Mistral Large: The New Rival to GPT-4
Plus: DeepMind’s new gen-AI model creates video games in a flash, Meta’s MobileLLM enables on-device AI deployment.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 219th edition of The AI Edge newsletter. This edition brings you “Mistral Large: The New Rival to GPT-4.”
And a huge shoutout to our incredible readers. We appreciate you😊
In today’s edition:
🌪️ Mistral Large: The new rival to GPT-4, 2nd best LLM of all time
🎮 DeepMind’s new gen-AI model creates video games in a flash
📱 Meta’s MobileLLM enables on-device AI deployment
💡 Knowledge Nugget: The AI battle rages on by
Let’s go!
Mistral Large: The new rival to GPT-4, 2nd best LLM of all time
The French AI startup Mistral has launched its largest-ever LLM and flagship model to date, Mistral Large, with a 32K context window. The model has top-tier reasoning capabilities, and you can use it for complex multilingual reasoning tasks, including text understanding, transformation, and code generation.
Due to a strong multitasking capability, Mistral Large is the world's second-ranked model on MMLU (Massive multitask language understanding).
The model is natively fluent in English, French, Spanish, German, and Italian, with a nuanced understanding of grammar and cultural context. In addition to that, Mistral also shows top performance in coding and math tasks.
Mistral Large is now available via the in-house platform "La Plateforme" and Microsoft's Azure AI via API.
Why does it matter?
Mistral Large stands out as the first model to truly challenge OpenAI's dominance since GPT-4. It shows skills on par with GPT-4 for complex language tasks while costing 20% less. In this race to make their models better, it’s the user community that stands to gain the most. Also, the focus on European languages and cultures could make Mistral a leader in the European AI market.
DeepMind’s new gen-AI model creates video games in a flash
Google DeepMind has launched a new generative AI model - Genie (Generative Interactive Environment), that can create playable video games from a simple prompt after learning game mechanics from hundreds of thousands of gameplay videos.
Developed by the collaborative efforts of Google and the University of British Columbia, Genie can create side-scrolling 2D platformer games based on user prompts, like Super Mario Brothers and Contra, using a single image.
Genie can be prompted with images it has never seen before, such as real-world photographs or sketches, enabling people to interact with their imagined virtual worlds-–essentially acting as a foundation world model. This is possible despite training without any action labels.
Why does it matter?
Genie creates a watershed moment in the generative AI space, becoming the first LLM to develop interactive, playable environments from a single image prompt. The model could be a promising step towards general world models for AGI (Artificial General Intelligence) that can understand and apply learned knowledge like a human. Lastly, Genie can learn fine-grained controls exclusively from Internet videos, a unique feature as Internet videos do not typically have labels.
Meta’s MobileLLM enables on-device AI deployment
Meta has released a research paper that addresses the need for efficient large language models that can run on mobile devices. The focus is on designing high-quality models with under 1 billion parameters, as this is feasible for deployment on mobiles.
By using deep and thin architectures, embedding sharing, and grouped-query attention, they developed a strong baseline model called MobileLLM, which achieves 2.7%/4.3% higher accuracy compared to previous 125M/350M state-of-the-art models. The research paper highlights that you should concentrate on developing an efficient model architecture rather than on data and parameter quantity to determine model quality.
Why does it matter?
With language understanding now possible on consumer devices, mobile developers can create products that were once hard to build because of latency or privacy issues when reliant on cloud connections. This advancement allows industries like finance, gaming, and personal health to integrate conversational interfaces, intelligent recommendations, and real-time data privacy protections using models optimized for mobile efficiency, sparking creativity in a new wave of intelligent apps.
Enjoying the daily updates?
Refer your pals to subscribe to our daily newsletter and get exclusive access to 400+ game-changing AI tools.
When you use the referral link above or the “Share” button on any post, you'll get the credit for any new subscribers. All you need to do is send the link via text or email or share it on social media with friends.
Knowledge Nugget: The AI battle rages on
In this article,
explores the potential impact of AI, particularly OpenAI's DALL·E 2, on creative jobs and the broader job market. The author suggests that DALL·E 2 marks the beginning of the automation of creative jobs, as it can produce high-quality images quickly and at a significantly lower cost than human designers.The article also discusses the broader implications of AI in fields such as software development, writing, and various business-related jobs, predicting that AI-powered algorithms will increasingly replace human labor in these areas. It will lead to significant disruptions in the job market and potentially benefit companies like OpenAI and Google while raising concerns about the displacement of creative professionals.
Why does it matter?
As AI creative tools like DALL-E 2 become more advanced and widely available at ultra-low costs, businesses could replace large numbers of human graphic designers, writers, and other creatives with AI. This could allow companies to dramatically cut costs but also puts the livelihoods of millions of creative professionals at risk. It remains to be seen whether new jobs will emerge to offset these losses.
What Else Is Happening❗
🤖 Qualcomm reveals 75+ pre-optimized AI models at MWC 2024
Qualcomm released 75+ new large language models, including popular generative models like Whisper and Stable Diffusion, optimized for the Snapdragon platform at the Mobile World Congress (MWC) 2024. The company stated that some of these LLMs will have generation AI capabilities for next-generation smartphones, PCs, IoT, XR devices, etc. (Link)
💻 Nvidia launches new laptop GPUs for AI on the go
Nvidia launched RTX 500 and 1000 Ada Generation laptop graphics processing units (GPUs) at the MWC 2024 for on-the-go AI processing. These GPUs will utilize the Ada Lovelace architecture to provide content creators, researchers, and engineers with accelerated AI and next-generation graphic performance while working from portable devices. (Link)
🧠 Microsoft announces AI principles for boosting innovation and competition
Microsoft announced a set of principles to foster innovation and competition in the AI space. The move came to showcase its role as a market leader in promoting responsible AI and answer the concerns of rivals and antitrust regulators. The standard covers six key dimensions of responsible AI: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. (Link)
♊ Google brings Gemini in Google Messages, Android Auto, Wear OS, etc.
Despite receiving some flakes from the industry, Google is riding the AI wave and decided to integrate Gemini into a new set of features for phones, cars, and wearables. With these new features, users can use Gemini to craft messages and AI-generated captions for images, summarize texts through AI for Android Auto, and access passes on Wear OS. (Link)
👨💻 Microsoft Copilot GPTs help you plan your vacation and find recipes.
Microsoft has released a few copilot GPTs that can help you plan your next vacation, find recipes, learn how to cook them, create a custom workout plan, or design a logo for your brand. Microsoft corporate vice president Jordi Ribas informed the media that users will soon be able to create customized Copilot GPTs, which is missing in the current version of Copilot. (Link)
New to the newsletter?
The AI Edge keeps engineering leaders & AI enthusiasts like you on the cutting edge of AI. From ML to ChatGPT to generative AI and LLMs, We break down the latest AI developments and how you can apply them in your work.
Thanks for reading, and see you tomorrow. 😊