• Everyday AI
  • Posts
  • Google groundbreaking AI updates, OpenAI continues to surprise & more – AI News That Matters

Google groundbreaking AI updates, OpenAI continues to surprise & more – AI News That Matters

ChatGPT Search is publicly available, DeepMind unveils Veo 2, Meta joins legal battle against OpenAI's for-profit shift and more!

Outsmart The Future

Sup y’all 👋

The past week in AI has been ….. it’s been a week. We had to give you bite-sized updates below for the AI News that Matters so we could fit it all in one newsletter without killing digital trees. 

Speaking of AI updates that matter… what do you wanna see for our livestream/podcast tomorrow? 

That’s a TON to cover.

What should we cover on tomorrow's show?

Login or Subscribe to participate in polls.

You’re in the driver’s seat. 

✌️
Jordan 

P.S. If you’re from Chicago, I’ll be a judge tomorrow at the GenAI Collective event at mHUB. Details here if you wanna join. 

Today in Everyday AI
6 minute read

🎙 Daily Podcast Episode: From Google quietly releasing real-time AI to OpenAI’s huge announcements, here’s the AI news that matters! Give it a listen.

🕵️‍♂️ Fresh Finds: Instagram’s head speaks on AI in social media, Trump’s Silicon Valley advisers eye AI censorship and why Marc Andreessen is scared about AI. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: ChatGPT Search is publicly available, DeepMind unveils Veo 2 and Meta joins legal battle against OpenAI's for-profit shift. For that and more, read on for Byte Sized News.

🚀 AI In 5: Google has released Deep Research and it’s pretty dang good! Is it better than Perplexity? Find out here

🧠 AI News That Matters: Why is Apple Intelligence being delayed? Will GPTSearch replace Google? And what do Meta’s LLM updates mean for the AI world? Here’s our breakdown of the AI news that matters. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about Google unveiling Agentspace, Microsoft launching Phi-4, Claude 3.5 Haiku is live, Google’s new NotebookLM Pro Check it here!

AI News That Matters - December 16th, 2024 📰

Did you see Google quietly released real-time AI.

Apple might be coming out with the cheapest AI phone to date.

And OpenAI and Google were like two heavyweights going 12 rounds, blow for blow with new AI releases.

Yikes.

This week was unofficially dubbed ‘AI Week’ by dorks like me online.

Didn’t have like 27 free hours to keep up? Don’t worry. We got you.

We bring you the AI News that Matters.

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

Also on the pod today:

• Microsoft’s AI model 🧮
• Klarna's Use of AI 🏢
• Eric Schmidt on AI 👤

It’ll be worth your 49 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

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

New AI Tool Spotlight – Vela is an AI agent that helps you invest in startups, ACE Studio is an AI vocal workstation for professionals and Voiser AI transcribes, summarizes and translates videos and recordings.

Trending in AI – Instagram’s head says social media needs more context because of AI.

AI in Politics – Trump’s Silicon Valley advisers are eyeing AI ‘censorship.’

Here’s why Marc Andreessen was ‘very scared’ after meeting with the Biden administration about AI.

AI in Governance – Lockheed has formed a subsidiary to help defense companies adopt AI.

AI in Media - Lisa Kudrow is criticizing Robert Zemeckis Film ‘Here’ as an “endorsement For AI.”

1. ChatGPT's Search Engine Rolls Out to All Users 🔍

OpenAI has announced that its AI-powered search engine is now available to all users, following its initial launch for paid subscribers in October. This upgrade brings a more traditional search experience on mobile, complete with location-based results that include images, ratings, and hours, making it easier for users to find local attractions and restaurants.

Additionally, the new feature allows for quicker access to links for specific sites like hotel bookings, enhancing user efficiency. For those looking to stay ahead, this development signifies a significant leap in how AI can assist in everyday decision-making and planning.

2. DeepMind Unveils Veo 2: A Game Changer in Video Generation 🎥

In a move to challenge OpenAI's Sora, Google DeepMind has launched Veo 2, a next-generation video-generating AI capable of producing over two-minute clips at resolutions up to 4K. While currently restricted to 720p and eight seconds in Google's VideoFX tool, Veo 2 promises enhanced realism with improved physics understanding and clearer visuals, making it a noteworthy contender in the AI space.

This advancement not only signals a leap forward for content creators but also raises questions about the ethical implications of AI in creative industries, especially regarding training data sourced from platforms like YouTube.

3. Meta Joins Musk in Legal Battle Against OpenAI's For-Profit Shift 🧑‍⚖️

Meta has thrown its support behind Elon Musk's bid to halt OpenAI's transition from a nonprofit to a for-profit entity. Citing potential "seismic implications for Silicon Valley," Meta urged California's attorney general to take action while suggesting Musk and former board member Shivon Zilis are fit to represent public interests.

This move underscores the intense competition within the AI sector, as Meta positions itself against a company that's rapidly evolving.

4. Google's Whisk: A New Way to Remix Images 🖼

Google Labs is currently testing Whisk, an innovative image generator that allows users to create unique visuals by combining three different images: one for the subject, another for the scene, and a third for the style. Powered by Imagen 3, Whisk enables users to remix their photos creatively, although results may vary in terms of physical traits.

Currently available only in the U.S., this tool could significantly impact how individuals and businesses present their brand through personalized imagery.

5. Ray-Ban Meta Glasses Get AI Upgrade 👓

Meta is making waves with its latest firmware update for Ray-Ban Meta smart glasses, introducing a suite of AI-powered features that include live conversations and real-time translation capabilities. With the new “live AI” functionality, users can engage in ongoing discussions with Meta AI without needing to use a wakeword, while also getting instant insights about their surroundings through the glasses' front-facing camera.

The update also enables translation of spoken English into Spanish, French, or Italian, making cross-language communication smoother than ever.

6. YouTube Gives Creators Control Over AI Training 👀

YouTube has rolled out a new feature allowing creators to permit specific third-party AI companies to train their models using their content. Creators can now opt in through YouTube Studio, selecting from a list of 18 companies, including heavyweights like OpenAI and Adobe. This change comes in response to ongoing concerns that AI companies have been using creators' content without consent or compensation.

While this setting enhances creator control, YouTube will still maintain its ability to train its own AI models on some content, keeping the balance in the evolving landscape of AI and content creation

Google Deep Research: Will this change the way we use the web?

Google has just released Deep Research and it’s REALLY good!

Could it beat out Perplexity? We think so.

We explain why as we break down this new update.

Unofficially dubbed by dorks like me as “AI Week,” Google and OpenAI went toe-to-toe in creating release-packed week of AI goodies. 

OpenAI's doing their "12 Days of" thing, dropping presents like a caffeinated Santa. Then Google woke up and chose violence, unleashing more updates in three days than we've seen in three years. 

Zero marketing strategy, seemingly.

And that's not even the wildest part. Servers melted, Siri got humbled, and somebody's building an AI to watch other AIs. For real.

There’s legit too much to cover this week, so our written recaps are gonna be a little shorter. 

(You hate the long stuff anywho, right?) 

Let's get into it, shorties.

1 – OpenAI’s Sora Hits Different 🎥

OpenAI's new video generation tool Sora just dropped, and their servers literally couldn't handle the heat. Available on both $20 and $200 plans, but only the ballers on Pro can make human videos. 

While Runway, Kling, and Luma Labs got more experience, Sora's ceiling is hitting different. Run ten generations and you might get some duds, but the best ones? Straight fire.

What it means:

This is the worst Sora will ever be. Think about that. Even with melted servers and first-day jitters, it's already making other video tools look like flip phones at an iPhone party.

At least when the generations come out on point. 

2 – Canvas Gets Democratized 🎨

ChatGPT's Canvas went free-for-all, with bonus Python powers built right in. Drop a spreadsheet, ask for pretty graphs, boom - done. No coding degree required.

What it means:

OpenAI's taking shots at Anthropic's artifacts feature. Making it free? That's not just generous - that's strategic warfare. 

And yeah, they definitely took a page outta Anthropic’s book for this one. More on that in a sec. 

3 – Advanced Voice Mode Grows Eyes 👀

Remember that Advanced Voice Mode from September? Now it's got video superpowers. Point your camera at math problems, whiteboards, whatever - your AI assistant's watching and learning.

What it means:

The tech's solid but we're all still figuring out the killer use case. Right now it's like having a really smart friend who's not sure how to help.

We gave it a spin, and weren’t super impressed. 

However, we’ve gotten better results in the past few days. Guessing the whole melting of servers just gave AVM some short-term problems. 

4 – Projects Finally Fixes Chat Chaos 📁

After what felt like forever, ChatGPT got folders. But nah - these aren't your boring old folders. Projects lets you share docs across chats, making custom GPTs look like they're still in training wheels.

What it means:

OpenAI fixed their biggest UI headache and snuck in a document handling system that's lowkey better (in some ways) than their custom GPTs.

5 – Apple Intelligence Gets Real 🏢

Apple and OpenAI got official with the joint release of Apple Intelligence, pretty much powered by ChatGPT. 

Here's the tea: Siri got exposed. When it can't handle your questions (which is most of the time), it straight up asks ChatGPT for help now with the latest official partnership between OpenAI and Apple. You'll need iOS 18.2 or Mac 15.2 for this humbling experience.

What it means:

Most of these features?

Nothing burger with extra mayo. The real story is Siri finally admitting it needs help.

Never thought a smart assistant would need an AI assistant. Lolz. 

6 – Budget iPhone Takes on AI 🏢

Apple's cooking up an iPhone SE 4 with all the AI goodies for half the pro price. We're talking $499-599 instead of dropping a rack and a half on the 15 Pro.

What it means:

The AI revolution won't be Pro Max exclusive. Apple's finally reading the room and realizing that the majority of the world might not wanna drop a used 2012 Honda Civic on a new phone every other year. 

7 – Schmidt Drops AI Reality Check 💣

Former Google CEO Eric Schmidt said we might need an AI kill switch in 2-4 years because it might be self-improving. You know… like AI building the next version of AI. 

His solution? Build another AI to babysit the first one. Peak tech logic.

What it means:

When the guy who ran Google says AI self-improvement is 2 years away instead of 50, it's time to pay attention.

8 – Klarna Goes Full Robot 🤖

Klarna shrunk from 4,500 to 3,500 humans this year. Their vibe? When people leave, AI fills the seat. Still got 50 engineering roles open though - turns out robots need handlers?

What it means:

The AI takeover isn't coming with warning sirens. It's the quiet replacement of "Currently Hiring" with "Currently Training."

9 – Devin Wants $500 to Code 💻

Cognition dropped Devin at $500 per month. Think junior dev who never sleeps, never complains, just builds. No constant prompting needed - give instructions, go grab coffee, come back to finished code.

Hopefully. 

What it means:

The dev world's about to get real interesting. This isn't just another coding assistant - it's the first real "fire and forget" development AI.

10 – Microsoft’s Tiny Beast 🎯

Phi-4 just proved size ain't everything. With only 14 billion parameters, Microsoft’s new small model is kinda embarrassing GPT-4's 1.8 trillion on math problems. David's out here dropping Goliath like pew pew pew. 

Who’s the shorty now!?

What it means:

The future of working with LLMs isn't one massive model. Our hot take? It's hundreds or thousands of specialized, domain-specific Small Language Models that do one thing perfectly.

11 – Gemini 2.0 Flash Flexes 🚀

Google saw all of OpenAI’s holiday announcements and said "hold my eggnog” then dropped Gemini 2.0 Flash. Their small model's outperforming the premium competition. The disrespect is real.

Could 2.0 Flash become a mainstream model while we wait for Google to bring us the Pro/Ultra variations of Gemini 2.0? Maybe. 

What it means:

Google finally found their rhythm and is shipping ships like they had governorship over a shipyard that ships ships. (That’s AI talk for they’re releasing a TON of new AI products and features.) 

And when your mini model's smoking the competition's best stuff, you're doing something right. Google seemingly took off like the first 11 months of 2024 but is coming in like a straight behemoth in the 11th hour. 

12 – AgentSpace Enters the Chat 🔍

Google Cloud launched AgentSpace, letting businesses create their own branded search agents. Works with Google Drive, Confluence, even Microsoft's SharePoint. 

The catch? Good luck getting access. 

What it means:

Google's enterprise game is leveling up, but their waitlist game is still stuck in 2020. Whomp whomp. 

13 – Project Astra Takes AR Seriously 👓

Google's pushing some new updates to the still-not-available Project Astra after debuting it months ago at their I/O conference. Think Gemini in your field of view, helping you navigate reality in real-time. In the short-term, we’ll see Project Astra available in Google-powered smartphones, but the end goal may be glasses/headsets. 

Think — this is Google’s ultimate competitor to OpenAI’s Advanced Voice Mode with Video. Everyone’s racing to put a friendly, know-it-all AI assistant in your pocket. 

What it means:

Cool tech, but are we really ready to wear computers on our faces all day? The Meta Ray-Bans are fire, but these chunky AR glasses? We'll see.

We would love to see Astra come to the Google iOS app, but we doubt that would ever happen. Guessing it needs some edge AI to work, hence the whole Google/Samsung phone thang.

14 – Android XR Changes the Game 🎮

Google, Samsung, and Qualcomm just announced Android XR - their answer to Apple's Vision Pro flop. Except this will be cheaper, lighter, and maybe actually useful, baked in with Gemini’s AI from front to back. 

What it means:

Mixed reality might work if you don't have to mortgage your house to try it. (Ahem. Looking at you $4,000 Apple headset that only like 5 dudes in Silicon Valley actually wanted.)

Looks like the Meta Quest might have a competitor finally. 

15 – NotebookLM Gets Premium 📝

NotebookLM Plus is here and Google is slooooooooowly rolling out with paid team features and 5x the limits. But the real MVP? The call-in feature letting you interrupt AI podcasts with questions.

Yeah. We want that.

What it means:

Google's RAG model just became a must-have for teams doing serious research.

Do you have access yet? We’re still waiting. 

Haven’t used NotebookLM yet? Go check out our latest review. 

16 – Project Mariner Sets Sail 🌊

Google's web automation AI is finally here (for "trusted users”…. Which we think is like 13 people cuz like no one has access.).

If Mariner sounds familiar, we’ve covered the Chrome agentic rumors when the project was previously codenamed ‘Jarvis.’ And yeah…. We called the name change on that one. 

One Chrome extension to rule them all - if you can get access.

What it means:

When this drops for real, your browser's basically got an intern that never sleeps if you have an extra window or computer for it to work. (It only works on an active Chrome tab, so you may not be able to multitask when Mariner is out Marinering.) 

17 – Deep Search Goes Hard 📚

Perplexity may be on watch. For real. 

Google's new Deep Research tool can literally visit/summarize hundreds of websites in a few minutes and save you like 1,432 pop up ads.

Perplexity's been pushing shopping ads while this Google’s new ‘Deep Research’ is going dang hard in the paint.

This is probably gonna end up being one of our top 10 AI tools/features of 2024. 

What it means:

Research game changed forever. While others are pushing products, Google built a research machine that actually listens.

18 – Google AI Studio Gets Real 🎬

Real-time voice and vision features finally dropped in AI Studio.

No waitlist, no drama.

Luckily on this one, Google just shipped it. Again, don’t go looking in the front-end of Gemini chat for this one. It’s an AI Studio exclusive. 

With this, you can talk to Gemini live, share your screen and chat, and also show it your video feed. Pretty cool. Just remember, AI Studio is free. And so is whatever data that you upload in there for Google to use. Lolz. 

What it means:

Google's proving they can sometimes ship features without a waitlist. Sometimes.

Numbers to watch

25%

Nearly a quarter (around 25%) of the US workforce was using generative AI weekly as of August 2024.

Now This …

What are your thoughts?

Vote to see live results

How much trust do you have in our reporting?

Login or Subscribe to participate in polls.

Reply

or to participate.