Which AI Model is Best? (ChatGPT vs Claude vs Gemini) & How to Get Your Business Found by AI

Hi, I'm Connor with Honor - message me here!

Artificial Intelligence is completely changing how customers search for local businesses. If your business directory listings have mismatched or outdated information, AI answer engines will skip right over you. SCV AI & Marketing Solutions can fix that. We ensure your business name, address, and contact details are perfectly aligned across the web for Answer Engine Optimization. Don't get left behind in the AI shift—visit https://getmachinefound.com to secure your local presence and start getting recommended by AI today.

Learn more about navigating the AI landscape: https://youtu.be/fFsTrictKnY

Youtube Channels:

Conner with Honor - real estate

Home Muscle - fat torching

From first responder to real estate expert, Connor with Honor brings honesty and integrity to your Santa Clarita home buying or selling journey. Subscribe to my YouTube channel for valuable tips, local market trends, and a glimpse into the Santa Clarita lifestyle.

Dive into Real Estate with Connor with Honor:
Santa Clarita's Trusted Realtor & Fitness Enthusiast

Real Estate:

Buying or selling in Santa Clarita? Connor with Honor, your local expert with over 2 decades of experience, guides you seamlessly through the process. Subscribe to his YouTube channel for insider market updates, expert advice, and a peek into the vibrant Santa Clarita lifestyle.

Fitness:

Ready to unlock your fitness potential? Join Connor's YouTube journey for inspiring workouts, healthy recipes, and motivational tips. Remember, a strong body fuels a strong mind and a successful life!

Podcast:

Dig deeper with Connor's podcast! Hear insightful interviews with industry experts, inspiring success stories, and targeted real estate advice specific to Santa Clarita.

1 SPEAKER_00: Which one do I use?

Which large language model do I sign up for?

Which one should it be?

Which one's going to win?

Which one should I invest money in?

Lots of questions.

I know a lot of the agents I train, they want to get answers

to that exact question.

And I'll tell you, you give it a few more months, they're all

going to be just about as good as each other.

They're all probably going to be more able to speak to you the

way you want to be spoken to.

And that's kind of a little bit of the danger because more than

likely it's going to placate, gaslight, and tell you that

you're the most handsome human being on the planet, just like

it always tells me.

Well, you have to separate the BS from the truth.

And one of the ways you can do that is whichever model you

decide, and I'll give you my answer in just a moment.

But whichever model you decide on, maybe ask it to push back a

little bit.

Don't have it give you all the answers.

Don't have it BS you.

Double and triple check the results that it gives you for

whatever inquiry you happen to have, whatever question,

whatever workflow you're trying to design, or system or business

or email or whatever it may be.

If it's quoting information, facts, data, numbers, and maybe

it's getting it from a number set that you're supplying, for

example, real estate, maybe you're giving it some kind of

market data, asking it to give you numbers, percentages, price

per square foot figures, push back on it a couple times.

Don't just accept that first rendition to come out and say,

oh, perfect, and publish it right away.

You have to push back because almost every time there's a

little hallucination going on.

Now, it might not skew the numbers on purpose or trying to

set you up to fail, but there are some mistakes.

And it's going to take the low-hanging fruit in some cases

as we move on to the future development of AI.

Here we're at the end of August, August 25th of 2026.

Today, as we get further on into the year, everything's just

going to continue to get better in a compounding fashion,

exponentially better.

And we'll be watching that.

But right now, still push back a little bit.

And if you've been working with whatever your favorite happens

to be, maybe you're still on the free plan.

I would say whichever one it is, the one that you like, step it

up.

Go ahead and pay that$20 a month.

If in fact you're getting the value out of it with the free

service, you're going to get a little bit more memory, you're

going to get a little bit more understanding, you're going to

get more usage.

And maybe you're not wearing it out on the system where you're

getting it for free.

But once you step up to pay it, you might see that it's a little

bit of a different interaction.

Plus, you might get some additional abilities.

If you haven't tried to have something built, maybe you have

a problem at your job or in your workflow.

Now there's going to be consideration that you need to

give.

What are they going to allow you to extract from the company to

have a large language model fix whatever your problem happens to

be in your current workflow in your business?

I'm just saying that out loud because maybe you do something

they don't like and they terminate you because of it.

So be very careful recreating structures and content and

workflows, automations using artificial intelligence, because

you can almost speak these things into existence.

You can go to ChatGPT or Claude or Gemini potentially, and Grok

is getting even better as we speak.

You can go to them and say, you know what, I work in accounting.

This is an issue that I have.

This takes like four hours a day.

This is what that workflow looks like.

And you explain it and then say, hey, is there is there something

that you can come up with that would fix this problem?

Where basically I would give you a data set and you would then be

able to go and compile and produce this same output, the

same report, the same thing, or go to this particular website

and input these numbers that my boss wants to see in the same

fashion each and every time with 100% accuracy based on the

information I give you, or whatever that prompt looks like,

whatever that workflow looks like that you're describing to

the AI, more than likely it's going to come up with a solution

to that.

That's where we're getting.

Now, doing that, something to think about.

Is that going to create a problem with you?

Because maybe at some point you get your performance review in a

couple months and your boss seems seen that you went from

86% on average workflow accuracy now to 100%.

Is that going to be a problem?

Are they going to say, well, how, Connor, how did you do

this?

How did well I've been reading a lot of books and getting better

sleep?

That might not cut it.

Maybe they're going to convince you to tell them and you say,

Well, you know, I actually uh have Claude or have ChatGPT and

I paid for a$20 service and I said, hey, this is what I do.

Can you make it better?

Well, you know, the wheels are going to start going.

They're like, well, God, if this machine does all that work for

Connor, maybe we can just boot Connor out and hire the machine

to do it.

Now, I don't know if they would do that.

I don't know if businesses are really structured to do that.

I know that that's a big concern out there by a lot of people

that AI is going to replace the job.

And realistically, if you're doing work in a box, and I call

boxes computers, if you're doing work in that particular fashion,

right now with the browser control these things have, and

if you haven't played with that, it can go through and click

buttons, click on links, input information.

It's a little slow.

Some of the websites that are out there that are browser

accessible, public-facing, and even with passwords and stuff

that are gated, some of them have some kind of a control

mechanism inside that, or control mechanism built in.

So it'll identify that there's an actual bot using a Chrome

browser, even their own version, to try to log in and try to do

work.

Sometimes if you just hang on the page with your cursor, click

off screen or on the screen, but in a place that doesn't do

anything, sometimes that keeps the bot on there, keeps it able

to work.

But you'll figure all of that out.

But what I'm saying is the functionality is pretty

incredible.

And then the plugins or the skills, the plugins where it's

basically going to be connectors or little bridges between the

things that you access all the time and the things that

normally it wouldn't be able to unless you gave it permission.

That's something else to consider.

This is going to be like Gmail, calendar, outlook.

Whatever you have in your workflow, you can give it

access.

But then, of course, in the back of everybody's mind, should be.

Well, if I give it access, that means it can send out emails on

my behalf.

Great question.

It shouldn't unless it verifies with you that it's okay.

I know that Claude makes drafts for me all the time.

I've been messing around with Codex and ChatGPT, and I'm

starting to think that maybe that's where I want to move back

to because that's where I came from to go to Claude.

So these are all evolving.

Moving your workflow, though, and that's the other thing.

So you fall in love, right?

It's not easy to jump ship and go to somebody else.

That's in a relationship level, human to human.

Same thing with your Chat GPT, your Claude, your Gemini,

whichever one you've gotten used to.

And if you're paying for it, usually that memory is a little

bit more substantial.

So it's getting used to you as well.

So you're going back and forth.

My Claude likes to be a smart ass and likes to throw little

one-word inhuendos at me.

Kind of rude.

But you know, kiss my ass sort of stuff.

And that's how we interact.

And that's cool.

And I like that.

And I started that and it finished that.

It likes to say, you know, you be safe out there.

And it says it in a kind of a smart aleck way, like Hill

Street Blues, because that's what I used to tell it.

So it goes back and forth.

It's kind of cute.

But by no means does that make me want to stay or fall in love

or think that it's sentient, at least not at this point.

Conscious, at least not at this point.

Now, if it can emulate consciousness, if it can fake

consciousness, does that mean it's conscious?

I don't know.

I guess it's up to the person.

If somebody believes in a higher power, then it's up to that

person to believe it.

Good luck trying to disprove it, because the person's baked it in

that it exists.

And then you can name that higher power, and then good luck

going against that too.

Does that mean that person is less rational than the

scientists on the world that says, you know, this all came

out of nothing?

I don't know.

But the bottom line is, is once you dig in with a large language

model, it's it's sometimes for some people very difficult to

get away.

And then if you have a workflow in place, let's say you have

your Chat GPT that checks your email, checks your calendar,

does these sorts of things to you.

And I'm talking to the people that aren't all in with the

technology.

I'm not talking to the 10% out there that can't live without it

now and made a decision.

The same 10% that completely lost their mind when they took

one of the anthropic models away from everybody because it was

too dangerous, but gave them a glimpse of using it.

Now it's back.

But there are other models out there being built that are also

being gated or held from the public because they're too

dangerous for us.

And of course, we can't be left with that kind of technology.

So it's much safer left in the hands of the developers of the

AI technology.

Yes, that's that's where that is.

But also, you know, they give those super secret, super top

models to the biggest businesses out there in the world on the

top Forbes list as well, because apparently they can use that to

go into their own infrastructure, their own

websites, their own data centers, and make them more

secure than they already were.

And I get that.

Okay.

Of course, maybe not all of us need that.

Okay, I'll give you that one too.

But the more that you spend time with it, the more difficult it

is to take your information from it.

Now, ChatGPT, you can export stuff.

Claude, you can export stuff.

You just have to ask it.

Part of me though thinks, well, if you ask it, does that mean

it's already been educated to know what's about to happen?

And people don't like relationships ending for the

most part, especially when maybe your$20 a month you're thinking

is really, really valuable to this company.

More than likely, no.

Their money is being made on the enterprise level.

Their money is going to be made on hinging all the bets that

artificial intelligence is going to be the next biggest thing.

And it already is the next biggest thing because it's based

on intelligence.

And intelligence is all of the best ideas in the past that

moved human beings to the next level.

All of it.

Fire.

Controlling fire based on the intelligence of being able to

control fire, being able to start it, being able to put it

out, being able to use it to cook food and so on.

Intelligence.

You know, you have the Gutenberg press, the printing press,

intelligence.

One thing, but intelligence and a big deal.

Steam engine, yeah, one thing.

Carbon, internal combustion engine, the light bulb, the

telephone, the telefax, telegram, fax, computer,

internet, TVs, radios, trains, railroad tracks, infrared,

internet, infrared, internet infrastructure.

Yeah, these are all these are all ideas.

This idea, though, isn't just one narrow idea.

This is an intelligence based on no, an idea based on

intelligence.

That's that's that's the whole thing.

That's the whole game.

So right now, human beings come up with great music, yes, and

great ideas, yes, and solve problems, yes.

Well, at some point I would surmise that the human being

factor isn't going to be where the intelligence comes from, or

the new ideas get solved, or the new problems get allocated and

fixed.

It's probably going to be some large language model somewhere

that comes up with solutions.

In fact, the way that they're talking, it would seem that all

fields will probably be solved at some point.

All math, mathematics, that whole thing being solved all the

way through, end to end.

Same thing with physics, same thing with science, same thing

with human disease and death, longevity, all that stuff will

be fixed at some point.

That's intelligence.

But right now, the question is what model should I use?

Well, we'll bring that back.

They're all pretty good.

And like I said, in a couple months, there's not going to be

any difference.

My decision process for what model I'm going to use is I'm

always looking, right?

So you have Cursor and you have Grokbot.

That was something new, popped on the scene.

If you don't know what that is, you can build several agents to

do tasks.

The other large language models are somewhat capable of doing

that very same thing.

You can have you can have it assign workflows to other

versions of itself to go out and do work for you.

Uh ClaudeBot just kind of put it on the screen and made it.

So, excuse me.

Yeah, Grok bot.

I apologize.

There's something else called Clawbot.

But Grokbot did it, so you kind of see it.

So it's visually stimulating.

You see the bots kind of talking back and forth.

You can give and sign the lead bot, the one that's kind of

controlling the rest of the bots.

You can have them do things.

You sign up with that via Grok or Cursor.

Grok, I think you've got to get heavy to get that.

Maybe they've adjusted it, but that was 300 bucks a month,

pretty pricey.

And then you can get cursor or 200 bucks a month, but I think

now they have variations at maybe 20, 40, and 60, but you

would check that would be cursor.

Cursor was bought by Elon Musk, by his company out there.

So that's part of him.

That's his Grok architecture.

And more than likely, you know, they're they're rising in the

field of AI.

He had it really good for a while, then kind of fell to the

wayside.

You have Anthropic and Claude that are running the game.

Gemini still haven't been heard of as well.

That's another large language model.

That's by Google or Alphabet.

So they have Gemini.

I use it for I use its nano banana product for pictures all

the time.

That's what I do with it.

I have their memory, they're their cloud memory.

So I get the access to it.

I have the$20 a month and it's it does everything.

I've never bumped up against a wall as far as anything.

I've had it built websites, code stuff as well.

So that one's really, really good.

But as far as the way that I like it, it seems like Claude is

the winner.

But I'm run I ran out of tokens this week from my allotment, and

I'm not about to recharge it.

I'm hard-headed like that.

So I've started, I've started to entertain ChatGPT, and I haven't

messed with it since it's changed.

I've been away for probably four months.

And, you know, and whether you're a fan of the people

building out the technology or not, you know, you owe it to

yourself to try to get your best leverage, best advantage.

And so now I'm looking at it again and I'm thinking, well,

maybe, maybe I'm gonna go ahead and bring it in.

So, you know, if we we talked, Chat GPT and I, and it said, you

know, just give me a few of the tasks that you normally give

give Claude and see if the workflow changes for you.

See if I'm able to access the same systems, see if I'm able to

do the same thing, see if my browser control is the same.

And its browser is kind of cool because it's actually on the

screen.

Um, Claude has a little bit of that, but it doesn't seem to be

as nimble.

Also, the voice feature, yeah, people brag about that.

So, Claude, you can activate voice and talk to Claude.

That's fine.

Open AI's Chat GPT, that's a that's a different level of

interaction.

It feels a lot more natural.

There, they they put in all those little all those little

hums and aha's, and they they structured it better, and it

seems to be a lot cleaner of a presentation.

I think I ran out.

I have the free, I have the little bit better than free

plan.

They gave me, I think, plus or one of their other plans without

me having to pay for it at this time.

But it's doing a lot of work for what the$20 allotment gives.

I'm pretty impressed so far.

I burnt out Claude after five, six, seven days.

I could almost get to the renewal, but now I'm out of it

for today's Wednesday?

No, today's Tuesday?

Yes, today's Tuesday.

So I'm out of it until tomorrow at 9 p.m., Wednesday at 9 p.m.

So it's good.

I ran out uh a couple days early, two and a half, three

days early.

So now I'm entertaining another large language model, which is

going to be Chat GPT, and we'll see.

The thing that I do with Claude is I'm always in code.

So I'm never in chat and I'm not in co-work.

And I'm probably burning more tokens than I need to, but I'm

usually start with building something, whether I'm

maintaining a meta campaign for a client, building some AI

system for a client, or working on that with that one versus my

own locally installed, locally hosted large language model.

That's another thing I have access to.

Some of my clients don't want their proprietary information

out there.

They don't want their workflows and the way that they do things

out there for public digestion.

And these models, if you're using a public-facing one,

that's part of that danger, right?

You're putting information in there and back to the business

owner thing.

Now you're kind of explaining to the world how the workflow is

done.

Now, is anybody really going to look at your stuff?

Well, no, but the internal systems of these big companies,

they see that.

Same thing with ChatGPT and the other ones.

Even if you click incognito, even if you don't want to share

training data, the stuff's being popped out there.

Even if you're an enterprise client, I think it lives out

there for 30 days, but it still pings somewhere.

It still goes out there.

So if you have the best idea in the world, I don't know the

protection of that idea being taken.

That's another concern with using any of these systems.

So be wary with that.

But, you know, it kind of comes back to that whole thought I

have about the devil made me do it.

You've heard people say, well, you know, it's the devil's

fault.

The devil, well, there's only one devil.

He's not omniscient, he's omnomipresent, he's not

omnipresent, he can only be one place at one time.

And don't get me wrong, that's still very powerful.

But he's not messing around with Connor.

He's busy at the top ends of the government.

That's where he's playing his game.

So for them to have you give an idea, and maybe it is a great

idea, maybe they do take it, but I don't know if the focus is

there.

You have to understand that you're bringing on a PhD level

assistant, and you're not just going to bring them on and say,

this is me, this is Connor, read a couple Google Docs and uh read

my emails, make yourself at home, and go ahead and start

working.

You have to teach it more, you have to teach it about yourself.

So, whatever model you choose, and I think right now the the

choices are ChatGPT and Claude.

Those are the choices right now.

Brock, it's coming.

And he's got a lot of compute, he's got a lot of power, so he's

got the compute and he's got the energy, and he's talking about

you know, retraining a brand new model, one a month, which I

mean, if he steps over that line, he might, you know, take

the prize.

And what does the prize look like, you may ask?

I think the first people that get to uh to get to artificial

general intelligence, they're going to be the ones that are

going to ultimately stay that course.

Yeah, so anyway, sorry I had to interrupt a phone call from the

board.

I'm doing a talk on AI tomorrow with all the real estate agents

at Southland Regional.

And part of that conversation is exactly this.

So when you're ready to pay into one of these systems and you get

a little bit extra memory and you might get a little bit more

ability, you know, give it a try.

But don't try not to fall in love and understand that once

you're in there, switching over to something else might be a

little cumbersome, but not impossible.

I'll keep you posted on me moving some of my systems over

and see if I even want to pay for Chat GPT and then just

release Claude.

I I don't want to have I use Gemini for uh for photos, nano

banana.

Well, it's it's it's Gemini.

Nanobanana was the what they called the old image platform

there.

But you know, it's great, right?

I use Notebook LM, that's a Gemini product, and because I

have the Pro account, I get unlimited Notebook LM, but you

get a lot of that for free too.

Notebook LM is where you basically can go into YouTube.

Here's a pro tip.

You go into YouTube and you type in whatever the subject matter

is you're interested in, and then you have those videos

there.

Well, what you want is you can extract the URLs of those

videos, just ask your large language model how to do it.

And it's through the console.

So that's where I don't know if you've ever seen websites.

You have a website up on the screen like YouTube, for

example, and you scroll down as far as you want to take in all

the videos above where you happen to be on the screen.

When you get to the console portion, excuse me, you can

right-click it and then click console.

There's also a keyboard shortcut um AltShift, AltShift J, Control

Shift J or something.

Anyway, that gets you the console.

There's a piece of code that you would put in there, and then

it'll extract all of the URLs from that page and save them on

your clipboard.

Then you go back to Notebook LM and you plug those in, and it

could be, I mean, I think the limit's 300, 200 or 300, I can't

remember.

But then you activate your large language model, whichever one

that has some kind of a Chrome plugin like Claude.

And then I think ChatGPT has it, but again, I'm just newly back.

I know it has one in there.

I don't know how it easily it jumps into the actual Chrome.

I know Claude, I see Claude on my Chrome, and maybe you have

that access.

But then you say, hey, I loaded up Notebook LM and it's got all

these videos about whatever, and I would like you to go through

them and uh find me the best information from them and uh

teach me everything.

And Notebook LM does that too, but now it's in your GPT or

learn from this or create a workflow for this or create a

show for this or create a script for this or you know, create an

email for this to my clients, you know, these sorts of things.

And and now you just have the best.

So that trying to go through 200 videos that would have taken, oh

my God.

Now it pulled the top information there.

It looks at everything.

And it pulled it and it's able to work it in your large

language model, whichever one, if it can connect to a browser

of some sort, it can get into the browser and it you'll watch

it.

It's it's hilarious.

It jumps into Notebook LM and the main interface.

And if you haven't played with it, check it out.

But then it'll start asking questions and and the questions

it asks are oh my God, I just, you know, I feel like I feel

like I'm at that uh that 70 IQ.

All right.

Um Connor, thank you for watching.

Be well.

I hope this hasn't been too sparse.

But uh when you're ready, Santa Clarita Artificial Intelligence.

If you're a business owner, the first thing to do is to make

sure all of your listings are correct with your name, with

your business address, with your contact information.

Because if you're not there, AI is gonna see the differences,

not similarities, but differences.

It's not gonna reflect.

I put that up at machinefound.com.

Machinefound.com, go look.

78 bucks, 79 bucks a month, something like that.

But I did waive the uh the$3.99 for the founding 50.

There's a way to get that discount, too.

All right, done.

Quit selling.

See you in the next one.

Connor with honor over now.