BIG AI updates from Google...

Hey there, AI-preneurs!  

Google just threw a massive AI party at Google Cloud Next, and we've snagged you a VIP pass, featuring a neat summary of the multiple updates today.

Plus, we've got some of the latest developments from Meta and Microsoft as well. 😸

Let's dive in!

Meta Platforms and Intel unveils new AI chips

Meta is introducing the next generation of its custom-made chips called the Meta Training and Inference Accelerator (MTIA) v2, designed specifically for its AI workloads.

Here’s the scoop:

• It will power the key recommendation and ranking models that are so crucial to keeping users engaged on Facebook, Instagram, and Meta's other apps.

• By controlling the entire technology stack, Meta can achieve greater efficiency with its custom silicon compared to commercially available GPUs.

• MTIA is an important part of Meta's long-term roadmap to build the most powerful and efficient infrastructure for its AI needs, including support for generative AI workloads in the future.

On the other hand...

Intel has unveiled more details on its Gaudi 3 AI accelerator, which is shaping up to be a real challenger to Nvidia's dominant H100 GPU. 

• Intel claims the Gaudi 3 can train certain large language models around 50% faster than the H100, all while costing a fraction of the price.

• The new Gaudi 3 AI accelerator will become available to major server vendors like Dell, HPE, and Lenovo in the second quarter of 2024.

• Intel is clearly gunning to break Nvidia's stranglehold on the accelerator market, and the Gaudi 3 could be the ticket. 

By touting both superior performance and a lower price point, Intel is positioning the Gaudi 3 as a compelling option for companies and organizations looking to scale their AI workloads and infrastructure.

Google Cloud Next 2024: Everything announced so far

Don’t have time to fly to Vegas? Don’t worry, we bring you the best of AI updates happening for the latest Google event.

Google is leveling up its AI game significantly with these 5 new launches and updates. Let’s get into it.

1. Google Vids: A new AI-powered video creation tool integrated into Google Workspace, allowing users to create videos alongside other Workspace apps like Docs and Sheets.

2. Gemini Code Assist: An enterprise-focused AI code completion and assistance tool, competing with GitHub's Copilot Enterprise.

3. Google Workspace updates:

   • Voice prompts for the "Help me write" feature in Gmail

   • Ability to polish email drafts in Gmail

   • Customizable alerts in Google Sheets

   • Support for tabs in Google Docs

   • $10/user/month add-ons for AI-powered meeting/messaging and security features

4. Imagen 2: An enhanced version of Google's image generation tool, now with capabilities like inpainting, outpainting, and text-to-video generation.

5. Vertex AI Agent Builder: A tool to help companies build AI-powered conversational agents.

What’s more? 

Gemini in Databases to simplify database creation, monitoring, and migration for Google Cloud customers. They are also renewing focus on data sovereignty through partnerships rather than building out Google's own solutions.

Gemini 1.5 Pro, Google's most capable generative AI model, now available in public preview on Vertex AI.

P.S. Upcoming Nvidia Blackwell platform support on Google Cloud in early 2025, with liquid-cooled GB200 servers for large language model training.

Microsoft AI opens London hub to access ‘enormous pool’ of talent

Microsoft is opening a major new AI hub in London, doubling down on its AI efforts in the UK. 

The hub will be led by AI expert Jordan Hoffmann and is part of Microsoft's £2.5 billion pledge to build AI capacity in the UK, including bringing 20,000 advanced GPUs by 2026.

And will focus on advancing language models, infrastructure, and tooling for foundation models.

Microsoft cites the UK's strong AI talent pool and commitment to responsible AI development as reasons for the new investment. 

The hub will collaborate with Microsoft teams and partners like OpenAI as the company seeks to hire exceptional AI scientists and engineers.

Microsoft’s global expansion plans:

The tech giant has also announced a major $2.9 billion investment in Japan over the next two years.

Furthermore, Microsoft is opening its first Microsoft Research Asia lab in Tokyo, tapping into Japan's research talent to advance areas like embodied AI, robotics, and scientific discovery. 

💣 Latest AI hits: Quick recap

• Google releases Imagen 2, a video clip generator. (Link)

• Microsoft allegedly pitched DALL-E to the military to assist in envisioning the battlefield. (Link)

• A new bill would compel tech companies to disclose any copyrighted materials that are used to train their AI models.  (Link)

• Samsung’s Galaxy AI is now supporting a total of 16 languages across AI features like live translation and interpreter.  (Link)

• Texas is replacing thousands of human exam graders with AI. (Link)

• Apple AI paper details how its multi-modal ‘Ferret-UI’ LLM can interpret phone UIs. (Link)

• AI editing tools are coming to all Google Photos users. (Link)

• Air Force secretary plans to ride in AI-operated F-16 fighter aircraft this spring. (Link)

AI art to sip your coffee with…

Prompt: Planets, Water, Space, Galaxies.

Tools to explore:

• Aboard’s AI-powered bookmarking app: Visual elegance of Pinterest, the organizational prowess of Trello, and the analytical capabilities of ChatGPT - all wrapped up in an intuitive, open-web-powered app.

• GummySearch: Unlock the power of Reddit for deep market research and audience analysis with niche exploration, product validation, content inspiration, and sales lead generation.

• AI Phone: Enhance call efficiency with AI transcription, privacy, and smart replies in real time.

That’s a wrap! See you on Monday for a fresh dose of AI Odyssey. 😎