Weekly Brief 23.-30 Nov - Week 48
- Kristine Lium
- 3 days ago
- 4 min read
Updated: 3 days ago
Gemini 3: Three Changes You’ll Actually Notice
Your Weekly Brief — clear, simple, and ready to share.
What to Watch This Week:
AI releases can easily feel overwhelming. Some people follow them closely. Others barely register that anything has changed.So here’s a simple starting point:
Google released Gemini 3 last week. You don’t HAVE to know the benchmarks — but here are three shifts that actually matter for everyday work.
Let’s make that as clear and human as possible..
This is the Weekly Brief is a short and condensed way for you to stay on top and part of the conversation - Every Week
Your executive briefing week 48
It reasons more like we do — and you choose how deeply it “thinks”
Humans don’t always think in the same way. Sometimes we give quick answers (“top of mind”). Sometimes we pause, consider context, compare options, and reason step by step.
Gemini 3 introduces a similar idea.Google calls it “thinking levels,” but what it really means is:
Quick mode → short, fast answers
Deep mode → slower, more structured reasoning
What human-style reasoning looks like in AI
Think of a simple task:
“We’re launching a new product. What should we focus on first?”
A quick answer might list best practices.A deeper answer would:
break the problem into steps
consider constraints
compare approaches
show why those choices make sense
That’s much closer to how humans reason when something matters.
What this helps you do
Analyse documents or decisions with clearer logic
Get more structured breakdowns
Ask bigger questions without getting vague answers
Important note
Deeper reasoning does not mean perfect reasoning.It simply means more steps, more structure, and more context awareness — with the same need for human judgement on top.
It finally understands mixed inputs — and can respond visually
Previous models struggled with multimodality because:
input formats were treated separately (text in one place, images in another)
context windows were too small to hold everything at once
the model couldn’t “connect” these formats into a single analytical chain
With Gemini 3, the underlying architecture and enormous context window remove many of these barriers.
Put simply:
You can give it mixed information in one place — text, screenshots, documents, photos, even video — and the model can analyse them together.
Example you can try today
Prompt:
“Here are a few slides from a workshop, a screenshot from our CRM dashboard, and a photo of notes from the whiteboard. Summarise the key challenges, turn them into three themes, and create a simple diagram showing how these themes connect.”What Gemini 3 can return
A short written summary
Three clear themes
A generated diagram (arrows, clusters, or a simple flow)
Optional: a suggested structure for a slide or concept draft
This is where the breakthrough becomes practical:
the model doesn’t just describe — it visualises. Something that previously required manually stitching tools together now fits in a single prompt.
It shows up where people actually work
Gemini 3 isn’t arriving as “another AI app. ”It’s being integrated directly into:
Google Search (AI Mode)
Gmail
Docs
Slides
Sheets
The Gemini mobile app
Enterprise tools via Vertex AI
For many people, this will be the first time advanced reasoning and multimodal support appear in tools they already use. Not in a separate tab, not in a special playground — but inside daily workflows.
What changes for everyday users
Search results may contain more visual answers
Emails can be summarised with mixed context
Docs and Slides can draft first versions from screenshots or files
Teams can collaborate around richer AI-generated outputs
This is where AI becomes less “something you try occasionally” and more “something quietly embedded in the workday.”
💬 Closing reflection
These three shifts —reasoning, multimodality, integration —are the clearest entry points for understanding why Gemini 3 is a meaningful update.
But they’re only the starting point.
🔮 What to watch
This is just touching on what it seems as Gemini 3 is making available so for next week's briefing will do a deeper dive to see
how Gemini 3 can run multi-step workflows
what “AI agents” actually do
how developers can build apps faster with tools like Antigravity
what this means for businesses, roles, and workflows
Part 1 is about noticing the change. and next week we'll look at what you can do with these new abilities at the tip of your fingers.
💡Weekly kick-off
👉 Your Weekly Brief — clear, simple, and ready to share.
Do you have any curious topics or interesting guests we should invite on the podcast?
Share your ideas by sending us an email:
Don't miss out on the next episode of the Simply Briefed podcast launching this Wednesday
What makes an idea worth betting on?
Together with Venture Capital Partner of Skyfall Ventures, Jon Kåre Stene, we are looking into the current speed of development and how innovative ideas pop up faster and more often than we've seen before. At the same time, trying to understand how to spot what are good ideas with the potential of actually making a difference and which ones are just ideas.
If you can't wait and need an episode to keep you going - don't miss the previous episode with Linn Solvang - How ready are we REALLY, for AI at work?
SIMPLY BRIEFED
The podcast that take complex topics and makes them easier to understand