Imagine this. It's Saturday morning. You're already tired. The week was, you know, long. And the weekend. Well, it's not looking much better. Oh, I know this feeling. You got to get your daughter to soccer at 9 a .m. sharp. You need to do the weekly grocery run. That's 45 minutes if you're fast. Then you got to hit the post office. But, and here's the tricky part, their windows only open from 10 to 1. And on top of all that, you have a hard stop. A Zoom call at
exactly 12 .00 p .m. This sounds like a nightmare. That sounds like my actual Saturday last week. Right. It's the classic chaotic Saturday. Feels impossible. Now, if you give that whole mess to most AIs today, they just fail. They just spit the list back at you. They don't see the friction. But there is one hidden switch one toggle buried deep in the settings that fixes it instantly. And that's just one of them. We're talking about buying back like 10 hours of your
week. Easy. Welcome to the Deep Dive. Today we're taking a look at a guide from Max Anne over at AI Fire. It's called 6 Best Gemini Settings to Enable in 2026 Power User Guide, specifically for Gemini 3 .0. Which is a beast of a model. But the idea here is really interesting. Max Ann says most people are driving a Ferrari like it's a golf cart. What do you mean by that? Break that down for me. Well, most of us use these incredibly powerful tools, you know, the hard
way. It's that rinse and repeat cycle. You type a prompt, you get a kind of meh answer, you get annoyed, you type it again. It's manual labor. It is. What this guide points out is that Google has buried six specific features inside Gemini. And they are all turned off by default. Off by default. That seems weird. Why would they hide their best stuff? Well, I think they want a super simple interface for most people. They want it to just feel like a chat. But these aren't just
tips. They're actual toggles in the settings. And if you flip them, you stop using a chatbot and start using a real proactive assistant. We're going to go through all six. Okay, let's unpack this. We can start with that chaotic Saturday problem. This brings us to the first setting. Thinking mode. Thinking mode, yeah. This is all about shifting from just speed to actual strategy. Most of us are used to the default, right? I ask, what's the capital of France? And it says
Paris before I can blink. We think that speed is the goal. Right. That's fast mode. And it's great for trivia. It's great for, you know, defined photosynthesis. But think about how our brains work. There's system one, thinking fast, gut reaction, and system two, which is slow and deliberate. When you ask a normal AI to solve a logic puzzle or debug code or plan that nightmare schedule, fast mode is a disaster. It's all system one. It just hallucinates because it's rushing to
give you an answer. It's choosing speed over being correct. Exactly. Thinking mode forces the AI to just pause. It literally stops. It maps out a plan, breaks the problem down, and it checks its own logic before it types a single word. Okay, so let's go back to that Saturday. Soccer at 9 -0 -0. Groceries. Post office. Zoom call at noon. What does thinking mode do differently? So fast mode just gives you the list back. It doesn't see the problem. Thinking mode looks
at it and goes, hold on. It spots the 40 -minute gaps. It realizes, hey, the grocery store is right next to the post office, so let's do those together to save a trip. And make sure you're not rushing back for the Zoom call. It builds a real logistical plan, not just a to -do list. It's doing the mental heavy lifting that usually wears a person out. That's it. It acts like a project manager. You'd use this for anything, with moving parts budgets, coding, planning a
trip. It's on Gemini Pro, so it's about $20 a month, but... I mean, if it saves you from one huge mistake, it's paid for itself. So are you saying the delay is actually the feature here? Yes. The silence prevents hallucinations and ensures logic. That's a huge shift in thinking. We're so used to instant that silence feels like a bug. But here, the silence is the work being done. It's the AI taking a deep breath before it answers. All right, let's get to the second
setting. This one fixes something that I think drives all of us crazy. The context problem. Every time I open a new chat, I feel like I have to reintroduce myself. I'm a podcaster. I focus on deep dives. Please don't give me shallow listicles. And please, please don't talk to me like a corporate
HR bot. Exactly. It's just friction. The fix is... custom instructions this is your digital id card for the ai it's buried in settings then settings and help and instructions for gemini think of it like a permanent cheat sheet for you so you set it up once and it just knows who you are from then on every single time it's like handing a new employee a user manual for how to work with you you tell it your role like freelance graphic designer you tell it your goal help me
build a stronger portfolio but here's the best part You tell it what you don't like. The anti -preferences. Yes. You can straight up say, no jargon, no formal, corporate speak, keep answers short and actionable. Max and even suggest you can add things like what you're studying or the kind of business you run. It seems like such a small thing, but the time you'd save not having to type for context over and over again must be huge. It completely changes the conversation.
You stop getting that generic one -size -fits -all advice and you start getting strategy that's actually built for you. And the best part is this one is free for everybody. So does this really stop the generic advice? Absolutely. It swaps generic fluff for tailored strategy. I like that. Swapping fluff for strategy. Okay, this next one. This is where I feel like the assistant idea really starts to become real. Workspace extensions. Oh, this is the big one.
This is what breaks it out of the chat box. The source uses a great analogy for this. It says using Gemini without extensions is like hiring a brilliant assistant, but then locking them in an empty room with just the notebook. And you're on the outside shouting through the door, hey, what did I say in that email last Tuesday? The assistant has no idea. They're smart, but they're completely blind. So how do we open the
door? You go into your settings, find connected apps, and you toggle Google Workspace to on end. And suddenly, Gemini can see. It can use the at workspace command to connect to your Gmail, your Drive, your calendar. Okay, give me a sense of the scale here. What can it actually do with that? Okay, so imagine this. You can type, summarize the key metrics from all of my Q4 marketing reports. Now think about that. You might have like... 50 PDFs scattered across a bunch of folders in
your drive. Gemini scans them all, instantly pulls out the data, and builds you a summary table. That is, wow. That is wildly powerful. I mean, it's basically magic. Seriously, it's a real whoa moment. Just imagine asking a human. to read a year's worth of PDFs and having the answer in, what, 10 seconds? Yeah. Or Gmail. You can say, find all my emails about Project Phoenix, group them by who sent them, and tell me which ones need a response today. It just
cuts through all the noise. It synthesizes your entire digital life. And again, this feature is free. It works on both the free and the pro plans. So it actually leaves the chat window to go get information. It sees your digital life to synthesize data. That capability alone feels like it changes what a file even is. If the AI can read everything, I don't have to be perfectly organized, which is a relief. It makes search feel obsolete and replaces it with answer. All
right, let's talk about maintenance. You know, I have to admit something here. I am absolutely that person. I have great intentions. I say I'm going to check my analytics every Monday morning. I say I'm going to audit my subscriptions every month. And then life just, you know, gets in the way. Work piles up. Yeah. And I just don't do it. I still wrestle with that simple consistency. It's a human flaw. That's where scheduled tasks comes in. This is the AI night shift. How does
that work? It's basically autopilot mode. You describe a task one time, set a schedule, and Gemini just runs it in the background for you. It doesn't wait for you to log in or ask. It just does it. So for my analytics example. You'd set it up once. Pull the weekly analytics from my drive every Monday morning and email me a summary. And that's it. Monday morning you wake up and the email's just sitting there. You didn't even have to remember to ask. Or like a daily
news briefing. Totally. Send me the top three tech stories at 9 a .m. Or a monthly budget check. Summarize my spending and flag any unusual expenses on the first of the month. It turns the chatbot into a system that works for you. This is a pro feature, but again, if it saves you from paying for one subscription you forgot about, it's worth it. So why not just use a calendar reminder for that? This doesn't just remind you. It does the work. That's a key distinction. A reminder is
another thing on my to -do list. This takes things off of it. Boom. Exactly. Okay. Moving on. We've talked about thinking, context, accessing files, scheduling. But what happens when you hit that knowledge wall? The knowledge wall. You know that feeling. The AI gives you a confident answer, but then you ask about something really new or super niche and it just starts to wobble or just make something up. Right. Because the standard model is just pulling from its training data,
which is like a frozen snapshot in time. It's not looking at the world right now. That's what deep research is for. This is the autonomous analyst. How is this different from thinking mode? Great question. Max and Audra has a really clear line. Thinking mode is like locking yourself in a room to solve a puzzle with the info you have. It's about internal logic. Deep research is like sending a junior analyst to the library to go find new information. It leaves the room.
It sends an agent out onto the live web. It reads dozens of articles, academic papers, whatever it needs to. It cross -checks the claims. Then, and this is the key part, it doesn't just give you a paragraph. It writes a structured report with links to its sources. So if I ask, what are the top five AI breakthroughs in the last 90 days? A standard chat might give you a vague
list from six months ago. Deep research goes out, finds the actual news, verifies the dates, synthesizes what it means for, say, a marketer, if that's your job, and then it hands you a whole document. It turns hours of Googling into a few minutes of waiting. And the cost on that? You get five reports a month on the free plan, but it's unlimited on Pro. So is this just a fancy Google search? No. It reads, synthesizes, and writes a structured report. It's the synthesis
that matters. Google gives you links. This gives you understanding. That's it. It does the reading for you. Now, we have to talk about the last setting. In an age where data privacy is, well, it's everything. Sometimes I hesitate to ask an AI certain things. You know, maybe it's a sensitive financial question or something about a client who is under an NDA. That hesitation is healthy. You should be careful. But it shouldn't stop you from using the tool. And that's why
temporary chats exists. This is the NDA protocol. What happens when you flip this switch on? So you go to activity controls or there's a little toggle at the top of the chat. When it's on, nothing is saved. Nothing is used for training the model. The second you close that window, the entire conversation just evaporates. So it's completely off the record. Totally off the record. It gives you all of Gemini's brain power, the thinking mode, the logic, all of it, but with
a zero digital footprint. It's perfect for medical questions, drafting legal ideas, financial planning. And it doesn't lose any intelligence in this mode. Full reasoning power, just zero digital footprint. That's really reassuring to know there's a kill switch for its memory. It's essential for any kind of professional use, I think. So we've gone through six settings. Thinking mode, custom instructions, workspace extensions, scheduled tasks, deep research, and temporary chats. The
hidden six. We've walked through the list one by one. But as we usually find on this show, the real power isn't in the individual pieces. Max Anne argues the big competitive advantage is what you do with them together. It's about chaining. Chaining. Right. Don't just use one at a time. Stack them. Think of them like Lego blocks of data. Walk me through what a chain would look like. Okay. Step one. Use deep research to get fresh data on a new market trend. Gives
you that report. Step two. You feed that report into thinking mode and tell it to develop a strategy based on your custom instructions. Right. Then step three. You set up a scheduled task to go out and update that strategy with new research every Monday morning while you're asleep. You're not just asking questions anymore. You're building a system. You're building a machine that works for you. That's the real difference between a regular user and a power user. Most people are
just chatting. You could be orchestrating. It really shifts the role of the person from being the doer to being the architect. Exactly. And the wildest part of all this, most people will never turn these on. They'll just stick with the defaults. Which means what? It means if you do it, you have an instant advantage. You're playing with a full deck while everyone else is missing a few cards. I think that's the big
idea to sit with. Yeah. These tools are incredibly powerful, but only if we actually take the time to configure them. Just open the settings, toggle just one of them. Start with custom instructions. It takes two minutes. Two minutes to save 10 hours a week. That's a pretty good trade. It's the best trade. Thank you for listening to this deep dive. Go explore those settings. We'll see you next time.