Trent Gillespie
December 04, 2025
OpenAI just hit the alarm button. Last week, Sam Altman sent an internal memo declaring "Code Red" — pausing advertising plans, shopping agents, and the Pulse assistant to focus entirely on ChatGPT's core experience. The catalyst? Google's Gemini 3 launch two weeks ago earned state-of-the-art benchmark scores and immediate deployment across 2 billion+ users.
The numbers tell the story: ChatGPT has seen a reported 6% drop in user traffic since Gemini 3's November 18th release. With ChatGPT's overall traffic share sliding from 87% to 74% over the past year — and Google's Gemini climbing from 6% to nearly 13% — Altman acknowledged his company is at "a critical time."
This isn't just competitive posturing. Three years after ChatGPT kicked off the AI revolution, the tables have turned. In 2022, Google declared its own "Code Red" over ChatGPT. Now OpenAI is scrambling while Google's stock hits all-time highs.
Let's be clear about OpenAI's recent track record. GPT-5 launched on August 7th with massive fanfare — Sam Altman promised "PhD-level intelligence" for everyone. What users got was different. The rollout was bumpy: users complained the model struggled with basic math and geography, it felt "flat" and "clinical," and many demanded access to the older GPT-4o they'd grown to prefer.
Altman went into damage-control mode, restoring legacy model access and promising fixes. OpenAI released GPT-5.1 in November with a "warmer, more conversational" personality. But the timing was brutal — it landed just days before Google dropped Gemini 3 and stole the spotlight entirely.
Service Spotlight
Rolling out AI without proper training is like handing someone car keys without teaching them to drive. We help organizations build AI capability at scale—from executive workshops that align leadership on strategy, to both online and in-person staff training that drives daily adoption, including our Jumpstart program to roll AI out in as little as thirty days.
The result? Your entire organization gains the foundation needed for safe, efficient AI use—not just a handful of early adopters.
Ready to move from experimentation to enterprise-wide adoption? Reply to start the conversation.
Google's Gemini 3 isn't just incrementally better. It topped the LMArena leaderboard with a record 1501 Elo score — the first model to break 1500. For now, it leads across many mathematics, creative writing, visual comprehension and other benchmarks. Their "Deep Think" reasoning mode scored 93.8% on GPQA Diamond (compared to GPT-5.1's 26.5%).
Perhaps more importantly, Google has distribution that OpenAI can only dream of. Gemini 3 launched simultaneously across Google Search (2 billion users), the Gemini App (650 million users), and Vertex AI for enterprise. When Salesforce CEO Marc Benioff publicly declared he was ditching ChatGPT for Gemini 3, calling it "insane" how much better it was, Wall Street took notice. Alphabet stock is up 66% this year.
And yes — "Nano Banana" is real. It's Google's image editor, and it's genuinely the best AI image generator on the market right now.
Yes, with nuance.
Here's our updated guidance on which AI application to standardize on for your company:
If you're a Google Workspace organization: Test Google Gemini seriously as your primary AI application. The integration with Gmail, Docs, Sheets, and Drive is now genuinely compelling, and the underlying model has caught up. The seamless experience of having AI built into tools your team already uses daily can be a real advantage.
If you're NOT a Google Workspace organization: ChatGPT remains a reasonable default, but hold it loosely. OpenAI is promising a new reasoning model next week that they claim beats Gemini 3 in internal evaluations. GPT-5.1 is better than GPT-5, and they promise image editing fixes soon. But the gap has closed dramatically, and their growth-at-all-costs approach hasn't provided the reliability businesses need for what is becoming a more and more critical workflow tool.
Anthropic's Claude is giving ChatGPT serious competition, particularly for technology organizations and enterprise customers. Their business customer base has grown from under 1,000 to over 300,000 in just two years. Claude Sonnet 4.5 is stable, thoughtful, and increasingly the choice for organizations that value reliability over hype. I now use Claude for about 50% of my work. Over the next few months, depending on how each platform performs, Claude may earn a higher recommendation from me than ChatGPT.
As for Microsoft Copilot? Although it's the simplest approach for Microsoft-based organizations, I'll be direct: using Copilot locks you into Microsoft for the AI era — the next 10+ years. If you are a Microsoft shop, and you don’t mind that, then Copilot may be a good fit for you. It’s simple and integrated.
However, almost uniformly, the business leaders I talk to consider Microsoft a costly burden to their companies, not an innovation enabler. With Copilot underperforming competitors functionally and showing lower user satisfaction scores, this is the perfect time to break out of Microsoft lock-in and try best-of-breed tools.
The most important strategic principle right now: build AI fluency, not platform dependency.
The landscape is shifting too fast for loyalty. Google went from embarrassing Bard failures to market-leading Gemini 3 in under two years. OpenAI went from untouchable to "Code Red" in months. Anthropic quietly built the enterprise darling. DeepSeek emerged from China matching GPT-5 benchmarks.
Any vendor you choose today may be third-best by mid-2026.
If you're leading AI adoption in your organization, here's what this competitive shake-up means for your strategy:
1. Train your team on AI fundamentals, not just one application.
Everyone needs to understand AI, and the user interfaces and functionality are becoming consistent. Choose the best-fit standard AI application for today, train your employees on it, and then allow your power users to develop fluency in ChatGPT AND Gemini AND Claude. Also, don’t hesitate to let your team use multiple AI tools when they follow your safety standards. From NotebookLM to Perplexity, there are many AI tools that are great for specific use cases and industries. Not letting employees use multiple AI tools — under the right conditions — is increasingly limiting team capabilities.
2. Integrate AI with your existing productivity stack.
The real unlock isn't the chatbot interface — it's integrating AI into your other systems to unlock new possibilities. If you're on Google Workspace, Gemini's native integration changes the equation to connect to your Google data. All the AI applications now have connectors to your various data stores: enable them and show employees how to use them. But avoid tight coupling to allow you to switch front-end AI applications if needed in the future.
3. Track outcomes, not just adoption.
Our research shows daily AI users achieve 20% productivity gains versus 5% for weekly users. Don't measure "how many people have accounts." Measure: Are they using it every day? Are they completing work faster? Are they producing higher-quality output? Then, create performance incentives to guide them to the right results.
The AI race just got a lot more interesting. OpenAI built this industry, but Google just proved that nobody's lead is permanent. For organizations operationalizing AI, that's actually good news — competition drives innovation, prices down, and options up.
The winners will be the companies that stay nimble, build multi-platform fluency, and focus on results rather than vendor loyalty.
What's your experience been with Gemini 3 versus ChatGPT? Hit reply and let me know.
About Trent: Trent Gillespie is an AI Keynote Speaker, CEO of Stellis AI, former Amazon leader, and advisor on building AI-Native, AI-Enabled businesses. Book Trent to speak to your group or book a call to discuss using AI within your business.
Tell us what you thought of today's email. |
Did someone forward this newsletter to you? If you're not already signed up, you can subscribe to AI SPRINT™ for free here.