I’ve been doing this for almost 3 years, made 6 figures, what you should know:
- SMALL businesses will NOT pay you 2k to be their AI guy
- you CAN extract 2-3k value with a bunch of automations
- you want to be aiming for 3-10M/year businesses
- you need clarity with business owner and clear ROI
- they do not care about AI
- some of their employees will be resistant to the change
- do NOT AUTOMATE SALES
- they just need more leads and speed to lead
Ask me more questions. Gurus yapping on podcasts are delusional.
CHRIS CAMILLO JUST LAID OUT HOW TO MAKE $500,000 A YEAR AS AN "AI GUY" FOR LOCAL BUSINESSES
His blueprint:
1. Walk into any HVAC, plumbing, or sprinkler business.
2. Ask where they're leaking money.
3. Build them an AI agent that answers after-hours calls, sends instant texts, and gets quotes out in real time.
4. Integrate it with their CRM for free.
5. Charge them $2K-$3K/month to be their "AI guy."
Repeat across 10-20 businesses.
"There are people right now doing this."
my honest review of the new flying Tesla Roadster
Pros
>from 0 to “I can see my house from here” in under 1s
>Schwarzenegger fits comfortably
>faster than NY cabs
Cons
>parking is now a federal offense
>soul briefly leaves body
>coffee everywhere
10/10
Thx .@elonmusk .@Tesla
hilarious Accenture news!
their employees now only get $29 of AI A MONTH:
Some were spending $1,000+
they said in an internal announcement that from October 1, anything over $29 is billed to client project.
this is the same company that tells YOU to go big on AI.
now you might say Accenture is being cheap. And indeed $29 is cheap for them for what they charge YOU
I believe in people's capabilities and I'd give them at least the $100 plan, either on Codex or Claude, but they're not completely wrong.
when we train 20+ person teams, we set a few rules.
one of them is measuring cost per outcome and relevant workflows, not cost per seat.
$1,000 a month is a bargain if it replaces a hire. we had a few of those situations after we set up AI champions.
and $29 can also be a waste if nobody uses it well.
What does your team spend on AI per person right now? and can they show ROI?
It's one of the things we set up in month 1.
P.S. comment or DM me "plan" and my team will send you our month 1 strategy to put AI to work that we used with 30+ SMB clients.
I argued with my partner today
so now I'm setting up AI workers for her:
Another day where AI surprises me!
I got into a discussion today with my partner because she's working hard to book multiple flights and accommodations for our travels.
context: I live a digital nomad life, traveling often as a speaker for conferences.
BUT today, instead of escalating this into an argument, I decided to spin up some Grok Bots and get them to handle our logistics.
Tracy Travel Planner
Harry Hotel Finder
Larry Logistics Pro
The AI labs keep showing these workflows where AI does travel bookings and flights and normally they never work in my experience
but today I was genuinely surprised!!
because what she did in 5 hours of work, these things did in 5 minutes.
It found the flight she found, it confirmed that it did not work, and then we had multiple options.
She double checked and they made sense, validated work. Usually working with clients I teach what works with AI, not ideas.
The final solution is, of course, to give her this little army of AI helpers that I built, that can help us travel more easily :)
When was the last time that an AI secretary booked a trip or flight for you?
If you're a busy founder or executive, this might be the time to test AI again or give it to your employees to make them 10 times as fast.
P. S. Since last week I’m also doing Grok Bot team trainings, banger product!
AI is starting to SURPRISE me.
After 15,000 hours of prompting:
3 episodes happened this week:
1. I came back from the World AI Summit conference, and one of my Grok bots told me to pick up a cute souvenir for my partner.
This is such a small detail that had such a huge impact because I would love to do this more often, and I always forget it.
It just came out of the blue with this!
It prompted me to remember that. It made such a pleasant impact on my life, so I am actually super grateful.
I sent a follow-up message to the Grok Bot, and it even gave me a heart reaction on it, Such a funky interaction.
I never expected it to reach this point.
2. Two days ago, it reminded me of a client session that I forgot to follow up with.
They sent me multiple availability times. I was traveling, and the email got buried.
It caught it again and told me, "Hey, you still didn't agree on a time yet. What should I send them? I checked your calendar, and this is what you have available."
Amazing. It was actually super beneficial that it caught what I missed.
3. For fun, I was doing the Rubik's Cube. I kept getting stuck at the final part, and I had to restart again.
Claude just created the visualization for me, a video anticipating the next 18 moves and helping me understand my mistake and be successful.
and I actually did it!!
We are not talking about prompting little chatbots anymore.
These are all helpful assistants to my life that are now working to make my life better proactively, and it's huge.
It does help that I set up all the systems.
I have multiple Grok bots checking on my stuff every 2 hours.
I have Claude custom setups and scheduled routines, and I also have Codex to help me with multiple coding tasks.
When they work together to monitor my day and help me live a better life, this is true leverage.
Did you also get these moments of AI delight lately? I'm curious.
@NickADobos Solid! Had a similar realisation that you have to do cool shit at the start of this year, to stay relevant post AI.
That’s how I started https://t.co/2Fft14spQF
New Sunday ritual: 60 seconds of AI learning
From 15 hours of top videos:
My new workflow is finding the most trending videos and then ask Opus to create explainer videos between 1 to 10 minutes.
This week: the 10 popular YouTube videos everyone keeps sharing from Hormozi, Naval, Huberman, Altman, YC and a few others
1 line per lesson this week:
1. Start from a problem you've lived.
2. Build something a few people LOVE.
3. Any weekly task is an automation you can sell.
4. Handle obstacles before you ask.
5. Delete "I think."
6. Ask, don't argue.
7. Open every story with a place.
8. Stuck? Think about fewer things.
9. Play wealth games, not status games.
10. Pay growth first.
11. Test yourself, then sleep on it.
12. Find the 5% that drives 95% of results.
13. Phone in another room. One clear goal.
Let me know the best video or book you consumed this week?
Rather than using 5% of some big honking package like Microsoft Office or Adobe Creative Suite, you'll probably soon just ask your agent for the 5% you actually use. Bespoke software.
World AI Summit: 2 days in Amsterdam, 10,000 people, 300+ speakers.
3 useful things I got:
1. Give your AI less, not more.
Viktor Botev from Iris. ai called it "minimum complete context." The smallest set of documents and rules a task actually needs. Dump your whole SharePoint into a Claude Project and the answers get worse.
I agree
2. No baseline, no AI project.
Marlabs had a simple rule. Before anything gets built, write down from what, to what, by when. Then find an owner who agrees their process will change. Most pilots I see skip at least one of those.
Made this mistake in the past.
3. The AI tools are not the hard part.
Last year the tents were full of people selling you the same agent, but the agent isn't your edge anymore.
How fast your team changes its habits around agents and compliance and orchestration is the market advantage, based on what I noticed working with 15 companies this year.
Are you compliant with AI EU Act? And does your team have reliable AI champions?
Opus 5.5 is AGI and I have proof
The most bizarre AI experiment of my life!
1. so this week I wanted to do the Rubik’s cube for the first time in my life.
YT tutorials were endless, hours
2. So I went to Claude and asked for a video tutorial, got 2 min and 37 seconds
3. I kept messing up step 7 and blew up the whole cube.
4. After I made this mistake 5 times it decided to create ANOTHER explainer video, ANTICIPATE the next 18 moves and give me the solution to follow and how exactly to spin it TO FIX IT.
IT WORKED!!
How many people do you know that can anticipate the rubicks cube moves?
It never made any mistake! I was the one making mistakes.
It’s midnight and I feel so strange while I’m typing this
Wow.
Claude video tutorial attached, I’ll write more about this whole bizarre thing tomorrow
Opus 5.5 is AGI and I have proof
The most bizarre AI experiment of my life!
1. so this week I wanted to do the Rubik’s cube for the first time in my life.
YT tutorials were endless, hours
2. So I went to Claude and asked for a video tutorial, got 2 min and 37 seconds
3. I kept messing up step 7 and blew up the whole cube.
4. After I made this mistake 5 times it decided to create ANOTHER explainer video, ANTICIPATE the next 18 moves and give me the solution to follow and how exactly to spin it TO FIX IT.
IT WORKED!!
How many people do you know that can anticipate the rubicks cube moves?
It never made any mistake! I was the one making mistakes.
It’s midnight and I feel so strange while I’m typing this
Wow.
Claude video tutorial attached, I’ll write more about this whole bizarre thing tomorrow
My goal: put AI to work for 100,000 people
And swim. Today I train 125 people:
Session 1 is 95 developers at a mid market US company rolling out Claude Code Enterprise.
The tool works fine but the danger is the bill imo with the tokenmaxxing hype BS
A team that size can burn $100k a month in tokens and have little to show for it.
So we're focused on two numbers:
1. Token spend that matches what actually ships.
2. Three use cases that make the company money or give people hours back.
Session 2 is 30 people learning the new Grok Bot.
These are agents with their own computer. They do the work, talk to each other, and keep going until the job is done.
The tech is very capable BUT
From what I’ve noticed, working with 50+ companies in the last 3 years, is that people's habits aren't ready yet… to orchestrate ai teammates reliably.
They still use AI like a search box. Ask, read, copy, paste. Working with agents is closer to managing a team, so today we start with the 3 habits that make that switch.
I spent my morning testing the workflows that worked for me and my clients and cutting anything confusing.
Also because I'm training humans, not AI and buying licences is easy. Changing how people work on a Monday is my actual job. Kinda like an AI productivity coach.
Tomorrow I'll share the 3 habits and what worked in both rooms, if you want to follow.
AND I’m working hard this week to … learn how to swim! You got tips? :)
If your team is rolling out Claude Code or using Grok Bots, what part worries you most or where do you need help?
Putting AI to work for 100,000 smart people
today we go to Amsterdam, AI World Summit
Wednesday and Thursday
10,000+ people, 300+ speakers, and a theme that leans hard into AI policy.
So expecting a lot of EU regulation talk… a meme at this point BUT!
Yesterday I spent the day with almost 100 people helping them put AI to work. 2 workshops for 2 companies.
the next 2 days will be full of people debating how much of that work with AI is actually allowed.
curious about what will Gary Marcus and Yann Lecun say at the conference
Mostly I'll be looking for the teams that are actually shipping.
If you're there, let's grab an espresso and laugh a little about EU AI regulation.
Are you going? Comment or DM me.
My goal: put AI to work for 100,000 people
And swim. Today I train 125 people:
Session 1 is 95 developers at a mid market US company rolling out Claude Code Enterprise.
The tool works fine but the danger is the bill imo with the tokenmaxxing hype BS
A team that size can burn $100k a month in tokens and have little to show for it.
So we're focused on two numbers:
1. Token spend that matches what actually ships.
2. Three use cases that make the company money or give people hours back.
Session 2 is 30 people learning the new Grok Bot.
These are agents with their own computer. They do the work, talk to each other, and keep going until the job is done.
The tech is very capable BUT
From what I’ve noticed, working with 50+ companies in the last 3 years, is that people's habits aren't ready yet… to orchestrate ai teammates reliably.
They still use AI like a search box. Ask, read, copy, paste. Working with agents is closer to managing a team, so today we start with the 3 habits that make that switch.
I spent my morning testing the workflows that worked for me and my clients and cutting anything confusing.
Also because I'm training humans, not AI and buying licences is easy. Changing how people work on a Monday is my actual job. Kinda like an AI productivity coach.
Tomorrow I'll share the 3 habits and what worked in both rooms, if you want to follow.
AND I’m working hard this week to … learn how to swim! You got tips? :)
If your team is rolling out Claude Code or using Grok Bots, what part worries you most or where do you need help?
@iamkylebalmer It’s interesting because it’s one of my worst fears. Ultimately I want to do it because the seaside is gorgeous and I wanna go swimming more :)