• Everyday AI
  • Posts
  • OpenAI launching agentic AI, Google reclaims top AI model ranking – AI News That Matters

OpenAI launching agentic AI, Google reclaims top AI model ranking – AI News That Matters

NVIDIA’s Blackwell is overheating, Perplexity launches ecommerce feature, Mistral unveils new AI models, ElevenLabs launches conversational AI bots and more!

Author

Everyday AI
November 18, 2024

Outsmart The Future

Sup y’all! 👋

Tomorrow’s my birthday. How can you celebrate? 

Make sure to join us live on the show tomorrow. 

It’s gonna be a BIG one! 

✌️

Today in Everyday AI
8 minute read

🎙 Daily Podcast Episode: OpenAI unveils its agentic AI. Google now has the top AI model. Microsoft will soon make its biggest AI announcement of the year. Here’s this week’s AI news that matters. Give it a listen.

🕵️‍♂️ Fresh Finds: Meta AI gets added to Ray-Bans in EU, NVIDIA and Google partner and Accenture and Microsoft create joint business. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: NVIDIA’s Blackwell is overheating, Perplexity launches ecommerce feature, Mistral unveils new AI models and ElevenLabs launches conversational AI bots. For that and more, read on for Byte Sized News.

🚀 AI In 5: We found some new features inside the Google Gemini Live app that no one is talking about. See it here

🧠 AI News That Matters: Will Microsoft’s next announcement change the AI game? Is OpenAI’s new agentic AI a game changer? And is NVIDIA still the company to beat? Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about Elon Musk looking to raise $6 billion for AI chips, Coca-Cola has an AI commercial mishap and a new writing feature from Claude? Check it here!

AI News That Matters - November 18th, 2024 📰

✅ OpenAI is releasing Agentic AI.
✅ Google now has the world’s top AI model.
✅ And Microsoft is hours away from its biggest AI announcements of the year.
 
Don’t waste time trying to keep up with AI on your own.

We bring you the AI news that matters.

Join the conversation and ask Jordan any questions on AI here.

Also on the pod today:

• NVIDIA's Role in AI 🟢
• AI in Medical Diagnostics 🧑‍⚕️
• Microsoft Ignite Event 🗓

It’ll be worth your 35 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

 Upcoming Everyday AI Livestreams

Tuesday, November 19th at 7:30 am CST ⬇️

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – Read AI For Gmail is a Chrome extension that summarizes your Gmail inbox, Gecko Security is an AI security engineer and Turbo Lens gives you OCR & insights from any image.

Microsoft – Accenture and Microsoft are launching a business AI unit.

Big Tech – NVIDIA is partnering with Google to accelerate Google Quantum AI.

Meta – Meta AI is rolling out to Ray-Ban Meta Glasses in France, Italy, Ireland and Spain.

Google – Google.org has invested $20 million to researchers using AI for scientific breakthroughs.

AI in Medical – Researchers from the National Institutes of Health (NIH) have developed an AI algorithm to help match potential volunteers to relevant clinical research trials.

AI in Media – David Attenborough reacts to an AI replica of his voice.

AI Startups - Former Google X employees have come out of stealth with TwinMind, an AI app that hears and remembers everything about you.

1. NVIDIA’s Blackwell Faces Overheating Hurdle 🫠️

NVIDIA's stock took a hit on Monday as concerns emerged over overheating issues with its next-gen Blackwell processors, particularly in high-capacity server racks. A report from The Information revealed that the AI chipmaker has had to repeatedly redesign server racks, raising alarms about potential delays in deploying its cutting-edge technology.

Despite these setbacks, an NVIDIA spokesperson reassured that such design adjustments are standard practice, done in collaboration with cloud service providers. As the company prepares to release its fiscal third-quarter results this Wednesday, investors are keenly watching for updates on these pressing challenges.

2. IBM and AMD Team Up for AI Power Boost 🚀️

IBM and AMD are set to launch AMD Instinct MI300X accelerators as a service on IBM Cloud in early 2025. This partnership aims to enhance performance and efficiency for generative AI models and high-performance computing applications, catering to enterprises' growing demand for robust AI solutions.

With 192GB of high-bandwidth memory, these accelerators promise to support larger model inferencing while optimizing costs.

3. Perplexity Launches ‘Buy with Pro’ Feature 🛒

Perplexity has unveiled its new AI-driven ecommerce feature, ‘Buy with Pro,’ allowing U.S. users to shop directly through its app with the lure of free shipping. This feature, previously known as ‘Pro Shop,’ integrates Shopify’s API to streamline purchases and enhance user experience, while Perplexity aims to grow its search query volume rather than monetize the shopping function immediately.

As it takes on giants like Amazon and Google Shopping, Perplexity is also launching a visual search tool, ‘Snap to Shop,’ allowing users to upload images for product recommendations.

4. Mistral Unleashes Powerful AI Upgrades 👀️

Mistral has rolled out significant updates to its Le Chat platform, now featuring web search capabilities with citations and a new "canvas" tool for content modification. With the addition of Pixtral Large, a 124 billion parameter model that excels in multimodal tasks, users can now analyze large PDFs and images, creating a more versatile experience for tasks like document analysis and automated workflows.

These advancements not only enhance user productivity but also position Mistral as a competitive player in the rapidly evolving AI landscape, which is crucial for professionals looking to leverage AI for career growth.

5. ElevenLabs Launches Conversational AI Bots 🗣

ElevenLabs has rolled out a platform for building customizable conversational agents, allowing users to tailor voice, tone, and length of responses. This new feature caters to the growing demand for interactive AI tools, with the company’s head of growth, Sam Sklar, noting that many clients were already leveraging similar functionalities.

Users can now integrate their own knowledge bases and select from various large language models, positioning ElevenLabs as a strong contender against giants like OpenAI and Google.

6. Sam Altman Joins SF Mayor-Elect Lurie’s Transition Team 🏛

San Francisco's mayor-elect Daniel Lurie has enlisted OpenAI CEO Sam Altman to lead his transition team, signaling a new era of collaboration between tech leaders and city officials. With public safety and innovation at the forefront, Lurie aims to reverse the exodus of tech talent from the Bay Area, a challenge exacerbated by rising crime and dissatisfaction with city governance.

Altman, recognized as a pivotal figure in the tech landscape, brings invaluable connections and insights that could help revitalize San Francisco's startup ecosystem.

Google Gemini Live: 3 Game-Changing Features Revealed!

There’s some new features inside the Google Gemini Live app on iPhone that no one is really talking about.

We reveal what they are and how to use them.

Find out in today's AI in 5.

There’s a new king of the AI hill. 

Yet, Microsoft is about to disrupt some things with their big AI announcements at Ignite. 

And maybe the biggest splash yet in AI this week? 

OpenAI’s agentic AI is reportedly launching in January. 

Sheeeeeeesh. 

The AI world never sleeps, even when we do. 

That’s why (almost) every Monday, we bring you the AI News that Matters. 

Enough chit chat shorties. 

Here’s what’s making waves and what you need to know. 

1 – Google Gemini Claims World's Top AI Model 👑

Google DeepMind's Gemini 1114 model just dethroned GPT-4o in the LM Arena on November 14th.

The model excels particularly in math and vision tasks.

Think Pepsi vs. Coke taste test, but for AI.

Users do blind comparisons in the LM Arena (formerly LMS Arena) without knowing which model they're testing.

The current leaderboard is dominated by OpenAI and Google models, with xAI's Grok 2 being the first outsider to break in.

OpenAI's already testing an anonymous new model in the arena. Google's got two more in the pipeline: "Mystery Gemini 3" and their spy-movie-worthy "Secret Chatbot."

Here's the catch – you can't access this powerhouse on Google's public site. Even Gemini's front-end chatbot runs on models 3-8 months behind their latest tech.

What it means:

Google finally showed up to throw hands, but OpenAI won't stay down. 

History shows that OpenAI will likely update their GPT4-o model within a week or 3 to reclaim the crown. But right now, Google Gemini’s 1114 is the world’s top model.  

The real problem? 

Google keeps their best models locked in AI Studio for dorky devs while competitors give front-end users access to their top tech. 

That strategy's gonna bite them, whether they realize it or not. 

2 – Medical Study: ChatGPT > Human Doctors 🏥️

A new JAMA study dropped some hard-to-swallow truth pill: ChatGPT-4 scored 90% accuracy in medical diagnoses. (Better than just the human doctors.) 

Human docs hit 74% without AI, barely improving to 76% with it.

Fifty doctors participated in the test using real patient cases from a 105-case set used since the '90s. This guaranteed ChatGPT couldn't train on the test data.

Dr. Jonathan Chen from Stanford found most doctors used AI like a basic search engine. Only a few realized they could input entire case histories for comprehensive analysis.

Dr. Adam Rodman, who helped design the study, noted doctors stayed confident in their diagnoses even when ChatGPT suggested better alternatives.

What it means:

The US healthcare system's about to get flipped. 

At least, we can hope as much. 

It seems like Big Healthcare’s egos are the biggest barrier to AI adoption in healthcare here in the U.S.. 

The tech's ready – but it doesn’t seem like the humans are yet.

3 – NVIDIA’s Q3 Earnings Drop Wednesday 💰️

The world's most valuable public company is about to show us the money. Wall Street's watching NVIDIA’s upcoming earnings report like a hawk after a 190% stock surge this year.

Investors want to see if NVIDIA can beat that $37 billion Q4 revenue projection. 

NVIDIA CEO Jensen Huang might spill details on next-gen Blackwell AI chips. Demand's already outpacing supply, which is a positive sign. 

What it means:

These aren't just numbers. These chips power literally every AI tool you use daily.

No chips = no AI revolution. That's why this Wednesday’s earnings report from NVIDIA matters more than your crypto portfolio.

It’s a sign of things to come for the entire tech industry. 

4 – Microsoft Ignite Kicks Off Tomorrow 🎯

Microsoft's biggest AI event of the year starts tomorrow (Tuesday) at 8 AM CST with Microsoft’s CEO Satya Nadella’s keynote.  

Make sure to join us tomorrow morning LIVE for all the deets, as we’ll be joining live from Ignite in Chicago. 

Their Autonomous AI Agents in Copilot Studio hits beta tomorrow, according to reports. This puts Microsoft square in the AI agent race with OpenAI, Google, and Anthropic.

What it means:

Will you or your team be at Ignite in Chicago? We will be. Hit us with a reply if you’d wanna chat AI with Jordan!

5 – ChatGPT's Desktop Apps Level Up 💻

OpenAI made it rain updates in its ChatGPT apps for both Mac and Windows users. 

Mac users can now send code directly from VS Code, Xcode, TextEdit, Terminal, and iTerm2 to ChatGPT. No more copy-paste juggling.

The feature doesn't let ChatGPT write back to these apps yet, but it can analyze and suggest improvements instantly.

Windows users finally get the ChatGPT app without a Plus subscription. That's huge for accessibility.

What it means:

These aren't just updates – they're the foundation for true AI-assisted development.

OpenAI's building bridges between your workflow and AI. The walls between coding and AI assistance are crumbling.

6 – OpenAI Proposes North American AI Alliance 🤝

According to reports, OpenAI's pushing for a NATO-style AI alliance with US allies, including Middle Eastern partners. 

The goal? Beat China in the AI race.

Their blueprint includes power infrastructure commitments, expedited AI economic zones, and naval nuclear reactor tech for data centers.

Sam Altman's personally pitching US officials on this plan. 

This isn't just about algorithms – it's about reindustrializing America with AI at its core.

What it means:

The AI race is now US vs China.

Full stop. Everyone else is playing for third place.

Power infrastructure is the new bottleneck. All the AI smarts in the world mean nothing if you can't keep the lights on.

7 – Gemini Gets Voice Upgrade on iPhone 🗣️

Google just launched Gemini as a dedicated iPhone app with their Gemini Live voice assistant. Android Authority found code suggesting bigger plans.

The upgrade will let Gemini Live read and analyze your files while chatting – a first for any AI voice assistant.

In our AI in 5 today, it looks like some of these features may already be functioning. (At least sometimes. Lolz) 

Current top voice AIs like ChatGPT's advanced voice mode can't touch your files or even handle typing mid-conversation.

What it means:

Voice AI is finally getting useful. Talk to your documents during your commute.

Google's about to leap past OpenAI in voice AI. The ability to access files while talking changes everything.

8 – OpenAI's "Operator" Drops in January 🤖

According to Bloomberg, OpenAI's agentic AI "Operator" launches January 2025 as a research preview and API tool. 

It's designed to handle complex tasks with minimal human input.

We're seeing a four-month AI agent rush: Anthropic's Computer Use is live, Google's Jarvis drops in December, Microsoft's CodePilot Studio launches tomorrow, and now Operator in January.

This isn't just another chatbot – it's AI that actually does stuff autonomously.

What it means:

This is the biggest shift in work since email. Your job description is about to change.

Love it or hate it, agentic AI work is gonna become as commonplace as a Slack or Teams message. Everyone’s gonna go agentic. 

The AI revolution isn't coming – it's here and it's actually doing your work while you read this.

⌚

Numbers to watch

$1 Trillion

Bloomberg says it will take at least $1 trillion to fund data centers, electricity supplies and communications networks required for AI.

Now This …

What are your thoughts?

Vote to see live results

How often do you read/watch/listen to Everyday AI?

Login or Subscribe to participate in polls.

Reply

or to participate.