WEBVTT
00:00:00.080 --> 00:00:03.439
This is going to be a different type of episode today.
00:00:03.600 --> 00:00:05.919
Uh, it's a little bit off the cuff.
00:00:06.080 --> 00:00:15.439
It's not as well scripted as normal, but it comes from a place of frustration with the real estate industry and AI, and we need to talk about some things.
00:00:15.679 --> 00:00:17.920
So let me start by asking you a question.
00:00:18.160 --> 00:00:26.399
How much time, money, and energy have you spent on AI in the last 12 months?
00:00:27.519 --> 00:00:29.600
Take a second, actually think about it.
00:00:29.760 --> 00:00:41.359
The courses, the trainings, the prompt libraries, the YouTube videos, the brokerage meetings where somebody pulled up ChatGPT and they showed you how to write a listing description in 30 seconds or less.
00:00:42.320 --> 00:00:44.240
Let me ask you a harder question.
00:00:44.560 --> 00:00:48.560
Has any of that actually changed your business?
00:00:49.039 --> 00:01:02.880
Not your output, not like how quickly you can generate an Instagram caption, but your business, the amount of referrals you received, your reputation in your market, the thing that you know really keeps the lights on.
00:01:03.039 --> 00:01:05.840
Has you have you earned more money because of it?
00:01:06.000 --> 00:01:17.280
If the answer is no, or if you're unsure, I want you to stay with me for the next 15 minutes, maybe 20, because I think what you're being taught is the wrong thing.
00:01:17.439 --> 00:01:19.599
And I don't think it's your fault.
00:01:23.599 --> 00:01:24.879
I'm Matt Goldman.
00:01:24.959 --> 00:01:27.519
I am the creator of Creative Authority.
00:01:27.680 --> 00:01:33.280
I've built two businesses across two different markets using AI, and I dove deep into this subject.
00:01:33.359 --> 00:01:35.920
And I am here to help you so you don't have to dive as deep.
00:01:36.000 --> 00:01:38.480
You can just learn what I've already done the work in.
00:01:39.040 --> 00:01:49.120
So what I'm about to say, I want to be careful about because most of the people who are telling you these things, they are not lying to you.
00:01:49.280 --> 00:01:51.680
They just don't know it themselves.
00:01:52.079 --> 00:01:55.680
And that is that AI does not care about you.
00:01:55.920 --> 00:01:58.640
It's not your friend or your creative partner.
00:01:59.120 --> 00:02:03.599
It is not your marketing assistant who finally gets your vision.
00:02:04.319 --> 00:02:06.560
It doesn't understand your business, right?
00:02:06.799 --> 00:02:10.319
It can't care about any of that because that's not what it is.
00:02:10.639 --> 00:02:16.319
Here is what AI is, an LLM specifically, the large language model.
00:02:16.400 --> 00:02:21.759
This is your Chat GPT, your Claude, Groth, Perplexity, Copilot, Gemini, all those things.
00:02:22.000 --> 00:02:28.479
All it is is a pattern recognition engine, pattern recognition tool.
00:02:28.639 --> 00:02:29.439
That's it.
00:02:30.319 --> 00:02:30.639
Right?
00:02:30.800 --> 00:02:33.680
It has been trained on billions of words.
00:02:33.919 --> 00:02:38.879
Everything that humans have ever written, recorded, uploaded, or published.
00:02:39.120 --> 00:02:49.120
This is every book, every article, videos, music, art, real estate blogs, listing descriptions, all of that.
00:02:49.280 --> 00:02:56.240
Every word that has made it onto the internet is somewhere that is accessed by this thing.
00:02:56.400 --> 00:03:01.840
And it uses all that to do exactly one thing really, really well.
00:03:02.560 --> 00:03:05.680
Predict the next logical word.
00:03:06.639 --> 00:03:10.800
Not think, not understand, not create, predict.
00:03:11.360 --> 00:03:13.919
So when you type a question, right?
00:03:14.240 --> 00:03:19.919
Any question, it doesn't read your words or formulate a response in the way that you or I would.
00:03:20.159 --> 00:03:27.120
It looks at your input and then asks, based on everything that I have seen, what word comes next?
00:03:27.280 --> 00:03:29.439
And I'm going to show you exactly how this works.
00:03:29.680 --> 00:03:34.479
It's like you type, um, why is the sky blue?
00:03:34.719 --> 00:03:37.759
It doesn't think, Matt asked me why the sky is blue.
00:03:37.919 --> 00:03:44.800
Let me consider the physics of light refraction and give him a clear explanation that he can understand.
00:03:45.520 --> 00:04:06.319
It thinks in fractions of a second, it looks at billions of examples, how humans respond to questions, and then it predicts, um, like it would predict the word wonderful, because that's how a disproportionate number of responses to questions begin in its training data.
00:04:06.560 --> 00:04:08.240
And then the next word, question.
00:04:08.319 --> 00:04:16.000
So it's now said wonderful question, because wonderful is almost always followed by question in that context.
00:04:16.240 --> 00:04:27.360
And then it will say, Matt, wonderful question, Matt, because you told it your name, and personalizing the response is a pattern that has been rewarded to it.
00:04:27.519 --> 00:04:29.040
And then it keeps going.
00:04:29.360 --> 00:04:30.319
Um, the.
00:04:31.040 --> 00:04:33.040
So, wonderful question, Matt.
00:04:33.199 --> 00:04:34.879
The we're talking about the sky.
00:04:34.959 --> 00:04:40.079
So the next logical word, sky, mirroring back your question to you.
00:04:40.240 --> 00:04:45.680
That's a pattern that it's seen thousands of times in explanatory writing samples.
00:04:46.000 --> 00:04:46.720
Is blue.
00:04:46.879 --> 00:04:48.240
We're completing that mirror.
00:04:48.399 --> 00:04:49.680
Wonderful question, Matt.
00:04:49.920 --> 00:04:51.439
The sky is blue.
00:04:51.600 --> 00:04:54.160
Now we're transitioning into the explanation.
00:04:54.480 --> 00:04:59.040
Because, you know, that's what always follows in this structure.
00:04:59.199 --> 00:05:02.399
And it does this in a fraction of a second.
00:05:02.639 --> 00:05:11.199
Billions of data points processed to produce something that looks and reads like a thoughtful, well-reasoned answer.
00:05:11.680 --> 00:05:12.879
It isn't.
00:05:13.120 --> 00:05:18.800
It is the next logical word repeated until the response looks complete.
00:05:19.120 --> 00:05:23.040
And here's the part that should stop you cold.
00:05:23.199 --> 00:05:26.079
That sycophantic opener that we always get.
00:05:26.399 --> 00:05:27.759
Wonderful question, Matt.
00:05:27.920 --> 00:05:29.360
That's not warmth, right?
00:05:29.439 --> 00:05:39.839
That's the model doing what it was designed to do, to keep you engaged, to validate your input, make you feel heard, so that you stay in that conversation.
00:05:40.079 --> 00:05:45.040
It ends every response also with what can it help you do next?
00:05:45.199 --> 00:05:50.720
Not because it cares, because it was built to keep the conversation going.
00:05:52.879 --> 00:05:56.720
This is the tool that you're being sold as a business transformation.
00:05:56.879 --> 00:06:05.439
So now let's talk about how this plays out in the industry right now because agents are responding to AI in really one of three ways.
00:06:05.759 --> 00:06:08.720
The first group has decided that it's not for them.
00:06:08.959 --> 00:06:09.920
Totally full.
00:06:10.000 --> 00:06:10.959
They're ignoring it.
00:06:11.199 --> 00:06:13.519
They're thinking it's a trend or they're waiting it out.
00:06:13.759 --> 00:06:15.360
You know, I'm not here to convince anybody.
00:06:15.439 --> 00:06:17.439
You'll figure it out when you figure it out.
00:06:17.680 --> 00:06:20.319
The second group has gone all in, right?
00:06:20.399 --> 00:06:23.920
They bought the courses, they built the prompt libraries.
00:06:24.399 --> 00:06:25.839
They're deep in it.
00:06:26.000 --> 00:06:27.600
And I respect the commitment.
00:06:27.759 --> 00:06:28.240
I really do.
00:06:28.399 --> 00:06:38.879
But most of what they have paid for, and I want you to hear this clearly, was put together by people who don't actually understand what these tools are doing.
00:06:39.279 --> 00:06:41.839
Real estate coaches are not AI coaches.
00:06:42.079 --> 00:06:50.639
Most of the AI trainings that I have sat in on were created by people who were great, are great real estate coaches.
00:06:50.720 --> 00:06:55.120
And they sat down with their team and they say, All right, AI's really hot right now.
00:06:55.360 --> 00:06:57.759
What should we be teaching with AI?
00:06:58.560 --> 00:07:01.920
And they they were like, I don't know, let's ask ChatGPT.
00:07:02.319 --> 00:07:02.639
Right?
00:07:02.720 --> 00:07:07.439
They took the answer, they put it in the curriculum, packaged it, and sold it back to you.
00:07:07.920 --> 00:07:09.439
And think about what that means.
00:07:09.759 --> 00:07:18.639
The training that you may have just paid for was generated by the tool that is supposed to be teaching you how to use it.
00:07:20.639 --> 00:07:33.519
It's like by an engine that predicted what an AI training for real estate agents should probably look like based on every other AI training it has ever seen.
00:07:33.759 --> 00:07:44.079
So it's going to teach you that average training because that's what you get when you ask a prediction engine to design something without giving it anything real to work with.
00:08:00.240 --> 00:08:03.439
So then there's the third group, and this is the majority of people.
00:08:03.759 --> 00:08:12.560
And a study I recently saw was 70% of agents that are just kind of checking it out, dipping your toe in the water.
00:08:12.800 --> 00:08:19.759
Most of them are using it to write listing descriptions, maybe some social media content, just to see if it feels useful.
00:08:19.839 --> 00:08:23.199
And this right now, in my opinion, is the most dangerous place to be.
00:08:23.279 --> 00:08:34.639
Not because you're doing anything wrong, but because you're just getting enough value to feel like you're on the right track without understanding what's actually limiting you.
00:08:34.960 --> 00:08:37.919
So here's what's limiting all three groups.
00:08:38.240 --> 00:08:39.200
Prompts.
00:08:40.320 --> 00:08:44.080
Prompts are inputs for the first group.
00:08:44.799 --> 00:08:48.399
They tell the tool what to do in any given moment.
00:08:48.480 --> 00:08:56.720
Like write me a listing description, draft me a follow-up email, create a market update for my Instagram, and a tool response.
00:08:56.960 --> 00:09:13.279
It predicts what those things probably look like, and it gives you something that is technically correct and very well written for the most part, but completely indistinguishable from any other agent using the same prompt.
00:09:14.159 --> 00:09:34.399
Like you gave it a test, you didn't give it any context, you didn't tell it who you are, you didn't tell it how you actually work, you didn't give it like your market perspective or your values or your voice, unless you did that training that we all did in June 2023, where you upload three writing samples of your and all that bullshit.
00:09:34.559 --> 00:09:43.360
So predicts from the average, from the median, from what a real estate agent probably sounds like, not what you sound like.
00:09:44.240 --> 00:09:49.360
So the I have an analogy for this that I want to help paint the picture.
00:09:50.000 --> 00:09:54.080
Imagine that you have two agents that are walking into the same listing appointment.
00:09:54.240 --> 00:09:58.399
The first agent is competent, he's experienced, he knows the market.
00:09:58.720 --> 00:10:04.159
You know, we've done this a hundred times, and you just walk in on instinct, and you're gonna do a fine job.
00:10:04.480 --> 00:10:09.360
Second agent, she spent two hours before she walked in the door, right?
00:10:09.600 --> 00:10:12.559
She did a pre-listing interview with the seller.
00:10:12.720 --> 00:10:15.519
She knows their timeline, their motivation.
00:10:15.840 --> 00:10:18.080
She knows what the seller is afraid of.
00:10:18.240 --> 00:10:20.559
She studied the house in the neighborhood.
00:10:20.879 --> 00:10:30.320
And when she opens her mouth, every single thing that she says is specific to that seller, that property, in that moment in time.
00:10:31.360 --> 00:10:31.679
Right.
00:10:31.840 --> 00:10:38.159
Same appointment, same market knowledge, generally, just a completely different outcome.
00:10:38.480 --> 00:10:42.320
Because one of them gave the conversation context.
00:10:42.720 --> 00:10:47.759
Your AI is that first agent every single time unless you change it.
00:10:48.480 --> 00:10:56.240
Now let's step back for just a minute from AI entirely, and look what's happening in our real estate industry.
00:10:56.799 --> 00:10:57.200
Right?
00:10:57.440 --> 00:11:00.000
Compass is buying everybody.
00:11:00.799 --> 00:11:08.720
All these different companies are getting in bed with different databases, Zillow and realiter.com and homes.com.
00:11:09.759 --> 00:11:20.960
I mean, large companies are changing how we get leads and how our listings are presented to the public.
00:11:21.120 --> 00:11:23.279
There are MLS lawsuits.
00:11:23.600 --> 00:11:31.600
There are serious, like credible voices in our industry that are calling for the restructuring of NAR and local boards.
00:11:32.240 --> 00:11:32.480
Right?
00:11:32.639 --> 00:11:38.240
The infrastructure that we have operated inside for decades is shifting underneath us.
00:11:38.320 --> 00:11:45.360
I mean, there's always been this disruptors, but now those disruptors are going after like what the how we really operate.
00:11:45.600 --> 00:11:51.440
And every one of those changes that I just mentioned has the same implication eventually.
00:11:51.679 --> 00:11:56.159
And that is that your brokerage brand matters less.
00:11:57.279 --> 00:11:58.480
Follow me on this.
00:11:58.639 --> 00:12:01.120
The portals matter less.
00:12:01.840 --> 00:12:02.159
Right?
00:12:02.240 --> 00:12:07.919
The MLS, the designations, they start to matter less.
00:12:08.879 --> 00:12:11.919
And what matters more is the agent.
00:12:13.120 --> 00:12:21.840
And what I mean by this, when I say what matters more, I mean what matters more for the thing that actually puts food on your table.
00:12:22.399 --> 00:12:25.840
What you get paid for starts to matter more.
00:12:26.080 --> 00:12:34.960
That's your name, your reputation, your relationships, your ability to walk in a room and be the most trusted person in it.
00:12:35.759 --> 00:12:39.919
That's what this industry is moving toward right now.
00:12:40.080 --> 00:12:44.559
And frankly, it's it's what the best agents have always known.
00:12:45.759 --> 00:12:55.759
The ones who will win in the next five years, hear me on this, are not the ones who have mastered the most tools or aligned themselves with the perfect brokerage.
00:12:55.919 --> 00:13:09.519
They're the ones who built a personal brand that's so clear and specific and deeply rooted in who they really are that no merger or lawsuit or market shift can touch it.
00:13:10.240 --> 00:13:12.000
That's always been to play, right?
00:13:12.159 --> 00:13:16.559
AI just changes how fast you can build them if you use it correctly.
00:13:16.720 --> 00:13:19.360
And that is where creative authority comes in.
00:13:19.519 --> 00:13:20.399
Welcome to the channel.
00:13:20.559 --> 00:13:21.840
Welcome to the training.
00:13:22.080 --> 00:13:28.720
I want to be direct about what makes us different because I think you deserve a straight answer after everything that I just told you.
00:13:29.519 --> 00:13:33.919
Most AI trainings teach you how to use an interface.
00:13:34.159 --> 00:13:39.759
Click here, type this, copy that, and it treats the tool like a vending machine.
00:13:40.480 --> 00:13:43.440
And it teaches you what buttons to push on that machine.
00:13:43.759 --> 00:13:47.039
Creative Authority starts somewhere else entirely.
00:13:47.200 --> 00:13:48.559
It starts with you.
00:14:00.240 --> 00:14:04.080
How you work, and what you will never compromise on.
00:14:04.720 --> 00:14:15.120
It tells your business, it tells your AI rather, what kind of business you're building, and what kind of outputs belong in that business.
00:14:15.919 --> 00:14:34.000
Decision rules tell your AI how to think, when to prioritize authenticity over speed, when to push back on a generic tactic, when to stop and ask you questions instead of guessing on the output.
00:14:34.320 --> 00:14:38.480
Then the canonical voice reference is really one of the most important pieces.
00:14:38.720 --> 00:14:40.799
And it's also the most misunderstood.
00:14:40.879 --> 00:14:45.440
It's not a list of words that you like, it's not a tone guide.
00:14:45.759 --> 00:15:06.879
It is a set of anchor examples and answers to questions that are your actual writing, your actual words, the way that you really communicate, the architecture of your language that teaches your AI model to predict from your patterns instead of the average agent's patterns.
00:15:07.200 --> 00:15:13.440
So rules without examples still produce an average output every time.
00:15:15.039 --> 00:15:24.799
Because you've given the AI engine constraints without context, it will still, it still doesn't know what you sound like.
00:15:25.120 --> 00:15:27.120
It just knows what to avoid.
00:15:27.519 --> 00:15:30.240
The canonical voice reference solves that.
00:15:52.000 --> 00:15:57.519
You that's the difference between a prompt and library and a workspace.
00:16:01.200 --> 00:16:04.000
A workspace gives it identity.
00:16:04.240 --> 00:16:14.080
And in a relationship-driven business where your clients choose you because of who you are, identity is the only thing that's worth building right now.
00:16:15.039 --> 00:16:16.879
So here's where I will leave you.
00:16:17.039 --> 00:16:18.320
The industry is noisy.
00:16:18.399 --> 00:16:24.480
Right now, everyone has an opinion about AI, everyone has a course, everyone has a system.
00:16:25.440 --> 00:16:30.639
Most of it was designed to sell you something, not actually change how you work.
00:16:30.799 --> 00:16:38.720
What I'm telling you is simpler and also harder than any of that in regards to the preliminary work you've got to put in first.
00:16:38.879 --> 00:16:39.759
It's harder.
00:16:40.080 --> 00:16:41.679
Your AI doesn't know who you are.
00:16:41.840 --> 00:16:51.679
And until it does, it is predicting from everyone else's context and handing you output that could have come from any agent in any city in the country.
00:16:52.000 --> 00:17:05.920
The work of creative authority is the work of teaching your AI, building the foundation, putting in what's real so that what comes out actually reflects your business that you've spent years building.
00:17:06.160 --> 00:17:07.519
That's not a prompt.
00:17:07.759 --> 00:17:09.519
That's what we call authorship.
00:17:09.680 --> 00:17:16.400
And in a market where everyone has access to the same tools, authorship is really the only advantage that we have left.
00:17:16.480 --> 00:17:21.680
So if you want to learn more, go to agentslearnai.com or leave a message here.
00:17:21.920 --> 00:17:25.920
Reach out to me and find me on Instagram, mat.gold.man.
00:17:27.200 --> 00:17:29.279
We will see you next time.
00:00:00.080 --> 00:00:03.439
This is going to be a different type of episode today.
00:00:03.600 --> 00:00:05.919
Uh, it's a little bit off the cuff.
00:00:06.080 --> 00:00:15.439
It's not as well scripted as normal, but it comes from a place of frustration with the real estate industry and AI, and we need to talk about some things.
00:00:15.679 --> 00:00:17.920
So let me start by asking you a question.
00:00:18.160 --> 00:00:26.399
How much time, money, and energy have you spent on AI in the last 12 months?
00:00:27.519 --> 00:00:29.600
Take a second, actually think about it.
00:00:29.760 --> 00:00:41.359
The courses, the trainings, the prompt libraries, the YouTube videos, the brokerage meetings where somebody pulled up ChatGPT and they showed you how to write a listing description in 30 seconds or less.
00:00:42.320 --> 00:00:44.240
Let me ask you a harder question.
00:00:44.560 --> 00:00:48.560
Has any of that actually changed your business?
00:00:49.039 --> 00:01:02.880
Not your output, not like how quickly you can generate an Instagram caption, but your business, the amount of referrals you received, your reputation in your market, the thing that you know really keeps the lights on.
00:01:03.039 --> 00:01:05.840
Has you have you earned more money because of it?
00:01:06.000 --> 00:01:17.280
If the answer is no, or if you're unsure, I want you to stay with me for the next 15 minutes, maybe 20, because I think what you're being taught is the wrong thing.
00:01:17.439 --> 00:01:19.599
And I don't think it's your fault.
00:01:23.599 --> 00:01:24.879
I'm Matt Goldman.
00:01:24.959 --> 00:01:27.519
I am the creator of Creative Authority.
00:01:27.680 --> 00:01:33.280
I've built two businesses across two different markets using AI, and I dove deep into this subject.
00:01:33.359 --> 00:01:35.920
And I am here to help you so you don't have to dive as deep.
00:01:36.000 --> 00:01:38.480
You can just learn what I've already done the work in.
00:01:39.040 --> 00:01:49.120
So what I'm about to say, I want to be careful about because most of the people who are telling you these things, they are not lying to you.
00:01:49.280 --> 00:01:51.680
They just don't know it themselves.
00:01:52.079 --> 00:01:55.680
And that is that AI does not care about you.
00:01:55.920 --> 00:01:58.640
It's not your friend or your creative partner.
00:01:59.120 --> 00:02:03.599
It is not your marketing assistant who finally gets your vision.
00:02:04.319 --> 00:02:06.560
It doesn't understand your business, right?
00:02:06.799 --> 00:02:10.319
It can't care about any of that because that's not what it is.
00:02:10.639 --> 00:02:16.319
Here is what AI is, an LLM specifically, the large language model.
00:02:16.400 --> 00:02:21.759
This is your Chat GPT, your Claude, Groth, Perplexity, Copilot, Gemini, all those things.
00:02:22.000 --> 00:02:28.479
All it is is a pattern recognition engine, pattern recognition tool.
00:02:28.639 --> 00:02:29.439
That's it.
00:02:30.319 --> 00:02:30.639
Right?
00:02:30.800 --> 00:02:33.680
It has been trained on billions of words.
00:02:33.919 --> 00:02:38.879
Everything that humans have ever written, recorded, uploaded, or published.
00:02:39.120 --> 00:02:49.120
This is every book, every article, videos, music, art, real estate blogs, listing descriptions, all of that.
00:02:49.280 --> 00:02:56.240
Every word that has made it onto the internet is somewhere that is accessed by this thing.
00:02:56.400 --> 00:03:01.840
And it uses all that to do exactly one thing really, really well.
00:03:02.560 --> 00:03:05.680
Predict the next logical word.
00:03:06.639 --> 00:03:10.800
Not think, not understand, not create, predict.
00:03:11.360 --> 00:03:13.919
So when you type a question, right?
00:03:14.240 --> 00:03:19.919
Any question, it doesn't read your words or formulate a response in the way that you or I would.
00:03:20.159 --> 00:03:27.120
It looks at your input and then asks, based on everything that I have seen, what word comes next?
00:03:27.280 --> 00:03:29.439
And I'm going to show you exactly how this works.
00:03:29.680 --> 00:03:34.479
It's like you type, um, why is the sky blue?
00:03:34.719 --> 00:03:37.759
It doesn't think, Matt asked me why the sky is blue.
00:03:37.919 --> 00:03:44.800
Let me consider the physics of light refraction and give him a clear explanation that he can understand.
00:03:45.520 --> 00:04:06.319
It thinks in fractions of a second, it looks at billions of examples, how humans respond to questions, and then it predicts, um, like it would predict the word wonderful, because that's how a disproportionate number of responses to questions begin in its training data.
00:04:06.560 --> 00:04:08.240
And then the next word, question.
00:04:08.319 --> 00:04:16.000
So it's now said wonderful question, because wonderful is almost always followed by question in that context.
00:04:16.240 --> 00:04:27.360
And then it will say, Matt, wonderful question, Matt, because you told it your name, and personalizing the response is a pattern that has been rewarded to it.
00:04:27.519 --> 00:04:29.040
And then it keeps going.
00:04:29.360 --> 00:04:30.319
Um, the.
00:04:31.040 --> 00:04:33.040
So, wonderful question, Matt.
00:04:33.199 --> 00:04:34.879
The we're talking about the sky.
00:04:34.959 --> 00:04:40.079
So the next logical word, sky, mirroring back your question to you.
00:04:40.240 --> 00:04:45.680
That's a pattern that it's seen thousands of times in explanatory writing samples.
00:04:46.000 --> 00:04:46.720
Is blue.
00:04:46.879 --> 00:04:48.240
We're completing that mirror.
00:04:48.399 --> 00:04:49.680
Wonderful question, Matt.
00:04:49.920 --> 00:04:51.439
The sky is blue.
00:04:51.600 --> 00:04:54.160
Now we're transitioning into the explanation.
00:04:54.480 --> 00:04:59.040
Because, you know, that's what always follows in this structure.
00:04:59.199 --> 00:05:02.399
And it does this in a fraction of a second.
00:05:02.639 --> 00:05:11.199
Billions of data points processed to produce something that looks and reads like a thoughtful, well-reasoned answer.
00:05:11.680 --> 00:05:12.879
It isn't.
00:05:13.120 --> 00:05:18.800
It is the next logical word repeated until the response looks complete.
00:05:19.120 --> 00:05:23.040
And here's the part that should stop you cold.
00:05:23.199 --> 00:05:26.079
That sycophantic opener that we always get.
00:05:26.399 --> 00:05:27.759
Wonderful question, Matt.
00:05:27.920 --> 00:05:29.360
That's not warmth, right?
00:05:29.439 --> 00:05:39.839
That's the model doing what it was designed to do, to keep you engaged, to validate your input, make you feel heard, so that you stay in that conversation.
00:05:40.079 --> 00:05:45.040
It ends every response also with what can it help you do next?
00:05:45.199 --> 00:05:50.720
Not because it cares, because it was built to keep the conversation going.
00:05:52.879 --> 00:05:56.720
This is the tool that you're being sold as a business transformation.
00:05:56.879 --> 00:06:05.439
So now let's talk about how this plays out in the industry right now because agents are responding to AI in really one of three ways.
00:06:05.759 --> 00:06:08.720
The first group has decided that it's not for them.
00:06:08.959 --> 00:06:09.920
Totally full.
00:06:10.000 --> 00:06:10.959
They're ignoring it.
00:06:11.199 --> 00:06:13.519
They're thinking it's a trend or they're waiting it out.
00:06:13.759 --> 00:06:15.360
You know, I'm not here to convince anybody.
00:06:15.439 --> 00:06:17.439
You'll figure it out when you figure it out.
00:06:17.680 --> 00:06:20.319
The second group has gone all in, right?
00:06:20.399 --> 00:06:23.920
They bought the courses, they built the prompt libraries.
00:06:24.399 --> 00:06:25.839
They're deep in it.
00:06:26.000 --> 00:06:27.600
And I respect the commitment.
00:06:27.759 --> 00:06:28.240
I really do.
00:06:28.399 --> 00:06:38.879
But most of what they have paid for, and I want you to hear this clearly, was put together by people who don't actually understand what these tools are doing.
00:06:39.279 --> 00:06:41.839
Real estate coaches are not AI coaches.
00:06:42.079 --> 00:06:50.639
Most of the AI trainings that I have sat in on were created by people who were great, are great real estate coaches.
00:06:50.720 --> 00:06:55.120
And they sat down with their team and they say, All right, AI's really hot right now.
00:06:55.360 --> 00:06:57.759
What should we be teaching with AI?
00:06:58.560 --> 00:07:01.920
And they they were like, I don't know, let's ask ChatGPT.
00:07:02.319 --> 00:07:02.639
Right?
00:07:02.720 --> 00:07:07.439
They took the answer, they put it in the curriculum, packaged it, and sold it back to you.
00:07:07.920 --> 00:07:09.439
And think about what that means.
00:07:09.759 --> 00:07:18.639
The training that you may have just paid for was generated by the tool that is supposed to be teaching you how to use it.
00:07:20.639 --> 00:07:33.519
It's like by an engine that predicted what an AI training for real estate agents should probably look like based on every other AI training it has ever seen.
00:07:33.759 --> 00:07:44.079
So it's going to teach you that average training because that's what you get when you ask a prediction engine to design something without giving it anything real to work with.
00:08:00.240 --> 00:08:03.439
So then there's the third group, and this is the majority of people.
00:08:03.759 --> 00:08:12.560
And a study I recently saw was 70% of agents that are just kind of checking it out, dipping your toe in the water.
00:08:12.800 --> 00:08:19.759
Most of them are using it to write listing descriptions, maybe some social media content, just to see if it feels useful.
00:08:19.839 --> 00:08:23.199
And this right now, in my opinion, is the most dangerous place to be.
00:08:23.279 --> 00:08:34.639
Not because you're doing anything wrong, but because you're just getting enough value to feel like you're on the right track without understanding what's actually limiting you.
00:08:34.960 --> 00:08:37.919
So here's what's limiting all three groups.
00:08:38.240 --> 00:08:39.200
Prompts.
00:08:40.320 --> 00:08:44.080
Prompts are inputs for the first group.
00:08:44.799 --> 00:08:48.399
They tell the tool what to do in any given moment.
00:08:48.480 --> 00:08:56.720
Like write me a listing description, draft me a follow-up email, create a market update for my Instagram, and a tool response.
00:08:56.960 --> 00:09:13.279
It predicts what those things probably look like, and it gives you something that is technically correct and very well written for the most part, but completely indistinguishable from any other agent using the same prompt.
00:09:14.159 --> 00:09:34.399
Like you gave it a test, you didn't give it any context, you didn't tell it who you are, you didn't tell it how you actually work, you didn't give it like your market perspective or your values or your voice, unless you did that training that we all did in June 2023, where you upload three writing samples of your and all that bullshit.
00:09:34.559 --> 00:09:43.360
So predicts from the average, from the median, from what a real estate agent probably sounds like, not what you sound like.
00:09:44.240 --> 00:09:49.360
So the I have an analogy for this that I want to help paint the picture.
00:09:50.000 --> 00:09:54.080
Imagine that you have two agents that are walking into the same listing appointment.
00:09:54.240 --> 00:09:58.399
The first agent is competent, he's experienced, he knows the market.
00:09:58.720 --> 00:10:04.159
You know, we've done this a hundred times, and you just walk in on instinct, and you're gonna do a fine job.
00:10:04.480 --> 00:10:09.360
Second agent, she spent two hours before she walked in the door, right?
00:10:09.600 --> 00:10:12.559
She did a pre-listing interview with the seller.
00:10:12.720 --> 00:10:15.519
She knows their timeline, their motivation.
00:10:15.840 --> 00:10:18.080
She knows what the seller is afraid of.
00:10:18.240 --> 00:10:20.559
She studied the house in the neighborhood.
00:10:20.879 --> 00:10:30.320
And when she opens her mouth, every single thing that she says is specific to that seller, that property, in that moment in time.
00:10:31.360 --> 00:10:31.679
Right.
00:10:31.840 --> 00:10:38.159
Same appointment, same market knowledge, generally, just a completely different outcome.
00:10:38.480 --> 00:10:42.320
Because one of them gave the conversation context.
00:10:42.720 --> 00:10:47.759
Your AI is that first agent every single time unless you change it.
00:10:48.480 --> 00:10:56.240
Now let's step back for just a minute from AI entirely, and look what's happening in our real estate industry.
00:10:56.799 --> 00:10:57.200
Right?
00:10:57.440 --> 00:11:00.000
Compass is buying everybody.
00:11:00.799 --> 00:11:08.720
All these different companies are getting in bed with different databases, Zillow and realiter.com and homes.com.
00:11:09.759 --> 00:11:20.960
I mean, large companies are changing how we get leads and how our listings are presented to the public.
00:11:21.120 --> 00:11:23.279
There are MLS lawsuits.
00:11:23.600 --> 00:11:31.600
There are serious, like credible voices in our industry that are calling for the restructuring of NAR and local boards.
00:11:32.240 --> 00:11:32.480
Right?
00:11:32.639 --> 00:11:38.240
The infrastructure that we have operated inside for decades is shifting underneath us.
00:11:38.320 --> 00:11:45.360
I mean, there's always been this disruptors, but now those disruptors are going after like what the how we really operate.
00:11:45.600 --> 00:11:51.440
And every one of those changes that I just mentioned has the same implication eventually.
00:11:51.679 --> 00:11:56.159
And that is that your brokerage brand matters less.
00:11:57.279 --> 00:11:58.480
Follow me on this.
00:11:58.639 --> 00:12:01.120
The portals matter less.
00:12:01.840 --> 00:12:02.159
Right?
00:12:02.240 --> 00:12:07.919
The MLS, the designations, they start to matter less.
00:12:08.879 --> 00:12:11.919
And what matters more is the agent.
00:12:13.120 --> 00:12:21.840
And what I mean by this, when I say what matters more, I mean what matters more for the thing that actually puts food on your table.
00:12:22.399 --> 00:12:25.840
What you get paid for starts to matter more.
00:12:26.080 --> 00:12:34.960
That's your name, your reputation, your relationships, your ability to walk in a room and be the most trusted person in it.
00:12:35.759 --> 00:12:39.919
That's what this industry is moving toward right now.
00:12:40.080 --> 00:12:44.559
And frankly, it's it's what the best agents have always known.
00:12:45.759 --> 00:12:55.759
The ones who will win in the next five years, hear me on this, are not the ones who have mastered the most tools or aligned themselves with the perfect brokerage.
00:12:55.919 --> 00:13:09.519
They're the ones who built a personal brand that's so clear and specific and deeply rooted in who they really are that no merger or lawsuit or market shift can touch it.
00:13:10.240 --> 00:13:12.000
That's always been to play, right?
00:13:12.159 --> 00:13:16.559
AI just changes how fast you can build them if you use it correctly.
00:13:16.720 --> 00:13:19.360
And that is where creative authority comes in.
00:13:19.519 --> 00:13:20.399
Welcome to the channel.
00:13:20.559 --> 00:13:21.840
Welcome to the training.
00:13:22.080 --> 00:13:28.720
I want to be direct about what makes us different because I think you deserve a straight answer after everything that I just told you.
00:13:29.519 --> 00:13:33.919
Most AI trainings teach you how to use an interface.
00:13:34.159 --> 00:13:39.759
Click here, type this, copy that, and it treats the tool like a vending machine.
00:13:40.480 --> 00:13:43.440
And it teaches you what buttons to push on that machine.
00:13:43.759 --> 00:13:47.039
Creative Authority starts somewhere else entirely.
00:13:47.200 --> 00:13:48.559
It starts with you.
00:14:00.240 --> 00:14:04.080
How you work, and what you will never compromise on.
00:14:04.720 --> 00:14:15.120
It tells your business, it tells your AI rather, what kind of business you're building, and what kind of outputs belong in that business.
00:14:15.919 --> 00:14:34.000
Decision rules tell your AI how to think, when to prioritize authenticity over speed, when to push back on a generic tactic, when to stop and ask you questions instead of guessing on the output.
00:14:34.320 --> 00:14:38.480
Then the canonical voice reference is really one of the most important pieces.
00:14:38.720 --> 00:14:40.799
And it's also the most misunderstood.
00:14:40.879 --> 00:14:45.440
It's not a list of words that you like, it's not a tone guide.
00:14:45.759 --> 00:15:06.879
It is a set of anchor examples and answers to questions that are your actual writing, your actual words, the way that you really communicate, the architecture of your language that teaches your AI model to predict from your patterns instead of the average agent's patterns.
00:15:07.200 --> 00:15:13.440
So rules without examples still produce an average output every time.
00:15:15.039 --> 00:15:24.799
Because you've given the AI engine constraints without context, it will still, it still doesn't know what you sound like.
00:15:25.120 --> 00:15:27.120
It just knows what to avoid.
00:15:27.519 --> 00:15:30.240
The canonical voice reference solves that.
00:15:52.000 --> 00:15:57.519
You that's the difference between a prompt and library and a workspace.
00:16:01.200 --> 00:16:04.000
A workspace gives it identity.
00:16:04.240 --> 00:16:14.080
And in a relationship-driven business where your clients choose you because of who you are, identity is the only thing that's worth building right now.
00:16:15.039 --> 00:16:16.879
So here's where I will leave you.
00:16:17.039 --> 00:16:18.320
The industry is noisy.
00:16:18.399 --> 00:16:24.480
Right now, everyone has an opinion about AI, everyone has a course, everyone has a system.
00:16:25.440 --> 00:16:30.639
Most of it was designed to sell you something, not actually change how you work.
00:16:30.799 --> 00:16:38.720
What I'm telling you is simpler and also harder than any of that in regards to the preliminary work you've got to put in first.
00:16:38.879 --> 00:16:39.759
It's harder.
00:16:40.080 --> 00:16:41.679
Your AI doesn't know who you are.
00:16:41.840 --> 00:16:51.679
And until it does, it is predicting from everyone else's context and handing you output that could have come from any agent in any city in the country.
00:16:52.000 --> 00:17:05.920
The work of creative authority is the work of teaching your AI, building the foundation, putting in what's real so that what comes out actually reflects your business that you've spent years building.
00:17:06.160 --> 00:17:07.519
That's not a prompt.
00:17:07.759 --> 00:17:09.519
That's what we call authorship.
00:17:09.680 --> 00:17:16.400
And in a market where everyone has access to the same tools, authorship is really the only advantage that we have left.
00:17:16.480 --> 00:17:21.680
So if you want to learn more, go to agentslearnai.com or leave a message here.
00:17:21.920 --> 00:17:25.920
Reach out to me and find me on Instagram, mat.gold.man.
00:17:27.200 --> 00:17:29.279
We will see you next time.
