• Everyday AI
  • Posts
  • ChatGPT Connectors: What they are and why you NEED to rely on them daily

ChatGPT Connectors: What they are and why you NEED to rely on them daily

Reddit talking bigger AI partnership with Google, Gemini and ChatGPT take home coding gold, Jensen Huang going bananas and more

Author

Everyday AI
September 17, 2025

Sup y’all 👋

I’ve had a hunch lately… the people want to learn more about AI Agents and how to actually use them.

Been thinking about dedicated more episodes and content here in the newsletter. But only if y’all want it.

What’s your take?

Should we start covering AI Agents more?

(Vote to see results)

Login or Subscribe to participate in polls.

Oh, today’s show was apparently a banger.

(Thanks Mike)

(Make sure to check the bonus at the end of this newsletter.)

✌️

Jordan

Outsmart The Future

Today in Everyday AI
8 minute read

🎙 Daily Podcast Episode: ChatGPT Connectors are the elementary RAG replacement cheat code that few people know. So we had to change that. Give it a listen.

🕵️‍♂️ Fresh Finds: NVIDIA faces Chinese chip ban, Meta is throwing weight and money into politics and Anthropic is shutting down AI use for surveillance. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Reddit talking bigger AI partnership with Google, Gemini and ChatGPT take home coding gold, Jensen Huang going bananas and more.  Read on for Byte Sized News.

🧠 Learn & Leveraging AI: So how do connectors work, and why should you RELY on them when using ChatGPT? Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about ChatGPT tightens teen safeguards, Fiverr cuts 30% of staff due to AI, Workday and Microsoft’s agentic collab and more. Check it here!

 The Last Lectures: Why colleges still running from AI in 2025 will eventually die 🎓️

Remember RAG? 🤔

When companies would spend multiple six or seven figures (and sometimes a year or more) trying to connect their enterprise data to LLMs?

Now there's an easier way.

ChatGPT connectors.

In a few seconds, you can get like 80% of RAG's power for like free 99.

Yet... so few people actually use them.

Let's change that and put AI to Work this Wednesday.

Also on the pod today:

• Connectors vs. RAG: Key differences 🔌
• Instant Google Drive integration ⚡
• Multi-connector business workflows 🔄

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 – Crustdata uses AI to let you find info on any person or company, Capalyze is an AI agent that scrapes, analyzes and presents data, Codewords uses AI to turn ideas into workflows.

Global AI Race — China has banned some tech companies from buying NVIDIA chips. See how that may shape the global economy.

LLM Progress — OpenAI’s models are getting so good, that apparently it’s harder for humans to find problems they can’t solve.

AI and Politics — Mark Zuckerberg is using Meta’s own super PAC to shape California AI policy and pressure rivals

AI Video — YouTube has rolled out new GenAI tools for Google shorts. More AI masterpieces or slop? We’ll have to wait and see.

AI and Surveillance — Anthropic’s surveillance ban sparks Washington backlash while OpenAI courts authorized use

AI and Education — Google just released a new AI tool to turn textbooks into interactive AI guides. See how it works. 

 

1. Reddit eyes richer AI tie-up with Google 🤝

Reddit is reportedly negotiating a deeper partnership with Google that would thread Reddit discussions more directly into AI Overviews and other Google AI products.

The talks reportedly include dynamic pricing so payouts from Google or OpenAI rise as Reddit’s data becomes more central to AI answers, building on earlier licensing deals worth roughly $60 million with Google and $203 million overall. If finalized, expect more Reddit threads to shape AI summaries and search results, which could shift visibility for brands, SEOs and creators who rely on Google for discovery.

2. Gemini and OpenAI clinch ICPC gold-level performance 🥇

An advanced version of Google Gemini 2.5 Deep Think hit gold-medal level at the 2025 ICPC World Finals on September 4 in Baku, solving 10 of 12 problems under contest rules and even cracking one problem no human team solved.

The system previously scored gold at the IMO, signaling a rapid climb in abstract reasoning across both math and competitive programming. Google DeepMind says the model combined multi-step reasoning, code execution, and reinforcement learning with parallel agents to iterate and verify solutions, and its run would have ranked second among university teams by total time.

OpenAI also announced on Twitter that they also received Gold status, but solved all 12 of 12 problems.

3. Jensen Huang Hypes Nano Banana in London 🍌

Nvidia CEO Jensen Huang used a London presser to gush about Gemini’s Nano Banana image tool, linking his enthusiasm to a reported 300 million image surge on Gemini in early September, according to Josh Woodward of Google Labs on X.

Huang also detailed his daily AI stack, saying he rotates among Gemini, Grok, Perplexity, and ChatGPT and has them critique each other’s outputs for serious work.

4. ElevenLabs launches Studio 3.0 with full-stack AI audio tools 🎙️

According to ElevenLabs, the new Studio 3.0 rolls out an all-in-one editor that blends AI voiceovers, auto-scored music, sound effects, captions, multilingual support, and video editing into a single timeline.

The update ties directly into its API and broader ecosystem, letting users assign speakers, fix flubs with text edits, isolate noisy audio, and generate custom music, which pushes AI audio from a niche add-on to a full production workflow. For creators and teams trying to grow a brand or business, this means faster turnarounds, fewer tools to juggle, and lower costs for podcasting, audiobooks, and video content.

5. Microsoft, Nvidia and tech heavyweights pour billions into U.K. AI buildout 💸

According to CNBC, Microsoft will invest $30 billion in the U.K. by 2028 to build AI infrastructure including what it calls the country’s largest supercomputer with more than 23,000 GPUs in partnership with Nscale.

Nvidia, Google, OpenAI, Salesforce and CoreWeave piled on with additional multibillion-dollar commitments, including Nvidia’s plan to deploy 120,000 Blackwell GPUs and Google’s new Waltham Cross data center, timed with President Trump’s state visit and a planned U.S.-U.K. tech pact on AI, Quantum and Nuclear.

🦾How You Can Leverage:

About 18 months ago, a RAG-enabled LLM system cost like half a million and took like 6 months. If you were lucky. 

Now? 

You can get 80% of the way there in 30 seconds and $20. Zero tech knowledge required. 

Mind blown emoji.  

While large language models of yesterday readily struggled with generic outputs and lacked your dynamic business context, today’s versions have simple Retrieval Augmented Generation on demand. 

Yet, so few companies we consult with use ChatGPT’s new connectors feature, or even know what they are. 

So, you know we had to change that with today’s ‘AI Working Wednesday’ series, giving ChatGPT Connectors the Deep Dive. 

Spoiler alert: this is a banger episode. 

Ready to dive in? 

Let’s get it. 

1 – Your enterprise is over-engineering everything 🔧

Companies are hiring consultants to teach employees "context engineering."

That's the fancy term for manually feeding ChatGPT better information before asking questions.

Meanwhile, connectors automate that entire process with like three clicks. Your CRM talks to your email talks to your calendar through ChatGPT's brain, creating business value you literally couldn't access before.

The real problem? Enterprises love complicated solutions even when simple ones exist. They'll spend $500K building custom RAG systems while ignoring the $20 feature that solves the same problem.

It's like buying a Ferrari and using it as a really expensive paperweight.

Try This: 

Check how many people on your team actually have connectors enabled right now. If you're on Enterprise, verify with your admin that connectors are allowed. Run one test with HubSpot and Gmail to summarize client status before your next meeting. Time that versus your normal prep routine and prepare to question every expensive workflow you've built.

2 –  Context switching is murdering your productivity 🧠

The biggest productivity killer isn't time spent in apps.

It's your brain constantly switching between search mode, analysis mode, and execution mode.

We demonstrated this with real podcast prep that normally takes three to six hours of scattered attention. Checking calendar details, hunting through email threads, researching guests, pulling materials from hundreds of poorly-named files.

Your brain gets hijacked every single time you open Gmail and see seventeen urgent messages. Then you open Calendar and spot twelve overdue tasks. Then you dive into research and lose two hours down internet rabbit holes.

With connectors, ChatGPT handled everything in four minutes while we grabbed coffee.

Focus preservation beats time savings every time. You make decisions instead of hunting for information. Your mental energy goes toward strategy instead of digital scavenger hunts.

Try This: 

Pick your most mentally exhausting weekly routine that involves multiple systems. Document every rabbit hole you fall into. Connect those apps to ChatGPT and write one prompt that handles the entire information workflow. You're aiming for like 90% reduction in context switching so you can redirect that brainpower toward actual business decisions.

3 – Your data silos are costing you millions 💰

Most companies have systems that can't talk to each other.

HubSpot doesn't know what's in your Gmail. Your calendar has no clue what's happening in Notion. SharePoint lives in its own little world.

This creates invisible productivity tax that compounds daily across every employee. People waste hours manually connecting information that should be connected automatically.

Connectors solve this by turning ChatGPT into a universal translator for your entire tech stack. Suddenly your CRM knows about your emails, your calendar talks to your project management, and your documents connect to everything else.

That's new business intelligence you literally couldn't access before. Patterns emerge across systems. Decisions get made with complete context instead of partial information.

Most enterprise teams are sitting on this goldmine and don't even know it exists.

Try This:

Identify your biggest data silo problem right now - like client info scattered across HubSpot, Gmail, and Teams. Connect those three systems to ChatGPT and ask it to give you a complete client picture before your next important call. Notice how different your preparation feels when you have unified context instead of fragmented information.

 🚨 Bonus Materials 🚨

What’s that… today’s episode wasn’t enough?

You want it all done for you?

OK, we created this insanely valuable ChatGPT Connectors Cheat Sheet. Just go share today’s episode on LinkedIn to get access.

Reply

or to participate.