- Everyday AI
- Posts
- Ep 669: OpenAI's Code Red: Is Google taking ChatGPT's Crown?
Ep 669: OpenAI's Code Red: Is Google taking ChatGPT's Crown?
Claude hints at front end Agents, Trump Approves Nvidia GPU sales to China, Meta’s new model in 2026 and more.
Everyday AI
December 09, 2025
Sup y’all 👋
No GPT-5.2 release as of hitting send, but TONS of AI updates over the past few days/weeks to dive into.
For our ‘AI at Work on Wednesday’ series tomorrow, what should we dive into?
AI at Work on Wednesday: What do you wanna learn more about?🗳️ Vote to see LIVE results 🗳️ |
See ya tomorrow,
✌️
Jordan
(Let’s connect on LinkedIn. Just tell me you’re from the newsletter!)
Outsmart The Future
Today in Everyday AI
8 minute read
🎙 Daily Podcast Episode: Reports suggest OpenAI has entered a code red as Google’s Gemini models surge ahead. What the heck is going on? Find out more in today’s show and give it a watch/listen.
🕵️♂️ Fresh Finds: Claude partners with Accenture, Gemini catching up with ChatGPT, Microsoft releases 7 AI trends that can reshape your 2026 and more Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: Claude hints at front end Agents, Trump Approves Nvidia GPU sales to China, Meta’s new model in 2026 and more. Read on for Byte Sized News.
🧠 Learn & Leveraging AI: We explain OpenAI’s reported ‘Code Red’ stance and tell you straight if Google is taking over the crown. And more importantly, what it means for you. Keep reading for that!
↩️ Don’t miss out: Trump Signing 'One Rule' Executive Order, Google soon to drop Nano Banana 2 Flash, Investors saying to buy in on Alphabet and more Check it here!
Ep 668: OpenAI's Code Red: Is Google taking ChatGPT's Crown?
OpenAI is (reportedly) in full panic mode. 🚨
All hands on deck, Code Red status.
So.... what happened?
How did OpenAI go from defining the AI category to getting beat by competitors they once trounced?
And, is it too late for them to turn it around? Or will Google permanently take the AI crown?
Tune in... we've got hot takes.
Also on the pod today:
• OpenAI’s “code red” panic 🚨
• Gemini 3 vs. GPT-5.1 showdown 🤖
• Market share: Google surging 📊
It’ll be worth your 46 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 – GLM-4.6V is an Open Source Multimodal Models with Native Tool Use, Kortex transforms Google NotebookLM into a dynamic, interconnected knowledge base. Kerno Helps you Automate your backend integration tests with AI
Claude Code x Slack — Now you can send bugs from Slack straight into Claude-powered coding sessions.
Anthropic Agent Mode — Claude hints at full agent mode as Anthropic tests task-first workflows
AI Coding — MediaTek NPUs quietly turn Chromebooks and phones into full on-device AI engines
Mixboard New Update — Mixboard just got a serious glow-up, with Nano Banana Pro now spinning your boards into full presentations. Want to see how wild your slides can get?
AI Shopping — ChatGPT turns into a grocery shopper—can agentic commerce finally fund OpenAI?
Microsoft Trends — AI is shifting from experiment to everyday partner, from “coworkers” to quantum. Peek at seven trends that could reshape how we work and innovate in 2026.
AI and Consulting — Right after signing a deal with ChatGPT, Accenture is rolling out a similar partnership to train 30,000 consultants on Claude.
1. Microsoft bets big on India’s AI future 🤑
Microsoft has just unveiled a massive US$17.5 billion investment in India over 2026–2029, signaling that the company sees India as its next big AI stronghold in Asia. The plan centers on building the company’s largest hyperscale cloud region in the country by mid-2026, expanding existing data centers, and rolling out new “sovereign” cloud options designed to keep sensitive data inside India’s borders.
Alongside the infrastructure build-out, Microsoft is doubling its AI skilling target to 20 million Indians by 2030 and wiring advanced AI into key government employment platforms that serve more than 300 million informal workers, aiming to plug AI directly into the country’s social and jobs ecosystem.
2. Trump clears limited Nvidia AI chip sales to China 🇨🇳
In a new move with big implications for global tech rivalry, President Donald Trump has approved Nvidia to sell its H200 AI chips to “approved customers” in China, while keeping its most advanced Blackwell and Rubin chips off the table.
The decision tries to walk a tightrope between protecting U.S. national security and keeping American chipmakers plugged into a crucial market that fuels their growth and manufacturing at home. Nvidia praised the move as a win for domestic production, while regulators at the Commerce Department will screen which Chinese buyers get access.
3. EU targets Google over AI-powered search and YouTube training 🎯
The EU has just launched a new probe into Google's AI search summaries and YouTube data practices, zeroing in on whether the company is using publishers' and creators' content to train its AI without fair pay or a real opt-out.
Regulators are especially worried that Google's AI Overviews and AI Mode could drain traffic and ad money from news sites, at a time when some publishers say their clicks from Google have collapsed since the feature arrived. The investigation also asks whether YouTube videos are silently feeding Google's broader AI models in ways that might hurt the very creators who supply the platform with content.
4. Meta’s AI Shakeup: New ‘Avacado’ Model, New Playbook 🥑
Meta is ramping up its AI rivalry with OpenAI and Google, shifting gears from its open-source Llama models to a secretive new LLM project reportedly codenamed Avocado, now set for release in early 2026.
After a series of high-profile hires and a $14 billion investment spree, CEO Mark Zuckerberg is betting big on new talent and a more closed approach to AI, hoping to regain momentum as competitors surge ahead. Internal shakeups and marathon workweeks reveal just how high the stakes are, with Meta slashing metaverse spending and refocusing on AI-powered products. As the dust settles, Wall Street and the tech world are watching closely to see if Meta’s new strategy can deliver a breakthrough in the race for AI supremacy.
5. Anthropic tests new “Agent Mode” for Claude 💻
Anthropic is quietly testing a new Agent Mode for Claude that shifts the assistant from casual chat into a more structured, task-focused workspace, signaling a significant product evolution that appears to be actively in development. Code references and new UI assets show a home-screen toggle that flips between classic chat and a more complex, grid-backed interface with a “let’s go” button, pointing to multi-step task flows and heavier use of tools and possibly browsing.
This positions Claude to compete more directly with AI platforms that already focus on automated task execution, while still keeping the advanced mode optional for users who want it.
Three years ago, Google hit the panic button when ChatGPT launched.
They were behind.
Last week, OpenAI hit the same button over Google.
Should heavy ChatGPT users and enterprises be worried?
According to reports, CEO Sam Altman's internal ‘Code Red’ memo told employees to freeze everything. The directive? Stop chasing features and get back to making OpenAI models actually good again.
The trigger was brutal.
Gemini 3 launched and immediately topped every major benchmark. Salesforce CEO Marc Benioff tweeted he was switching after three years of daily ChatGPT use. His post hit 3.2 million views. Wells Fargo now rates Google as the AI race leader for the first time since 2016.
But here's what nobody's talking about.
OpenAI's real problem isn't Google. It's that 76% of their paying customers are using a model that ranks 15th on third party benchmarks while Google serves number one.
The once-uncontested king of the hill in OpenAI now faces an immediate uphill battle.
So on today's show, we broke down the self-inflicted wound behind OpenAI's panic, why spreading across five businesses is bleeding them dry, and the hidden data that explains why they'll probably win anyway.
And most importantly, what it means for you.
1. The default model problem nobody mentions 🎯
Everyone keeps saying ChatGPT "feels worse than before."
They're right. They're just blaming the wrong thing.
Here’s one of the biggest reasons that ChatGPT has started to slip that no one is talking about: the default chat model is no longer top-tier.
OpenAI's thinking models are legitimately competitive. But ChatGPT’s default ‘Auto’ mode often routes chats to a model that doesn’t think much. As we dished on today’s show, OpenAI’s baseline GPT-5.1 mode ranks 15th on LLM Arena.
The previous GPT-5 chat version? Twenty-first. Yikes.
Since shifting to this strategy in August, the majority of users (even paid ones) are likely using a CRAZILY inferior model and don’t even know it.
Pre-thinking models, OpenAI’s default chat model GPT-4o was top shelf. Not anymore.
Google Gemini serves you their best model by default. OpenAI makes you dig for theirs.
That's not a Google problem. That's a product decision that's costing OpenAI the narrative. (But it at least saves them compute?)
Try This
Open ChatGPT right now and check what model it's actually serving you.
If you're getting instant responses, you're on the 15th-place model.
Switch to thinking mode for anything that touches real business decisions.
You've been driving a rental while paying for a Ferrari.
2. Five businesses, zero focus 📉
OpenAI isn't just an AI company anymore. They're a friggin conglomerate.
Frontier models. Consumer ecosystem with 800 million users. Enterprise platforms. Custom chip development. Trillion-dollar infrastructure buildout. Five unrelated businesses hemorrhaging cash at once.
Projected 2025 spending: $22 billion. Revenue: $13 billion.
Another reason that likely lead to the ‘Code Red’ revelation? OpenAI is no longer just a chatbot company. They’re a company competing with everyone from Google and Anthropic to NVIDIA and Microsoft.
Could it pay off? For sure.
Is it a reason they’ve lost their footing atop the AI hill? Absofrigginlutely.
Anthropic picked one lane and expects to break even by 2028. OpenAI reportedly won't see profit until 2030.
The company chasing everything could get temporarily lapped by their AI rival doing one thing well.
Try This
Count the AI tools your team touches weekly.
ChatGPT for general stuff, Copilot for code, Perplexity for research, plus whatever corporate mandates.
Pick the two delivering 80% of your value and go deep.
Shallow across five beats deep in two exactly zero percent of the time.
3. The data that says ChatGPT wins anyway 📈
We’ll cut it to you straight shorties.
Much of this is just AI nerd talk, fueled by Silicon Valley gossip circles and overblown tech journo coverage.
Ignore the panic for a second. Look at the stickiness numbers.
According to Similarweb, ChatGPT has a 33% stickiness rate. Gemini sits at about 5%. No other company comes close to ChatGPT, which is why they’ve been winning the user race. 800 million weekly active users can’t be wrong: ChatGPT has the features, modes and connections that make users return all week.
One more future reason OpenAI aint giving up its top spot?
Ads.
Yeah, yeah, yeah. Millions are gonna belly ache. But only 5% of ChatGPT’s users are reportedly paid users. Which means that 760 million weekly active users on free plans are likely to start seeing ads at one point soon. (After the Code Red dust settles)
And when that happens? We see OpenAI making more bank than…. Banks.
Think of it like this: The average Google search: 3.4 words. Average ChatGPT query: 60 words.
That's not search. That's a relationship. Users share context, preferences, and work patterns over months.
Oh, and throw in the fact that Enterprise message value jumped 8x this year. And 36% of US businesses run ChatGPT Enterprise versus 14% for Anthropic.
Yeah, OpenAI’s not leaving their market share throne anytime soon.
ChatGPT built something stickier than a better model. That's why ads are coming despite the backlash and how they’ll be able to scale beyond many of their competitors.
Try This
Pull your ChatGPT history and notice what you've already shared.
Your job, your preferences, your communication style are all baked in there.
Set up custom instructions and projects so that context compounds into smarter outputs.
That's your moat if you stay. And your switching cost if you leave.
Reply