WEBVTT

00:00:00.000 --> 00:00:03.259
It's January 21st, 2026, and I want you to just

00:00:03.259 --> 00:00:05.820
think back for a second. Think back to last year,

00:00:05.879 --> 00:00:09.439
to 2025. Oh, wow. Do we have to? We have to,

00:00:09.519 --> 00:00:14.679
because honestly, it was a bit of a mess, wasn't

00:00:14.679 --> 00:00:17.399
it? A total mess. It was just this fever dream

00:00:17.399 --> 00:00:19.800
of hype and demos. Right. It felt like every

00:00:19.800 --> 00:00:22.719
week there was some new game -changing tool.

00:00:22.780 --> 00:00:24.579
You'd see it on X, you'd sign up for the waitlist,

00:00:24.660 --> 00:00:27.910
and then you'd get access. It was basically vaporware.

00:00:28.050 --> 00:00:30.449
Or it was just a toy. That was the worst part.

00:00:30.530 --> 00:00:32.450
You play with it for 10 minutes and you realize,

00:00:32.570 --> 00:00:35.170
OK, this can't actually fit into a real professional

00:00:35.170 --> 00:00:37.450
workflow. And you just never touch it again.

00:00:37.549 --> 00:00:39.609
It was the year of the demo. Exactly. The year

00:00:39.609 --> 00:00:42.390
of overpromising. But now, sitting here today.

00:00:43.469 --> 00:00:45.729
It feels like the dust has finally settled. We

00:00:45.729 --> 00:00:47.929
aren't just chatting with bots for fun. We're

00:00:47.929 --> 00:00:50.829
actually watching these complex workflows happen

00:00:50.829 --> 00:00:53.450
in real time. I mean, we're seeing a four -hour

00:00:53.450 --> 00:00:55.789
video shoot get remixed into a full campaign

00:00:55.789 --> 00:00:58.789
blog, social email in like, what, five minutes?

00:00:59.149 --> 00:01:01.509
And that's the headline for 2026. Yes. We've

00:01:01.509 --> 00:01:04.769
moved from just experimenting to real execution.

00:01:05.170 --> 00:01:07.670
So today, we're going to do a deep dive into

00:01:07.670 --> 00:01:10.950
this guide that really claims to be the playbook

00:01:10.950 --> 00:01:14.219
for right now. It's called... The Top Five AI

00:01:14.219 --> 00:01:19.719
Skills for Marketers in 2026 by Max Anne. And

00:01:19.719 --> 00:01:21.400
what I like about it is that it's not sci -fi.

00:01:21.480 --> 00:01:23.319
We're not talking about flying cars here. No,

00:01:23.400 --> 00:01:25.700
this is the baseline. This is, if you're a professional

00:01:25.700 --> 00:01:27.640
marketer today, this is your new job description.

00:01:27.840 --> 00:01:30.439
If you aren't doing this stuff, you are already

00:01:30.439 --> 00:01:32.959
falling behind. Okay, so we've got five key areas

00:01:32.959 --> 00:01:36.000
to roadmap. First is content remixing with Gemini

00:01:36.000 --> 00:01:39.579
3. Then we have visual branding with a tool that

00:01:39.579 --> 00:01:43.540
has my favorite name ever. Nano Banana Pro. I

00:01:43.540 --> 00:01:45.060
still can't get over that. It sounds like something

00:01:45.060 --> 00:01:46.739
you'd put in a smoothie, but it is seriously

00:01:46.739 --> 00:01:48.579
powerful. We'll get to that. Then there's AI

00:01:48.579 --> 00:01:51.560
video with VO and Sora, the rise of AI agents,

00:01:51.640 --> 00:01:54.079
which is a huge mental shift. And finally, this

00:01:54.079 --> 00:01:56.840
thing called vibe coding. Vibe coding. Yeah,

00:01:56.900 --> 00:01:58.599
it sounds like something a teenager does on TikTok,

00:01:58.680 --> 00:02:01.420
but it's actually one of the most practical skills

00:02:01.420 --> 00:02:03.780
on this whole list. Okay, let's jump in with

00:02:03.780 --> 00:02:06.659
number one. Content remixing. The source says

00:02:06.659 --> 00:02:10.479
we've shifted from create once, use once to infinite

00:02:10.479 --> 00:02:12.800
remixing. But we've been talking about repurposing

00:02:12.800 --> 00:02:15.439
for years. What's actually different now? The

00:02:15.439 --> 00:02:17.800
engine is different. The game changer is Gemini

00:02:17.800 --> 00:02:20.620
3. And to get why, you kind of have to look at

00:02:20.620 --> 00:02:24.520
how we did this back in, say, 2024. OK, so take

00:02:24.520 --> 00:02:27.439
us back to the old days. Back then, if you wanted

00:02:27.439 --> 00:02:30.960
an AI to summarize a video, you had to transcribe

00:02:30.960 --> 00:02:33.889
it first. Yeah. You were feeding it text. a script

00:02:33.889 --> 00:02:35.689
right you didn't know if you were whispering

00:02:35.689 --> 00:02:37.629
or shouting it didn't see the slide you were

00:02:37.629 --> 00:02:39.849
pointing to on screen it was basically blind

00:02:39.849 --> 00:02:42.270
and deaf just processing words so it lost all

00:02:42.270 --> 00:02:45.400
the nuance the emotion all of it exactly But

00:02:45.400 --> 00:02:48.180
Gemini 3 has what they call native multimodal

00:02:48.180 --> 00:02:50.099
understanding. It's not reading a transcript.

00:02:50.599 --> 00:02:53.460
It is watching the video file. It's processing

00:02:53.460 --> 00:02:56.060
the raw data, the audio, the pixels, the gestures.

00:02:56.280 --> 00:02:58.319
It understands it like a person does. It gets

00:02:58.319 --> 00:03:00.300
the context. It gets everything. And Max Anne

00:03:00.300 --> 00:03:02.560
gives this amazing workflow example. You have

00:03:02.560 --> 00:03:04.520
a four -hour video recording. Maybe it's a huge

00:03:04.520 --> 00:03:06.759
workshop or a long interview. In the old days,

00:03:06.800 --> 00:03:10.180
that was a week of work. Easily. For sure. Scrubbing

00:03:10.180 --> 00:03:12.259
through timelines, finding clips. Now, you just

00:03:12.259 --> 00:03:15.039
drop the YouTube URL into Gemini 3. No transcription.

00:03:15.300 --> 00:03:17.080
Yeah. And you give it a very specific prompt.

00:03:17.259 --> 00:03:21.539
Watch this. Extract 10 insights, write a 1 ,200

00:03:21.539 --> 00:03:25.199
-word blog, 10 LinkedIn posts, and a three -email

00:03:25.199 --> 00:03:28.439
sequence. And it just does that. In about five

00:03:28.439 --> 00:03:31.539
minutes. Wow. That is a massive compression of

00:03:31.539 --> 00:03:34.659
time. But is the quality any good? We've all

00:03:34.659 --> 00:03:37.900
seen those AI blog posts that sound so robotic.

00:03:38.800 --> 00:03:41.099
Delve into the landscape of... In today's fast

00:03:41.099 --> 00:03:43.419
-paced digital age. Yeah. We all know that sound.

00:03:43.539 --> 00:03:45.439
And that's a fair point. The guide is honest

00:03:45.439 --> 00:03:47.240
about it. It says the output is good because

00:03:47.240 --> 00:03:49.599
it understands the context. But this is a big

00:03:49.599 --> 00:03:51.819
but. It still needs a human. It takes about one

00:03:51.819 --> 00:03:54.500
hour for a person to review, refine, and polish

00:03:54.500 --> 00:03:56.400
everything. It's not a magic button. It's a lever.

00:03:56.599 --> 00:03:59.099
Exactly. But you're comparing one hour of polishing

00:03:59.099 --> 00:04:01.400
to, what, three days of drafting? That's the

00:04:01.400 --> 00:04:04.370
real ROI. It's about speed. The source also brings

00:04:04.370 --> 00:04:06.909
up this idea of a smart stack, saying you shouldn't

00:04:06.909 --> 00:04:08.830
just use Gemini for everything. What's that about?

00:04:08.969 --> 00:04:11.889
This is so important. We all want one tool to

00:04:11.889 --> 00:04:14.289
do it all, but different models just have different

00:04:14.289 --> 00:04:18.050
personalities. Gemini 3 is your analyst. It has

00:04:18.050 --> 00:04:20.170
this huge context window. It can hold that whole

00:04:20.170 --> 00:04:22.470
four -hour video in its head. It's the heavy

00:04:22.470 --> 00:04:25.110
lifter. It does the logic. Okay, so Gemini builds

00:04:25.110 --> 00:04:28.689
the skeleton. Right. But for the polish, for

00:04:28.689 --> 00:04:31.949
the creative flair, The guide says to move the

00:04:31.949 --> 00:04:34.509
text over to Claude. Claude just has a better

00:04:34.509 --> 00:04:38.029
ear for human nuance. It writes less like a corporation

00:04:38.029 --> 00:04:40.470
and more like a person. Interesting. So Gemini

00:04:40.470 --> 00:04:43.250
for thinking, Claude for writing. That's a good

00:04:43.250 --> 00:04:45.370
way to put it. And you keep ChatGPT around for

00:04:45.370 --> 00:04:47.829
just the fast, everyday stuff, quick questions.

00:04:47.930 --> 00:04:50.430
It's not about one perfect tool. It's about stacking

00:04:50.430 --> 00:04:53.089
them. Let me play devil's advocate here. If the

00:04:53.089 --> 00:04:55.730
AI is doing the analysis and the AI is doing

00:04:55.730 --> 00:04:58.740
the writing. What's left for the marketer? Are

00:04:58.740 --> 00:05:00.899
we just pushing buttons? So does this replace

00:05:00.899 --> 00:05:04.000
creativity or just distribution? It replaces

00:05:04.000 --> 00:05:06.180
the drudgery. Creativity moves up to the strategy

00:05:06.180 --> 00:05:08.660
level. Creativity moves up the ladder. I like

00:05:08.660 --> 00:05:10.759
that. Okay, let's move on to the second skill.

00:05:10.860 --> 00:05:13.220
Now we get to talk about Nano Banana Pro. Finally.

00:05:14.199 --> 00:05:16.779
I've been waiting, Nano Banana. Visual branding.

00:05:17.620 --> 00:05:19.360
Remind us why this was such a headache before

00:05:19.360 --> 00:05:22.300
now. Why was visual consistency so hard? Oh,

00:05:22.319 --> 00:05:24.920
it was a nightmare. If you used Midjourney or

00:05:24.920 --> 00:05:29.459
Dali back in 2024, you know this pain. You generate

00:05:29.459 --> 00:05:32.000
a character, let's say a mascot for a coffee

00:05:32.000 --> 00:05:34.740
brand, a bear in a hoodie. Sure, a bear in a

00:05:34.740 --> 00:05:36.930
hoodie, classic. And it looks great. But then

00:05:36.930 --> 00:05:39.670
your next prompt is, OK, now show that same bear

00:05:39.670 --> 00:05:42.310
sitting at a desk. And suddenly it's a different

00:05:42.310 --> 00:05:44.050
bear. The hoodie is a different color. The face

00:05:44.050 --> 00:05:46.629
is wrong. It just morphed every single time.

00:05:46.769 --> 00:05:48.370
You could never build an identity because you

00:05:48.370 --> 00:05:50.920
couldn't get the same asset twice. It was a slot

00:05:50.920 --> 00:05:52.819
machine. And Nano Banana Pro, which is really

00:05:52.819 --> 00:05:56.180
just the Gemini 3 Pro image model, it solves

00:05:56.180 --> 00:05:58.560
this. It does. It has true character consistency.

00:05:58.939 --> 00:06:01.980
You can lock in a specific face, a specific outfit,

00:06:02.180 --> 00:06:04.480
and it stays the same across different scenes.

00:06:04.800 --> 00:06:07.319
The AI remembers who the bear is. What about

00:06:07.319 --> 00:06:09.800
text? That was always the big tell for AI images,

00:06:09.980 --> 00:06:12.899
the garbled words on signs. Gone. Text rendering

00:06:12.899 --> 00:06:15.439
is finally clean. You can put your slogan on

00:06:15.439 --> 00:06:17.959
a T -shirt in the image, and it's actually spelled

00:06:17.959 --> 00:06:21.250
correctly. That's huge. The guide talks about

00:06:21.250 --> 00:06:24.129
a workflow it calls the Brand Bible. How does

00:06:24.129 --> 00:06:26.009
that actually work? So this is the pro move.

00:06:26.209 --> 00:06:28.029
You don't just start from scratch every time.

00:06:28.110 --> 00:06:31.470
You create this master document, your Brand Bible.

00:06:31.930 --> 00:06:34.490
It has your hex codes, your fonts, your logo,

00:06:34.569 --> 00:06:36.970
your mood board images. And you just feed that

00:06:36.970 --> 00:06:39.709
to the AI? Every single time. You're priming

00:06:39.709 --> 00:06:41.970
the model. The guide says you can upload between

00:06:41.970 --> 00:06:45.759
1 and 14 reference images. So for our bear. The

00:06:45.759 --> 00:06:47.740
bear in the hoodie. You define it once, minimalist

00:06:47.740 --> 00:06:51.420
style, blue palette, hashtag 1E8085. You save

00:06:51.420 --> 00:06:53.699
that reference image. Then next week, you don't

00:06:53.699 --> 00:06:55.920
describe the bear again. You just say, using

00:06:55.920 --> 00:06:58.680
this character, generate a scene of them celebrating

00:06:58.680 --> 00:07:00.860
a win. And it looks like your brand because it's

00:07:00.860 --> 00:07:02.339
locked to that reference. It looks like your

00:07:02.339 --> 00:07:04.379
brand. And the cost change is just absurd. You

00:07:04.379 --> 00:07:07.399
used to pay thousands for a photo shoot, for

00:07:07.399 --> 00:07:09.759
models, a location. And now it's $20 a month

00:07:09.759 --> 00:07:12.660
for a tool. Exactly. It completely democratizes

00:07:12.660 --> 00:07:15.040
high -end visual branding. Okay, but here's the

00:07:15.040 --> 00:07:17.319
question that races. If I can do this for 20

00:07:17.319 --> 00:07:20.439
bucks, so can my competitor. So if everyone has

00:07:20.439 --> 00:07:24.360
perfect visuals, how do you actually stand out?

00:07:24.860 --> 00:07:27.379
If everyone has perfect visual consistency for

00:07:27.379 --> 00:07:29.899
20 bucks, how do you stand out? You stand out

00:07:29.899 --> 00:07:32.279
by having a better brand Bible and taste. Taste.

00:07:32.660 --> 00:07:35.279
The human eye is the advantage. Okay, let's shift

00:07:35.279 --> 00:07:39.339
to the third skill, AI video. This feels like

00:07:39.339 --> 00:07:41.139
the one that really just crossed the good enough

00:07:41.139 --> 00:07:43.290
threshold this year. Oh, yeah. This is the big

00:07:43.290 --> 00:07:45.290
one. This is where the magic is happening. The

00:07:45.290 --> 00:07:48.889
guide points to two main tools, VO 3 .1 and Sora

00:07:48.889 --> 00:07:51.129
2. Are they basically the same thing or are there

00:07:51.129 --> 00:07:52.589
real differences? Oh, they're very different.

00:07:52.709 --> 00:07:55.209
It's really a case of physics versus vibes. Physics

00:07:55.209 --> 00:07:57.730
versus vibes. I like that. Break it down. Okay,

00:07:57.769 --> 00:08:01.089
so VO 3 .1 is from Google DeepMind. It's the

00:08:01.089 --> 00:08:03.930
king of photorealism. It gets how light works,

00:08:04.009 --> 00:08:06.290
how gravity works. If you need a product shot

00:08:06.290 --> 00:08:08.689
of a soda can and you want it to look absolutely

00:08:08.689 --> 00:08:12.019
real, you use VO. It respects the laws of nature.

00:08:12.180 --> 00:08:15.100
Sora 2 from OpenAI is more of the creative artist.

00:08:15.339 --> 00:08:18.600
It's way better for stylization, for abstract

00:08:18.600 --> 00:08:21.120
concepts, for things that need to feel kind of

00:08:21.120 --> 00:08:25.160
dreamy. It bends reality, but in a good way.

00:08:25.500 --> 00:08:28.160
And there's a quick mention of Kling AI 2 .6.

00:08:28.339 --> 00:08:31.040
Yep. It's noted as an S tier tool right up there

00:08:31.040 --> 00:08:33.259
with the others in terms of quality. Now, the

00:08:33.259 --> 00:08:36.799
guide has a golden rule for video. It says to

00:08:36.799 --> 00:08:39.279
always start with images. Why? Why not just type

00:08:39.279 --> 00:08:41.940
in a prompt? Because text is just too ambiguous.

00:08:42.240 --> 00:08:45.159
If I tell you to imagine a woman in a cafe, you're

00:08:45.159 --> 00:08:46.700
picturing something totally different from what

00:08:46.700 --> 00:08:49.019
I am. Different lighting, different mood. Sure.

00:08:49.039 --> 00:08:50.700
I'm thinking a rainy Paris street. You might

00:08:50.700 --> 00:08:53.100
be thinking a bright Starbucks. Exactly. So if

00:08:53.100 --> 00:08:55.450
you just give that text to the AI. you get a

00:08:55.450 --> 00:08:58.350
random result, a lottery. But if you generate

00:08:58.350 --> 00:09:00.950
the image first with nano banana, you lock it

00:09:00.950 --> 00:09:03.029
in. You get the lighting perfect, the face perfect.

00:09:03.210 --> 00:09:06.210
That image becomes the anchor. So the video model

00:09:06.210 --> 00:09:08.649
isn't inventing the scene. It's just animating

00:09:08.649 --> 00:09:10.549
what you've already approved. Precisely. It's

00:09:10.549 --> 00:09:13.309
image to video. That's the secret. Okay, walk

00:09:13.309 --> 00:09:16.649
me through that workflow. Step one, generate

00:09:16.649 --> 00:09:19.710
your hero images. Get them perfect. Step two,

00:09:19.909 --> 00:09:23.669
animate them with VO. Add a slow camera orbit,

00:09:23.750 --> 00:09:26.710
maybe. Step three, create what the guide calls

00:09:26.710 --> 00:09:29.190
lifestyle moments. Someone using the product,

00:09:29.330 --> 00:09:31.950
smiling. Then you just cut it all together for

00:09:31.950 --> 00:09:34.730
social. It sounds super efficient, but the guide

00:09:34.730 --> 00:09:37.529
is also really honest about what AI video can't

00:09:37.529 --> 00:09:39.850
do yet. Yeah, and this is so important to remember.

00:09:40.049 --> 00:09:43.029
The biggest limitation is complex human conversations.

00:09:43.450 --> 00:09:45.669
Just people talking. Yeah, it still looks off.

00:09:45.789 --> 00:09:48.029
The lips move, but the micro expressions, they

00:09:48.029 --> 00:09:50.330
aren't there. It's deep in the uncanny valley.

00:09:50.429 --> 00:09:52.269
It just makes viewers feel uncomfortable. What

00:09:52.269 --> 00:09:54.970
else? Long narratives. Anything over 30 seconds

00:09:54.970 --> 00:09:57.750
starts to drift. The AI forgets what the character

00:09:57.750 --> 00:10:00.169
was wearing. And precise brand compliance, like

00:10:00.169 --> 00:10:02.190
getting a logo perfectly right on a moving shirt,

00:10:02.309 --> 00:10:04.649
is still a bit hit or miss. So we aren't filming

00:10:04.649 --> 00:10:06.629
the Super Bowl commercial with this yet? No,

00:10:06.710 --> 00:10:09.009
this is for the daily content feed beast, not

00:10:09.009 --> 00:10:12.009
cinema. Right. It feeds the algorithm. Okay,

00:10:12.049 --> 00:10:14.350
skill four. This feels more technical, but the

00:10:14.350 --> 00:10:17.830
source calls it a 10x multiplier. AI agents and

00:10:17.830 --> 00:10:20.889
automation. This is that huge shift from chatting

00:10:20.889 --> 00:10:24.149
with a bot to building a system. I think people

00:10:24.149 --> 00:10:26.730
get a little lost on the term agent. How is it

00:10:26.730 --> 00:10:30.669
different from just using ChatJPT? Well, a chatbot

00:10:30.669 --> 00:10:33.330
is passive. It just sits there and waits for

00:10:33.330 --> 00:10:36.049
you to talk to it. An agent is active. It runs

00:10:36.049 --> 00:10:38.350
in the background. It watches for things to happen.

00:10:38.370 --> 00:10:40.690
It makes decisions. And it takes actions without

00:10:40.690 --> 00:10:42.990
you holding its hand. It's like a digital employee.

00:10:43.169 --> 00:10:45.370
Exactly. A digital intern who never sleeps and

00:10:45.370 --> 00:10:47.919
never complains. The source gives this specific

00:10:47.919 --> 00:10:50.860
example using a tool called N8n. Can you walk

00:10:50.860 --> 00:10:52.399
us through that? I think this is where people

00:10:52.399 --> 00:10:55.360
get intimidated. Yeah, it looks scary, but it's

00:10:55.360 --> 00:10:57.279
really just a flow chart. You start with a blank

00:10:57.279 --> 00:10:59.720
canvas, and you drag these little nodes onto

00:10:59.720 --> 00:11:01.940
it. Okay. So your first node is the trigger,

00:11:02.059 --> 00:11:04.679
let's say, a chat command. The second node is

00:11:04.679 --> 00:11:08.299
the brain, the AI agent. You connect it to OpenAI

00:11:08.299 --> 00:11:11.100
and tell it, write an HTML blog post about this

00:11:11.100 --> 00:11:13.620
topic. Okay, so far that's pretty standard. But

00:11:13.620 --> 00:11:16.029
this is where it gets cool. The next node is

00:11:16.029 --> 00:11:18.830
an action Google Docs node. The agent literally

00:11:18.830 --> 00:11:21.429
creates a new document in your Google Drive.

00:11:21.909 --> 00:11:24.710
Then another node inserts the content. And the

00:11:24.710 --> 00:11:27.870
last node uses Gmail to email you the link. So

00:11:27.870 --> 00:11:30.110
in practice, I could type a topic into a Slack

00:11:30.110 --> 00:11:32.330
channel, and three minutes later I get an email

00:11:32.330 --> 00:11:34.409
with a link to a finished Google Doc. That's

00:11:34.409 --> 00:11:36.210
it. You didn't copy -paste anything. You didn't

00:11:36.210 --> 00:11:38.169
open a new tab. It just happened. Do you suck

00:11:38.169 --> 00:11:41.379
silence? Whoa, I mean... That sounds like actual

00:11:41.379 --> 00:11:43.860
magic. It really feels like it. And the guide

00:11:43.860 --> 00:11:46.620
compares this skill to knowing Excel. Excel.

00:11:46.919 --> 00:11:49.620
Yeah. 20 years ago, if you were in finance, you

00:11:49.620 --> 00:11:53.000
had to know Excel. The guide says marketers who

00:11:53.000 --> 00:11:55.879
understand automation tools like Make, ZP or

00:11:55.879 --> 00:11:59.220
N8N. They're the new Excel experts. It's becoming

00:11:59.220 --> 00:12:01.620
mandatory. This sounds like engineering, not

00:12:01.620 --> 00:12:03.759
marketing. Is the line blurring? The line is

00:12:03.759 --> 00:12:06.240
gone. The modern marketer is a systems architect.

00:12:06.500 --> 00:12:09.740
A systems architect. That's a heavy title. We're

00:12:09.740 --> 00:12:11.759
going to take a very short break. Mid -roll sponsor

00:12:11.759 --> 00:12:15.100
read. We are back and we are at the final skill.

00:12:15.320 --> 00:12:17.559
Skill number five, vibe coding. Vibe coding.

00:12:17.960 --> 00:12:21.039
I just love that term. It sounds so casual. Just

00:12:21.039 --> 00:12:23.740
doing some vibe coding. What is it really? It's

00:12:23.740 --> 00:12:25.879
really the democratization of making software.

00:12:26.179 --> 00:12:28.919
It just means using plain English to tell an

00:12:28.919 --> 00:12:31.820
AI what kind of tool to build. You don't need

00:12:31.820 --> 00:12:33.340
a computer science degree. You don't need to

00:12:33.340 --> 00:12:35.720
know Python. You just need to be able to describe

00:12:35.720 --> 00:12:38.559
the vibe or the function of what you want. So

00:12:38.559 --> 00:12:40.759
English is the new coding language. Exactly.

00:12:40.899 --> 00:12:42.919
And the only syntax you need to know is clarity.

00:12:43.320 --> 00:12:46.259
The guide mentions tools like ClaudeCode and

00:12:46.259 --> 00:12:49.259
Cursor. What are people actually building with

00:12:49.259 --> 00:12:51.120
this stuff? Well, we're not building the next

00:12:51.120 --> 00:12:53.799
big banking app or a secure OS. Right. Please

00:12:53.799 --> 00:12:56.240
don't do that. No, don't vibe code your security.

00:12:56.419 --> 00:12:59.399
We're building useful, maybe even disposable

00:12:59.399 --> 00:13:02.200
internal tools. Can you give me an example? Okay.

00:13:02.240 --> 00:13:04.419
Say you need a dashboard to track a campaign.

00:13:04.559 --> 00:13:06.879
Normally you'd file a ticket with IT. You'd wait

00:13:06.879 --> 00:13:09.259
six weeks. And by then the campaign's over. Of

00:13:09.259 --> 00:13:12.000
course. With vibe coding, you open cursor and

00:13:12.000 --> 00:13:14.850
you just say. Build me a dashboard that pulls

00:13:14.850 --> 00:13:17.970
data from these three CSE files and visualizes

00:13:17.970 --> 00:13:19.929
it with a bar chart. Make the background dark

00:13:19.929 --> 00:13:22.230
mode. And it just writes the code. It writes

00:13:22.230 --> 00:13:24.549
the code, shows you a preview of the app, and

00:13:24.549 --> 00:13:26.289
lets you talk to it. If the chart is the wrong

00:13:26.289 --> 00:13:28.509
color, you just tell it, make the chart green.

00:13:29.090 --> 00:13:32.190
You iterate in minutes, not weeks. It just removes

00:13:32.190 --> 00:13:34.490
that bottleneck completely. That's incredibly

00:13:34.490 --> 00:13:37.269
empowering. But the guide does mention what not

00:13:37.269 --> 00:13:40.090
to build. Besides security, are there other limits?

00:13:40.809 --> 00:13:42.669
Yeah, it's mainly about mission critical stuff.

00:13:43.110 --> 00:13:45.570
If the company would shut down if this tool breaks,

00:13:45.769 --> 00:13:49.049
don't vibe code it. Yeah. But for a quick competitor

00:13:49.049 --> 00:13:53.309
scraper, an ROI calculator for sales, it's perfect

00:13:53.309 --> 00:13:55.669
for that. This raises a really interesting point

00:13:55.669 --> 00:13:58.250
for me. If the barrier to entry is just describing

00:13:58.250 --> 00:14:01.529
what you want, what's the actual skill that's

00:14:01.529 --> 00:14:04.190
scarce now? If the barrier to entry is just describing

00:14:04.190 --> 00:14:06.929
what you want, what is the scarce skill? Clarity.

00:14:07.479 --> 00:14:10.399
The ability to clearly articulate exactly what

00:14:10.399 --> 00:14:12.720
needs to be built. Clarity of thought. That's

00:14:12.720 --> 00:14:14.200
it. If you have fuzzy thinking, you're going

00:14:14.200 --> 00:14:17.039
to get fuzzy code. The AI does exactly what you

00:14:17.039 --> 00:14:19.379
tell it. If you can't explain the logic clearly,

00:14:19.620 --> 00:14:21.860
it can't build the tool. So it actually forces

00:14:21.860 --> 00:14:24.799
you to be a better thinker. 100%. Clarity of

00:14:24.799 --> 00:14:26.779
thought is the new coding. We've covered a lot

00:14:26.779 --> 00:14:30.360
of ground. Content remixing, nano banana, AI

00:14:30.360 --> 00:14:34.080
video, agents, and vibe coding. Max Anne ends

00:14:34.080 --> 00:14:36.200
this guide with this conclusion about a chasm.

00:14:36.320 --> 00:14:38.120
Yeah, this is the big takeaway at the end. He

00:14:38.120 --> 00:14:40.759
splits all marketers into two groups, Group A

00:14:40.759 --> 00:14:42.779
and Group B. Okay, break that down for us. Who's

00:14:42.779 --> 00:14:45.500
in Group A? Group A are the people using AI as

00:14:45.500 --> 00:14:47.740
a force multiplier. They're building the agents.

00:14:47.860 --> 00:14:50.120
They're vibe coding dashboards. They aren't working

00:14:50.120 --> 00:14:52.139
harder. They're just working smarter. They move

00:14:52.139 --> 00:14:54.220
10 times faster because they have all this leverage.

00:14:54.440 --> 00:14:57.809
And Group B? Group B is stuck. They're the ones

00:14:57.809 --> 00:15:00.590
still manually copying and pasting data between

00:15:00.590 --> 00:15:03.529
spreadsheets. They're waiting three days for

00:15:03.529 --> 00:15:06.129
a designer to make a thumbnail. They are doing

00:15:06.129 --> 00:15:08.129
the rote work that machines have already solved.

00:15:08.309 --> 00:15:11.309
That is a very stark contrast. It is. And the

00:15:11.309 --> 00:15:14.029
guide's main point is that the barrier between

00:15:14.029 --> 00:15:16.730
group A and B isn't a university degree anymore.

00:15:17.289 --> 00:15:20.789
It's not about who knows C++A. Then what is it?

00:15:20.909 --> 00:15:22.789
It's curiosity. That's the only requirement.

00:15:22.929 --> 00:15:26.139
If you're curious enough to just open N8N. And

00:15:26.139 --> 00:15:29.419
try to connect two nodes. Even if you fail at

00:15:29.419 --> 00:15:31.779
first, you're putting yourself in group A. If

00:15:31.779 --> 00:15:33.220
you're just waiting for someone to give you a

00:15:33.220 --> 00:15:36.000
manual, you're going to stay in group B. That's

00:15:36.000 --> 00:15:38.840
a powerful place to leave things. But I also

00:15:38.840 --> 00:15:41.059
know, for you listening right now, if you try

00:15:41.059 --> 00:15:43.039
to do all five of these things this week, you're

00:15:43.039 --> 00:15:44.840
just going to burn out. Oh, absolutely. Don't

00:15:44.840 --> 00:15:47.279
do that. That's a recipe for disaster. I mean,

00:15:47.299 --> 00:15:49.700
I still wrestle with prompt drift myself when

00:15:49.700 --> 00:15:51.740
I try to do too much at once. So what's the real

00:15:51.740 --> 00:15:54.679
first step? Just pick one. Seriously. Blah, blah,

00:15:54.740 --> 00:15:57.080
blah. Maybe it's the nano banana images. Just

00:15:57.080 --> 00:15:58.940
go try to create a brand character and see if

00:15:58.940 --> 00:16:00.519
you can make it look the same in two different

00:16:00.519 --> 00:16:03.299
pictures. That's it. Or maybe it's the N8A agent.

00:16:03.659 --> 00:16:06.399
Just try to automate one simple email. Just get

00:16:06.399 --> 00:16:08.700
your hands dirty. Exactly. Don't stay in the

00:16:08.700 --> 00:16:10.740
audience. The only way to really learn any of

00:16:10.740 --> 00:16:13.879
this is to just do it. The chasm is here. The

00:16:13.879 --> 00:16:17.379
tools are here. The only question left is, which

00:16:17.379 --> 00:16:19.879
side are you standing on? That is the question.

00:16:20.240 --> 00:16:22.259
Thanks for listening to The Deep Dive. We'll

00:16:22.259 --> 00:16:23.100
see you on the next one.
