Playing this video loads it from YouTube (Google), which may store cookies on your device and receive your IP address. Click to play.
Video: "9 Google AI New Features That Take Gemini Out of the Chat Box" by Julian Goldie on YouTube.
What changed on 23 September
Before connected apps, Gemini was a chat interface. You typed a prompt, got a text answer, then did the actual work yourself - copying the output into Airtable, writing it into a contract template, pasting it into your project management tool. It was useful as a thinking aid, not as something that completed tasks.
Connected apps change that. Gemini now has direct write access to the tools in the integration list. It can create a task in Airtable without you opening Airtable. It can draft a contract in PandaDoc and populate it from context you have already given. It can push a content update to a Squarespace or Webflow site. The chat box is still there; what comes out the other end now includes actual actions, not just text.
Julian's old-way vs new-way comparison
Julian Goldie ran six common weekly jobs through both approaches. The old workflow was the same for all six: write a prompt, get a text response, open the relevant app, manually transfer the output, do the final step yourself. Functional, but each job had four or five steps where only one or two were actually thinking work.
The new workflow collapses most of that. He described how for project tracking - previously: ask Gemini what to add, copy the output, open monday.com, create the task manually - the new flow is: tell Gemini to add it, and the task appears. Same result, two fewer steps, none of the context-switching. He ran the same pattern across content drafting, contract prep, website copy updates, research collation, and supplier outreach, with broadly similar results.
In practice, a few of the connections are smoother than others. Webflow and Squarespace publishing works but requires confirmation before Gemini commits any changes. PandaDoc contract drafting works well for standard templates; highly customised documents still need manual review. None of that is surprising for an initial rollout.
What's genuinely new versus what was already possible
Worth knowing: most of what Gemini connected apps offers was already achievable with a well-configured Zapier or Make workflow. The difference is that Gemini interprets natural language and decides which action to take, rather than you having to define the trigger-action logic up front. That is a real reduction in setup time for ad-hoc tasks. For high-volume repeating workflows, a proper automation build is still more reliable - an AI deciding what to do each time introduces variability you may not want in a process that runs fifty times a day.
The connected apps rollout also covers Peloton and a handful of consumer apps alongside the business tools. The Peloton integration in a B2B product is a reasonable sign that this is a broad platform play, not just a productivity tool - Google wants Gemini embedded in daily life across categories, not just in the workday.
The Googlebook and what it means
Alongside the connected apps rollout, Google launched the Googlebook laptop in the US on 4 October 2026, with the UK, Canada, France, Germany, and Australia following on 5 October. The Googlebook has Gemini built in at the OS level - not as a browser tab, but as an always-available layer across every app. It is the connected apps capability in hardware form: Gemini aware of what you are working on and ready to act on any of the connected services without switching context.
For UK businesses, the Googlebook is available in stores from today. Whether it makes sense over a standard laptop depends entirely on whether you use the Google and connected-app ecosystem heavily. If you live in Google Workspace, Squarespace, and monday.com, the integration is genuinely useful. If you use Microsoft 365, it is considerably less compelling.
Where this connects to NordSys
Connected apps is what an automation flow looks like when an AI model provides the decision layer rather than a trigger-action rule. The practical result - Gemini reads a brief, picks the right app, creates the task or draft or publish - is the kind of workflow a properly set-up automation could already handle for your business. Whether you use Gemini, Make, Zapier, or a custom agent stack, the underlying question is the same: which of your repetitive tasks should a machine be doing? Our Programming & Automation service is built to answer that for real businesses.
See our Programming & Automation service →