Google just officially released Gemma 2, introducing ultra-efficient 9B and 27B open models for developers. The 27B variant punches above its weight by matching models twice its size while running full precision on a single Nvidia H100, and the 9B edges out Llama 3 8B. It is available now under a commercial-friendly license.
Meta has reportedly discussed integrating its generative AI models into Apple Intelligence alongside OpenAI's ChatGPT. Apple has also held talks with Google, Anthropic, and Perplexity, potentially creating a hub where AI companies sell premium subscriptions directly to iPhone users. Negotiations are still ongoing and could fall through.
Apple just confirmed it won't roll out Apple Intelligence, iPhone Mirroring, or SharePlay screen sharing to EU users this year due to regulatory uncertainties from the Digital Markets Act. The company claims the DMA's strict interoperability rules could force compromises on user privacy and data security. Apple is working with the European Commission to find a workaround.
Microsoft just released Florence-2 on Hugging Face under a permissive MIT license, cramming multi-task vision capabilities into 232M and 771M parameter models. It uses text prompts to handle object detection, captioning, and segmentation, outperforming models a hundred times its size in zero-shot tests. This gives developers a single, lightweight tool instead of juggling multiple specialized vision pipelines.
Microsoft just officially launched its new Surface Laptop and Surface Pro Copilot+ PCs starting at $999. Powered by Snapdragon X series processors and dedicated NPUs, the machines enable on-device experiences like real-time Cocreator image generation and system-wide Live Captions. It marks Microsoft's first major hardware push to make on-device Windows AI a practical reality.
Apple just released 20 new Core ML models and four datasets on Hugging Face, including FastViT and DepthAnything. Built specifically to run on-device without an internet connection, they let developers build fast computer-vision features while keeping user data strictly private.
The Qwen team just launched Qwen2, an open-source model family spanning five sizes from 0.5B to 72B parameters. Trained on 27 additional languages, the flagship Qwen2-72B goes toe-to-toe with Llama-3-70B across major benchmarks, while context lengths stretch up to 128K tokens. Except for the 72B tier, all smaller models are now released under the Apache 2.0 license.
Apple just sent out official invites for its WWDC 2024 keynote on June 10, setting the stage for its big leap into consumer AI. The company is widely expected to unveil iOS 18, with reports pointing to AI upgrades across photo editing, messaging, email, and a significantly smarter Siri.
Google just rolled out a suite of AI features for its Chromebook Plus lineup, including system-wide writing help, AI-generated wallpapers, and the Magic Editor in Google Photos. Even better: buying a new device gets you a full year of the Google One AI Premium plan for free, unlocking Gemini Advanced.
The UK data watchdog just launched an inquiry into Microsoft's new Recall feature, which takes screenshots of Copilot+ PC screens every few seconds to build a searchable history. While Microsoft insists all data stays encrypted locally and is strictly opt-in, privacy experts warn that capturing passwords and sensitive records without filtering poses a serious security risk.
Microsoft just made its lightweight Phi-3 model family available to developers, bringing Phi-3-mini to Azure AI alongside small and medium versions. The company also previewed Phi-3-vision, a 4.2-billion-parameter multimodal model built for chart and image reasoning. Why it matters: Powerful compact models like this pave the way for real-world AI PC deployments.
Microsoft just unveiled Recall for Copilot+ PCs at Build 2024, using AI to fix Windows' notoriously clunky search. Running over 40 local multimodal small models, it lets you search virtually anything you've seen—from past emails to specific meeting slides—using vague contextual clues. Everything runs offline on-device, rolling out on June 18.
Apple is reportedly teaming up with OpenAI to integrate the startup's chatbot technology into iOS 18, with a formal announcement planned for WWDC on June 10. While Apple plans to use homegrown models for on-device and cloud tasks to upgrade Siri, it lacks a true ChatGPT equivalent. Borrowing OpenAI's engine gives Apple an immediate answer to user demand while it races to catch up.
Google just overhauled Android with a suite of new AI capabilities powered by Gemini and Circle to Search. Students can now circle math and physics problems for step-by-step explanations, while Gemini's new overlay can answer questions about YouTube videos. Even better, multimodal Gemini Nano will run on-device to spot phone scams in real time.
Google just dropped a teaser for an unreleased multimodal AI interface running on Android, right after OpenAI launched GPT-4o. The video shows the assistant using a phone's camera to understand its surroundings in real time and answer questions about the Google I/O stage setup. OpenAI may have had the spotlight for only 24 hours, though real-world performance remains to be seen at the keynote.
Apple just rolled out a major update to Logic Pro for Mac and iPad, packing new AI features into its music production suite. Creators get new AI Bass and Keyboard players to jam alongside Drummer, plus a Stem Splitter tool that breaks mixed audio files into separate tracks. The catch: flagship tools like Stem Splitter require Apple Silicon (M1 or newer).
Google just unveiled the Pixel 8a, packing its Tensor G3 chip and signature AI features into a $499 device. It brings onboard Gemini, Circle to Search, and camera tools like Best Take and Audio Magic Eraser. Why it matters: Google is democratizing its mobile AI suite without forcing buyers into pricey flagship territory.
Apple just unveiled its next-generation M4 chip, debuting inside the redesigned iPad Pro. Built on second-generation 3nm technology, the processor features Apple's fastest Neural Engine yet, claiming to outpace the NPU of any AI PC today. Along with hardware-accelerated ray tracing and dynamic caching, Apple is positioning the tablet as an on-device AI powerhouse.
Apple is reportedly testing a generative AI photo editing feature dubbed "Clean Up" for its next-generation operating systems. The tool replaces the existing Retouch brush, allowing users to paint over and remove larger unwanted objects directly inside the Photos app. It aims to bring Photoshop-level cleanup directly to everyday users, though it remains in internal testing.
Apple is reportedly preparing a suite of AI features for iOS 18 powered by its Ajax LLM, including text summarization in Messages and Safari. Simple tasks will run on-device with strict privacy warnings, while complex queries could tap server-side power from Google or OpenAI. Details are expected at WWDC on June 10th.
