Google's AI Context Caching Cuts Costs By 75%
Plus: LMSYS's new Multimodal Arena compares compare vision language models, Apple's Vision Pro gets an AI upgrade.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 308th edition of The AI Edge newsletter. This edition features Google’s announcements about Vertex AI updates.
And a huge shoutout to our amazing readers. We appreciate you😊
In today’s edition:
🚀 Google announces advancements in Vertex AI models
🤖 LMSYS's new Multimodal Arena compares top AI models' visual processing abilities
👓 Apple's Vision Pro gets an AI upgrade
📚 Knowledge Nugget: Social mobility and GDP and some predictions on the possible effect of AI by
Let’s go!
Google announces advancements in Vertex AI models
Google has rolled out significant improvements to its Vertex AI platform, including the general availability of Gemini 1.5 Flash with a massive 1 million-token context window. Also, Gemini 1.5 Pro now offers an industry-leading 2 million-token context capability. Google is introducing context caching for these Gemini models, slashing input costs by 75%.
Moreover, Google launched Imagen 3 in preview and added third-party models like Anthropic's Claude 3.5 Sonnet on Vertex AI.
They've also made Grounding with Google Search generally available and announced a new service for grounding AI agents with specialized third-party data. Plus, they've expanded data residency guarantees to 23 countries, addressing growing data sovereignty concerns.
Why does it matter?
Google is positioning Vertex AI as the most "enterprise-ready" generative AI platform. With expanded context windows and improved grounding capabilities, this move also addresses concerns about the accuracy of Google's AI-based search features.
LMSYS's new Multimodal Arena compares top AI models' visual processing abilities
LMSYS Org added image recognition to Chatbot Arena to compare vision language models (VLMs), collecting over 17,000 user preferences in just two weeks. OpenAI's GPT-4o and Anthropic's Claude 3.5 Sonnet outperformed other models in image recognition. Also, the open-source LLaVA-v1.6-34B performed comparably to some proprietary models.
These AI models tackle diverse tasks, from deciphering memes to solving math problems with visual aids. However, the examples provided show that even top models can stumble when interpreting complex visual information or handling nuanced queries.
Why does it matter?
This leaderboard isn't just a tech popularity contest—it shows how advanced AI models can decode images. However, the varying performance also serves as a reality check, reminding us that while AI can recognize a cat in a photo, it might struggle to interpret your latest sales graph.
👓 Apple's Vision Pro gets an AI upgrade
Apple is reportedly working to bring its Apple Intelligence features to the Vision Pro headset, though not this year. Meanwhile, Apple is tweaking its in-store Vision Pro demos, allowing potential buyers to view personal media and try a more comfortable headband. Apple's main challenge is adapting its AI features to a mixed-reality environment.
The company is tweaking its retail strategy for Vision Pro demos, hoping to boost sales of the pricey headset. Apple is also exploring the possibility of monetizing AI features through subscription services like "Apple Intelligence+."
Why does it matter?
Apple's Vision Pro, with its 16GB RAM and M2 chip, can handle advanced AI tasks. However, cloud infrastructure limitations are causing a delay in launch. It's a classic case of "good things come to those who wait."
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: Social mobility and GDP and some predictions on the possible effect of AI
Is AI the great equalizer or the ultimate divider? Greg Baker dives into the relationship between wealth and social mobility, uncovering some surprising trends. Using data from Wikipedia and a bit of AI, Baker shows that wealthier nations generally offer more opportunities for people to climb the social ladder. But here's where it gets interesting: he predicts AI could shake things up.
Baker suggests two possible scenarios: AI could either become a great equalizer, giving anyone who can harness its power a shot at moving up in society, or it could widen the wealth gap if it becomes a prized resource that only the rich can access. Either way, he claims, we're in for some significant societal shifts.
Why does it matter?
As AI continues to reshape our economy, understanding its potential impact on social mobility is crucial. Whether AI becomes a ladder to success or a new barrier, policymakers and business leaders must prepare for the societal ripple effects.
What Else Is Happening❗
📞 Amazon's Q AI assistant for enterprises gets an update for call centers
The update provides real-time, step-by-step guides for customer issues. It aims to reduce the "toggle tax" - time wasted switching between applications. The system listens to calls in real-time and automatically provides relevant information. (Link)
💬 WhatsApp is developing a feature to choose Meta AI Llama models
Users will be able to choose between two options: faster responses with Llama 3-70B (default) or more complex queries with Llama 3-405B (advanced). Llama 3-405B will be limited to a certain number of prompts per week. This feature aims to give users more control over their AI interactions. (Link)
⚡ Bill Gates says AI's energy consumption isn't a major concern
He claims that while data centers may consume up to 6% of global electricity, AI will ultimately drive greater energy efficiency. Gates believes tech companies will invest in green energy to power their AI operations, potentially offsetting the increased demand. (Link)
🍪 Amazon is investigating Perplexity AI for possible scraping abuse
Perplexity appears to be scraping websites that have forbidden access through robots.txt. AWS prohibits customers from violating the robots.txt standard. Perplexity uses an unpublished IP address to access websites that block its official crawler. The company claims a third party performs web crawling for them. (Link)
🤖 Microsoft AI chief claims content on the open web is "freeware"
Mustafa Suleyman claimed that anything published online becomes "freeware" and fair game for AI training. This stance, however, contradicts basic copyright principles and ignores the legal complexities of fair use. He suggests that robots.txt might protect content from scraping. (Link)
New to the newsletter?
The AI Edge keeps engineering leaders & AI enthusiasts like you on the cutting edge of AI. From machine learning to ChatGPT to generative AI and large language models, we break down the latest AI developments and how you can apply them in your work.
Thanks for reading, and see you tomorrow. 😊