Last Week in AI - A Weekly Unwind
From 20-April-2024 to 26-April-2024
It was yet another thrilling week in the AI field with advancements that further extend the limits of what can be achieved with AI.
Here are 10 AI breakthroughs that you can’t afford to miss 🧵👇
15T Tokens of the Finest Data the Web Has to Offer 🌐
Llama 3 models are trained on over 15 trillion tokens. Seems impossible to get such a huge dataset right? It’s not though. FineWeb is a dataset consisting of more than 15 trillion tokens of cleaned and deduplicated English web data from CommonCrawl dumps since 2013. This dataset not only surpasses the previously available RefinedWeb in terms of cleaning and deduplication processes but the models trained on this dataset outperform those trained on other datasets. It is made freely available to encourage ongoing development and research in AI.
Microsoft to Invest $1.5B in a UAE-based AI Company 🤝
Microsoft has announced a $1.5 billion investment in G42, an AI firm based in the United Arab Emirates. This financial move involves G42 expanding its use of Microsoft Cloud and fully integrating its AI services with Microsoft Azure to co-develop AI solutions tailored to sectors like healthcare, government, and energy. The collaboration will focus on developing advanced AI solutions across various industries in the underserved areas of the Middle East, Central Asia, and Africa, utilizing Microsoft Azure.
Microsoft's Phi-3 Beats Llama 3 8B 💪
Microsoft has released Phi-3 family of models, the next generation of Microsoft’s small but mighty models. This model is small enough to run directly on your smartphone, packing an impressive performance into a compact size. Despite just 3.8 billion parameters, the mini model competes with larger models like Llama 3 8B, Mistral 7B, Gemma 7B, Mixtral 8x7B, and even GPT 3.5 in some tasks.
Meta Quest OS Now Available to Third-party Hardware Makers 👓
Meta has announced a major shift in its strategy, opening up its operating system that powers its Meta Quest headsets, to other hardware makers. This means we can expect a wider variety of devices to choose from, all running on the same platform and able to access the same apps and experiences. This move can make mixed reality more accessible and appealing to a wider audience. Companies like ASUS, Lenovo, and even Xbox are already working on new headsets powered by Meta’s OS, called the Meta Horizon OS.
The Autonomous Vehicle That Communicates as It Drives 🚘
Wayve has introduced LINGO-2, an advanced AI-powered driving model that enhances transparency in autonomous vehicles by explaining its decisions in real-time using natural language. This model improves upon its predecessor, LINGO-1, by not only commenting on driving scenarios but actively controlling the car and providing relevant commentary as it drives. LINGO-2 also responds to user commands and adapts its behavior accordingly, allowing for interactive and customizable driving experiences. It’s designed to foster trust and acceptance by answering questions about its actions and the driving environment, making autonomous driving more accessible and understandable.
Adobe releases Firefly Image 3 AI Model 🦋
Adobe has released Adobe Firefly Image 3 Foundation Model, its next generation of creative generative AI model. Image 3 not only delivers high-quality images but also helps you ideate fast with prompt auto-complete suggestions and generates images within seconds. It features better lighting, more precise positioning, and a greater variety of moods and expressions in people’s rendering. It also lets you transfer style and structure from a reference image. The model is available to use in Photoshop and the Firefly web application.
Snowflake’s Generative AI Model with Just $2 Million 🦾
Building and using powerful AI models for businesses has always been very expensive and resource-intensive, costing millions of dollars. Snowflake has developed its own model, “Snowflake Arctic” with 480B parameters, trained with a budget of just $2 million! Arctic is designed for tasks that businesses care about, like generating code, writing SQL queries, and following complex instructions. It has a unique hybrid of Mixture of Experts (MoE) and a dense model, bringing the best of both architectures. This makes it way more affordable to train and use than other models out there.
UK probes into Big AI Companies like Microsoft and Amazon 🧑⚖️
As Microsoft and Amazon intensify their investments and partnerships in the AI sector, with Microsoft backing OpenAI and recruiting top executives from Inflection AI, and Amazon deepening its stakes in Anthropic, the UK's Competition and Markets Authority (CMA) is investigating these deals to ensure they do not stifle competition or unfairly concentrate power in the hands of a few, focusing particularly on ‘foundation models’ critical to diverse AI systems. This scrutiny aims to preserve competition and innovation in the UK's AI industry, with possible outcomes ranging from approval to legal action against these tech giants.
New Contender in Humanoid Robotics for Everyday Tasks 🦿
Mentee Robotics has introduced a bipedal humanoid robot, MenteeBot, designed for both residential and industrial environments. This robot incorporates advanced AI to enable mobility, cognitive skills, and the ability to learn tasks not initially programmed. It can take instruction from its owner and communicate effectively, utilizing an LLM. Scheduled for limited release by the end of 2024, MenteeBot can perform a variety of tasks, operates up to five hours on a single charge, and has human-like mobility and hand functions, setting it apart as primarily a household assistant.
Apple’s LM That Might Power Generative AI on iPhones 📱
Apple has released OpenELM, a series of open-source small language models designed to run on devices like iPhones. The models come in four variants—270M, 450M, 1.1B, and 3B parameters—using a technique called “layer-wise scaling” to deliver high performance efficiently. Comparative assessments reveal that while the OpenELM-1.1B model outperforms competitors in some standard tasks, its real-world applicability is in question, as shown by its lower scores in the MMLU metric. Running AI models like OpenELM locally on devices is important for data privacy, keeping user data on the device. It also ensures that AI features are available offline.
Which of the above AI development you are most excited about and why?
Tell us in the comments below ⬇️
That’s all for today 👋
Stay tuned for another week of innovation and discovery as AI continues to evolve at a staggering pace. Don’t miss out on the developments – join us next week for more insights into the AI revolution!
Click on the subscribe button and be part of the future, today!
📣 Spread the Word: Think your friends and colleagues should be in the know? Click the ‘Share’ button and let them join this exciting adventure into the world of AI. Sharing knowledge is the first step towards innovation!
🔗 Stay Connected: Follow us for AI updates, sneak peeks, and more. Your journey into the future of AI starts here!
Shubham Saboo - Twitter | LinkedIn ⎸ Unwind AI - Twitter | LinkedIn