The argument in one line.
The real AI edge is not prompting skill but decision clarity — knowing which work belongs entirely to AI, which needs AI as a co-pilot, and which must stay human to be worth anything at all.
Read if. Skip if.
READ IF YOU ARE…
- A business owner or operator who has started using AI but feels overwhelmed trying to figure out what to delegate versus what to keep.
- A founder or team lead who suspects they are wasting senior time on tasks AI could handle, but lacks a framework to sort them.
- Anyone who has heard that AI will replace their job and wants a clear answer on what the machine actually cannot do.
SKIP IF…
- You are a developer looking for a technical AI integration guide — this is a strategic framework talk, not a build tutorial.
- You want a deep dive on any single AI tool; this is tool-agnostic and stays at framework altitude throughout.
The full version, fast.
The GAIN Matrix splits every task by two axes: easy/hard for humans, easy/hard for computers. Give quadrant tasks to AI entirely (meeting notes, email triage, CRM updates). Accelerate tasks where AI buys back days of manual work (research, data analysis, market breakdowns). Integrate when the output needs human taste, vision, or care to be any good. Reserve No AI for leadership, persuasion, coaching, and networking — the moments that require your presence, not a prompt. The person who understands the problem best wins, not the person with the most AI tools.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Where the time goes.
00:00 – 01:22
01 · Cold open + problem frame
AI is easy to win with if you stop thinking about it the wrong way. Winning = man plus tool, not just tool.
01:22 – 01:43
02 · The GAIN Matrix intro
Introduces the 2x2 whiteboard framework: Give, Accelerate, Integrate, No AI.
01:43 – 04:09
03 · G — Give to AI
Tasks easy for both humans and computers: meeting notes, inbox triage, CRM, reports. Rule of 3 Rs for when to automate.
04:09 – 06:57
04 · A — Accelerate with AI
Hard for humans, easy for computers: research, data analysis, market scans. The 10-80-10 rule. AI buys back days, not minutes.
06:57 – 11:11
05 · I — Integrate with AI
Hard for both: creative work, product prototyping, strategy. AI generates options; human has taste, vision, care. Workbook example.
11:11 – 13:40
06 · N — No AI
Easy for humans, impossible for machines: leadership, persuasion, coaching, networking. The irreplaceable quadrant.
13:40 – 14:51
07 · Wrap + CTA
Every task lives in one of these quadrants. Instagram DM for free AI tech stack.
Lines worth screenshotting.
- AI doesn't solve problems — the person using it solves problems. AI just does stuff.
- The person who understands the problem best wins in the AI era, not the person who understands the tool best.
- Winning with AI is not about prompting better — it is about deciding which tasks should never reach your desk at all.
- 80% done by AI is 100% awesome — perfecting what AI started is faster than starting from scratch yourself.
- The 10-80-10 rule: 10% human ideation, 80% AI execution, 10% human quality pass.
- AI can generate 2,700 versions of anything — it still takes a human with taste to pick the right one.
- The Rule of 3 Rs: only automate tasks that are Repetitive, Rules-based, and Return more time than the build costs.
- No AI is not a limitation — it is the quadrant where you become irreplaceable.
- Business is not B2B. Business is H2H — human to human. AI removes the friction that keeps humans apart.
- Leadership, persuasion, coaching, and networking are nearly impossible for machines — protect them as your competitive moat.
- Most people are busy because they are stuck between tasks that belong to them and tasks that belong to AI.
- AI without experience or taste produces mediocre output — the Integrate quadrant only works if you already know what good looks like.
- The GDP labor data already shows which jobs AI is eating fastest — the ones that live in the Give and Accelerate quadrants.
Every task maps to one of four AI relationships.
WHAT TO LEARN
The question is never whether to use AI — it is which of four collaboration modes applies to this specific task right now.
- Not all tasks deserve the same AI treatment: some should be fully automated, some accelerated, some co-directed, and some kept entirely human.
- The Rule of 3 Rs (Repetitive, Rules-based, Returns time) is a fast filter: if all three are true, hand the task to AI today without overthinking it.
- The 10-80-10 split prevents both over-delegation and under-delegation — start the brief yourself, let AI do the bulk, finish it yourself.
- Creative and strategic work still requires human taste and vision: AI generates options but cannot judge which option fits the context, the culture, or the moment.
- Leadership, persuasion, coaching, and networking are structurally human — not because AI lacks raw capability but because the value comes from the human being the one who shows up.
- The most dangerous trap is staying stuck between two quadrants: doing tasks that belong entirely to AI, or over-delegating tasks that need your judgment in the loop.
- Irreplaceability in an AI economy is built in the No AI quadrant — the more you protect time for presence-based human work, the harder you are to automate away.
Terms worth knowing.
- GAIN Matrix
- A 2x2 decision framework that maps tasks by how easy they are for humans (vertical axis) and for computers (horizontal axis), producing four quadrants: Give, Accelerate, Integrate, and No AI.
- Give quadrant
- Tasks that are easy for both humans and computers — the work you should automate or delegate to AI immediately, such as meeting notes, email summaries, and CRM updates.
- Accelerate quadrant
- Tasks that are hard for humans but easy for computers — research, data analysis, pattern detection — where AI buys back large blocks of human time.
- Integrate quadrant
- Tasks that are hard for both humans and computers — creative work, strategy, prototyping — where AI generates options but a human must direct, judge, and finalize.
- No AI quadrant
- Tasks that are easy for humans but hard for computers — leadership, persuasion, coaching, relationship building — the irreplaceable human work.
- 10-80-10 rule
- A workflow split: the human contributes 10% ideation and framing, AI does 80% of the execution, and the human returns for the final 10% quality and judgment pass.
- Rule of 3 Rs
- A test for automation: only build the workflow if the task is Repetitive (at least weekly), Rules-based (same inputs/outputs each time), and Returns more time than the build costs.
- H2H (Human to Human)
- The speaker's reframe of B2B — the idea that all business ultimately runs on human relationships, and AI's job is to remove the friction that keeps humans from connecting, not to replace the connection itself.
Things they pointed at.
09:09bookBuy Back Your Time (Dan Martell)
Lines you could clip.
00:22
“AI doesn't actually solve problems. The person using it solves problems. AI just does stuff.”
01:20
“The person who understands the problem the best, they'll actually win. Not the person who understands the tool the best.”
06:57
“80% done by somebody else or AI is a 100% awesome.”
11:11
“No AI is actually the quadrant that makes you a weapon — irreplaceable.”
11:22
“I don't believe business is B2B. I think business is H2H, human to human.”
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
metaphoranalogy
Winning with AI in 2026 is actually pretty easy. Most people are just thinking about it the wrong way. And I'm not gonna tell you how to get better at prompting or use this AI tool.
That won't get you anywhere unless you use a simple framework. I've built and scaled several companies past a million dollars in under six months through Martell Ventures. And the most important thing to understand first and foremost is AI doesn't actually solve problems.
The person using it solves problems. AI just does stuff. For the longest time, winning has been man plus tool, man plus laptop, man plus bike, and it still is man plus tool.
So as long as that's still the equation, then your job is to be a better man that uses the tool. Like if I just gave you a computer and you didn't know how to use it, it wouldn't make you any money. And a lot of people think, now that I've got AI, I should be rich.
Unfortunately, AI isn't gonna make you rich. You being the person using it and the tool, that's how you get rich. And the truth is some tasks belong to you and some tasks belong to the AI.
And if you're stuck between both of them, you'll just feel really busy, but you won't be making progress. In this AI era, the person who understands the problem the best, they'll actually win. Not the person who understands the tool the best.
And to figure out what problems you should be solving with AI, I've come up with a simple framework. I call it the game matrix. There's four letters in four quadrants.
The left side is easy for computers. The right side is hard for computers. The top is easy for humans.
The bottom is hard for humans. Each quadrant will show you how to collaborate with AI to get things done. And the easiest one to start with is give it to AI.
This is easy for humans, but also incredibly easy for computers. Think about all the stuff that you started using ChatGPT for. Things like synthesizing text or looking over emails or updating your CRM or creating spreadsheets.
Things that you could do, not hard, but the AI does it way better, way faster. Yesterday, I was working with my finance guy and he was working on a new proposal and I was like, bro, stop writing. Just talk and give it to AI.
It'll analyze the financials, it'll analyze your words, it'll put it together and it'll get us a table that gets us clear on the decisions we have to make to move it forward. Why are we trying to do stuff that's so wildly easy for AI that takes our time just because we can do it? These are just a small tedious task that AI can do and you just want to give it to it and automate it.
I'm talking about things like meeting notes. Right? I use granola.
It's one of the best tools out there, but essentially every meeting is recorded. It's summarized. It's tagged.
It's added to notion. It is the easiest thing. The other one is pulling reports.
Okay? Think about the ability to have dashboards of all the data about your life and your business in one place so you can see if anything's going off, follow the numbers or what you expect them, and it's just a really easy thing for you to do in autonomy. And then the third is to summarize your messages.
Okay? Think Slack. Think email.
Think text messages. You can do a lot of stuff with AI that most people have no idea about. And these are just some examples.
And I know the dopamine hit of clearing out your inbox and getting to inbox zero can feel awesome. But it's still work and you don't have to do it. Let's let AI do it.
So for example, this morning I asked my agent to go check all of my messages, my email, my calendar, my Slack and tell me the top three things that I need to prioritize today based on my goals. It told me I had a YouTube shoot and a real shoot, so I gotta save my energy. It told me that Andy needed a call to review the podcast that I'll be doing on my San Francisco trip, and also said that team needed sign off on a new feature that we're adding to our Apex voice system.
Now I know exactly what I need to do today without scrolling through the countless message, without getting distracted, without doing stuff that I don't need to do. So how do you know when a task belong in the g quadrant? I use what's called the rule of ours.
First, is it repetitive? Is it something I have to do at least once a week? Second, is it rules based?
Does the task have the same inputs and outputs every single time? The third is does it generate a return? Does it take me more time to run this whole automation than just do it every time?
The last thing I want to do is spend sixty hours to build something that takes me less than two minutes a week to do. So if the answer is yes to all three, repetitive, rule based, and gives me a return on my time, then I hand it to the AI today. And look, you can use a ton of different AI tools for this quadrant.
Some are some are awesome. So to help you out, I put together my entire AI tech stack, every tool I use in every business that I own to run a super efficient business. It's a 100% free.
Just DM me the word YouTube stack on Instagram and I'll send it right over to you. Now, that's the easy stuff and with it, you could probably free up a couple hours of your week. Now, we need to move on to the next quadrant where the real work happens.
Quadrant number two, accelerate. I know my handwriting sucks but just listen to my words. Okay?
Accelerate. Now this is where it's hard for humans but easy for computers. Really complicated processes we're trying to find a signal in the noise.
AI strives at this thing. And the cool part is you can accelerate all that work with AI where AI buys you back days of work. Here are just a few things that you can use AI to accelerate with.
First off is research. I use this probably the most every day. I have an idea, I go to my AI.
I go, hey, do some research, find some people, tell me what you think. Hey, I found this link for this cool idea. Tell me based on what I'm using today, is it helpful?
Should I add it? Do deep research projects around strategic decisions. Look at my life.
Tell me is there anything I'm missing? The other one is data analysis. If you're starting off as a business owner for data analysis, think about it.
You have Facebook ads, you can get it analyzed. You have social media data, get it analyzed. You have contracts, why don't you ask it to go analyze all your legal contracts to see if you signed something you probably shouldn't have.
You shouldn't be spending hours on spreadsheets, especially creating anything from scratch. That is where AI shines. Market breakdowns, figure out what's going on in your industry, what's going on in the market.
Every morning, I have multiple prompts that scan and research and figure out what's new in the news. It tells me all the things I should be paying attention to as an innovator in my space. Your job as a leader is to be able to see around the corner.
The news is out there. Have the AI go analyze it. When you think of tasks in this quadrant, these are the three questions you get asked yourself.
One, is it looking for a pattern? Because AI can spot patterns faster than any human. Two, would it take you hours to do manually or would you normally hire really expensive specialists to do this?
Think about like visual designers, researchers, consultants. Fuck that. .
These are the signs that you should accelerate that type of work where you're processing it with AI. I see a lot of people not even accelerate because they're like, I just don't think the output is very good. And they're like, fine.
Then use my ten eighty ten rule. 10% ideation, the input, the prompt, the innovation, have it do the 80% research, have it go do some of the work, and then come in at the very end for the last 10% where you give it the thing that makes it really awesome.
At least the AI took 80% of the work off of your plate. 80% done by somebody else or AI is a 100% awesome. So so far, we've been giving AI just the easy stuff.
This next quadrant is where we start adding the human element to the mix. Quadrant number three, integrate with AI. This is where it's like still hard for computers and hard for humans.
AI can help you get some things done, but the human has to still be the director. And unfortunately, if you don't have experience or taste, it's actually really hard to get great work out of this. And the truth is is the AI isn't perfect.
Did you know if you asked the AI that you're gonna go wash your car and the car washes 50 meters away from where you live and you asked it, should you take your car or walk? It would actually say you should walk. So right off the bat, it can't be left to do really important creative strategic work on its own.
It doesn't have the context. It doesn't have the nuance and this is where having it help you calibrate makes a lot of sense. AI can generate 2,700 versions of anything, but it still takes a human to look at it and go, that feels right.
That is the way I would do that. So here are some things to integrate with AI. Number one is creative work.
I really don't know many people in today's world. I don't care if it's media or ads or brainstorming or design or brand where they're not using AI to come up with illustrations, examples to help shape the idea, to give them something visual they can look at.
Most music today uses AI in some aspect of the production of the beats of the sound. It still takes a human with an ear for what the world or the market or the culture wants to decide how you're gonna finalize and shape a song. Next is digital products.
This for me is like building out MVPs. Think about the prototypes. Think about the apps.
If you haven't tried building something like an app or an artifact or some kind of visualization, you're missing the opportunity. That still requires taste, still requires vision to direct it, but the AI can execute. It can write the code.
It can publish it online. It's kinda wild what it can do for you. And the third is strategy.
Strategy equals sequencing. AI can suggest options but AI doesn't know your personal preference.
It doesn't know other people's psychology. It doesn't understand the situation, the landscape of your world. So it can give you ideas but you still have to make the decision.
So partnering with AI to give you those options is huge opportunity. For example, when I wanted to create a workbook for my book, Buy Back Your Time, I asked the AI, go through my book, pull out the frameworks, and build me something I could look at. The first version was fugly.
I couldn't believe that it would design these diagrams this way, that the workbook was laid out that way, but it at least gave me something to work with that I could prompt and change and prompt and change. Eventually, I got it to a place where I'm like, yes, this looks good. Give it to my designer.
He makes it look great. Now I have a completed workbook. The whole idea is AI is your partner in very creative and strategic work.
So how do you know if a task belongs in this quadrant? There's three checks. One, does it require taste?
Your personal opinion. AI can generate options like ads. It can generate emails, but it still requires you to pick the right one.
Two, does it require vision? This is the strategy. This is the sequencing.
The AI can execute, but it still needs to be told do it in this sequence. That's why most developers build apps in plan mode.
It shows them what the plan is, then you can change the plan, the sequence. It gives you options. The human still decides the sequence or the phases it's gonna build in, then the AI could build it.
And last, does it require care? AI can help you draft messages but you may be the person that needs to deliver it because humans still want to hear from other humans.
How many times have you been really pissed and you wanna send a message and you know you shouldn't? So you sit there and you work with AI to help it craft you 17 different variations. Am I the only one?
That's just me? And then finally you find one that says the thing the way that you know the person's gonna receive and you're like, oh my gosh, crisis averted. That's this box.
I don't think AI replaces humans. It actually takes away the things that have been keeping us apart. So if you answered yes to any of these, integrate it with AI.
So you've handed AI the easy stuff. That's the give. Then you gave it the stuff to accelerate things that take you a lot of time, but it can do it really easily.
Then we went to the eye and we've integrated so we start collaborating with it on some pretty creative work. Now this last quadrant is where you want to spend all of your time. Quadrant number four is no AI.
These are the things that are easy for humans and wildly hard for computers. These are the moments that need your presence, not a prompt. I don't believe business is b to b.
I think business is h to h, human to human. And like I said, the more we can give AI to do the stuff that keeps us away from being more human, the better the human experience is. No AI is actually the quadrant that makes you a weapon irreplaceable.
Now taste, vision, care don't just guide the action, they are the center of it. So here are some things I think you should never outsource to AI. One, leadership.
People don't follow computers. They follow people. People inspire them.
If you're a leader, by definition, it means you have followers. I've never met followers of an AI. There might be some weird people out there, but we'll leave that alone.
I think that your opportunity to show up for people, to believe in them, to communicate, to invest in them, to build the people, that is a very human thing to do, not an AI thing to do. Number two is kind of like persuasion. I think that there are certain situations when you're doing certain negotiations that you need a person involved.
When we're talking business development, we're talking partnerships. Yes, you can use computers for a lot of sales processes, but I'm a big fan.
It's like if it's important to your business, you show up. You look at them in the eye. You sit knee to knee.
When you're negotiating with somebody, that human part is what matters. Number three is coaching. Right?
It's different than training and it's different than just learning. Coaching means you sit down and you help guide somebody to become more. Is This where I say, you build the people and the people build the business.
The best way to do that is to coach. Have them come into your work and shadow you. And number four is networking.
This is where you go out and you talk to people. Hand to hand, voice to voice, and you build relationships with them.
You find ways to add value. Most people's value is in their network. Your network equals your net worth.
Nobody's gonna trust their relationships with an AI. They'll trust you with it because if you're somebody that adds value, they know if they introduce you to somebody, you're gonna add value to their life. Not your AI, not your email response, not your templates.
They wanna talk to you. And this quadrant right here comes down to one question. Does this revolve around your humanity?
If a human is on the other end and they need to feel you, not a machine, then it's no AI. Me on stage is way different than you watching a video of me on stage.
You hanging out with me in this room, way different than you just watching this video. That is where no AI comes into play. So there's four quadrants we gotta give to AI, we gotta accelerate with AI, we gotta integrate it into our life, and then there's no AI.
And any task on your plate lives in one of these quadrants. Because here's the big idea. AI is reshaping every industry, every job, every business.
There's a thing called the GDP val that actually shows you where AI is affecting what jobs at what percentages and how it's gonna continue to evolve. There's just no future where the AI models don't get better and start doing more of the things for you. Learning this and working your way up to no AI where you're the most envisioned, the most taste, the most care, that's where you compete.
So drop a comment below and let me know what's the thing you would love to be doing more of that AI can't help you with because that is your focus and everything else, let's give it to AI. And remember, I put together an entire AI tech stack, every tool I use to run my $100,000,000 business. If you want it, just find me on Instagram and message me YouTube stack, and I'll send it right over to you.
And if you wanna see why I'm going all in on AI this year, click here, and I'll see you on the other side.
The Hook
The bait, then the rug-pull.
Every framework for AI productivity starts with the tools. This one starts with a different question: which tasks should never reach you at all? The GAIN Matrix reframes the AI race — not as a prompting competition but as a sorting problem — and the answer to that sorting problem is what separates the buried from the free.
Named ideas worth stealing.
01:23model
GAIN Matrix
- Give
- Accelerate
- Integrate
- No AI
A 2x2 grid mapping tasks by ease for humans (vertical) vs ease for computers (horizontal). Tells you exactly where every task on your plate belongs.
Steal forDelegation conversations, team AI onboarding, personal productivity audit
06:39model
10-80-10 Rule
10% human ideation sets the brief, AI handles 80% of the work, human returns for the final 10% quality pass. Prevents both over-delegation and under-delegation.
Steal forOnboarding clients or team members to AI-assisted workflows
03:48list
Rule of 3 Rs
- Repetitive
- Rules-based
- Returns time
Three questions to test whether a task should be automated. All three must be yes before you invest in building the workflow.
Steal forSOPs, automation audits, ops reviews
09:45list
Three Checks for Integrate
- Does it require taste?
- Does it require vision?
- Does it require care?
If yes to any one, the task belongs in the Integrate quadrant — not fully delegated, not fully owned. Human is the director, AI is the executor.
Steal forCreative briefs, design reviews, strategic planning sessions
How they asked for the click.
VERBAL ASK
04:09product
“DM me the word YouTube stack on Instagram and I'll send it right over to you”
Mentioned twice — once mid-video at 04:09 and once at the end at 14:24. Low friction, no opt-in page. Converts interest generated by the framework into an Instagram DM lead.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
AFFILIATECommission earned if you click.
OTHER LINKSAlso linked in the description.
Visual structure at a glance.
