has upgraded its Ray-Ban smart glasses with an AI assistant, enabling voice-activated information and translations. Priced at $300, the glasses are currently available in the U.S. and Canada with a wider release planned soon. (Link)

๐Ÿค– Synthesia Unveils 'Expressive Avatars' Revolutionizing Digital Communication. Synthesia's new 'Expressive Avatars' can mimic human emotions for corporate videos and presentations. Their AI model, used by top companies, creates realistic avatars, enhancing business communication and training experiences. (Link)

๐Ÿ” UK Regulator Probes Big Tech's AI Moves for Anti-Competitive Concerns. The UK's Competition and Markets Authority is investigating Microsoft and Amazon's recent AI partnerships and acquisitions, examining their impact on competition and innovation in the AI industry. (Link)

๐Ÿ’„ Estรฉe Lauder and Microsoft Unveil AI Innovation Lab for Beauty Industry Advancements. Estรฉe Lauder and Microsoft have launched the AI Innovation Lab to harness generative AI for the beauty industry, aiming to enhance customer experiences and accelerate product innovation through Azure OpenAI services. (Link)

๐Ÿ’ก U.S. Government Grants $6.1 Billion to Micron for Semiconductor Production. The Biden-Harris Administration has awarded Micron Technology Inc. a $6.1 billion grant to boost U.S. semiconductor production. This is part of a larger effort under the CHIPS and Science Act to enhance American tech innovation and security. (Link)

Enjoying this newsletter?

Subscribe to get more content like this delivered to your inbox for free.