0:02 Imagine it's a few years from now. AI
0:04 can build almost anything, write almost
0:06 anything, and do most of the tasks
0:09 people used to get paid for. In that
0:09 world, [music]
0:12 I got to ask, what skill is still
0:14 valuable? I've been thinking about this
0:15 non-stop, [music]
0:18 and I've narrowed it down to six core
0:20 skills. None of them require a fancy
0:22 degree or connections, and all of them
0:25 could be started this weekend. And every
0:26 single one of them gets [music] more
0:30 valuable as AI gets better, not less. By
0:31 the end of this episode, you'll learn
0:34 the six most valuable skill sets on the
0:37 planet with some tips to get started.
0:38 I'm going to go through each of them
0:40 super clearly, so you'll understand
0:42 where the world is going, what you need
0:43 to learn,
0:43 >> [music]
0:45 >> and just concretely
0:48 how to get started. Let's not waste any
0:49 time, [music] and let's go right through it.
1:00 >> So, the first skill that I think is the
1:02 most valuable skill right now is
1:05 people who can set up agents properly,
1:09 manage them, and run local AI models.
1:11 This is basically the grown-up version
1:13 of prompt engineering. So, a lot of
1:15 people learned how to type a good prompt
1:18 into ChatGPT, which is useful, but the
1:20 next layer is being able to design a
1:23 little AI employee that has contacts,
1:25 that has tools, that has permissions,
1:28 that has memory, that has a goal, and a
1:31 way to check its own work before it
1:34 bothers you. That skill, that little
1:37 skill, is going to be valuable because
1:39 most companies are about to have the
1:41 exact same problem.
1:44 They're going to have 10 AI tools, 50
1:47 workflows, a bunch of half-working
1:50 automations, and nobody understands how
1:52 to turn that into an operating system,
1:54 which is what they really want. The
1:56 person who can walk in and say, "Hey,
1:58 here's a customer support agent, here's
2:00 a research agent, here's the sales
2:03 follow-up agent, here are all are all
2:05 the rules, here's what it's allowed to
2:07 do, here's where it needs approval,
2:09 here's how we know if it's working."
2:12 That person becomes really, really hard
2:13 to replace.
2:16 And the local AI part, which I've talked
2:17 about on this channel before, I've done
2:19 a couple videos on it,
2:22 is important because there are certain
2:24 workflows where privacy or cost, you
2:26 know, the prices of these models are
2:29 going up and up, latency or control
2:32 matter a lot. If you could run these
2:34 models locally with something like
2:36 Ollama, LM Studio, you start to
2:39 understand what can happen on your own
2:42 machine and what needs the cloud and
2:44 what needs to touch private docs and
2:46 what should stay behind the wall and how
2:48 this all interact. So, I think this
2:50 whole idea around local is going to be
2:53 more and more important as time goes on.
2:55 Even if, cuz a lot of people, you know,
2:57 in the comments will say, "Well, local
2:59 models are going to going smaller." Even
3:02 if local models go smaller, you learn
3:04 the architecture for the future. You
3:08 learn which jobs need a giant brain and
3:10 which jobs just need a reliable worker
3:13 that never sleeps. So,
3:14 what would I do? What's a rep I would do
3:17 to learn this skill set? The first rep I
3:18 would do
3:19 is simple.
3:22 Build a daily briefing agent for
3:25 yourself. Give it three sources. Give it
3:28 your calendar. Give it a folder of notes
3:31 and give it a few a few saved links. And
3:34 its job is to tell you what matters
3:36 today, what decisions are waiting for
3:40 you, and what follow-ups you owe people.
3:43 Then, you can add one rule.
3:46 You can say it has to show sources and
3:48 ask for approval before sending
3:51 anything. That one project teaches you
3:54 context, it teaches you retrieval, it
3:56 teaches you tool use, it teaches you
3:59 permissions, and it teaches you evals.
4:01 So, it might sound small and maybe a
4:04 boring first agent to build, but I I
4:07 believe that that is basically the shape
4:11 of every serious agent inside a company.
4:13 The mistake people, I think, make is
4:16 they try to build an all-knowing agent
4:17 first, these really big agent pro projects.
4:19 projects.
4:20 The better move is basically just to
4:23 build small agent to start, get it to be
4:26 very valuable, schedule it, have a clear
4:30 success metric, um and, you know, the
4:32 success metric might be did it save me
4:34 10 minutes? Did it catch something I
4:36 would have missed? Did it produce
4:39 something I would have actually used?
4:40 Um if the answer is that is yes, then
4:43 you're learning the skill.
4:45 So, that's the first
4:47 super super valuable skill. And by the
4:48 way, this is in no particular order.
4:50 These are just the six that, you know,
4:52 the more you know, the better. The
4:55 second skill is marketers who know how
4:57 to build distribution.
4:59 I think this one is underrated because
5:01 people confuse distribution with posting.
5:03 posting.
5:05 Uh distribution is way more deeper than
5:08 just like posting on social media. It's
5:11 knowing where attention already lives,
5:14 what people are already anxious about,
5:16 what language they use when they
5:18 describe the problem, and how to turn
5:21 that into trust before you ask them to
5:24 buy anything. And in AI world, we all
5:26 know building products are, you know, is
5:29 really easy. Uh building demand is just
5:31 getting more and more important. So,
5:32 when anyone can ship a landing page or
5:34 an app,
5:37 you know, or build a SaaS, uh the bottle
5:39 the bottleneck moves to the question of
5:41 can you make people care? And that's
5:42 where someone who's really good at
5:46 distribution uh is a pro app. So, the
5:48 marketers who are going to win in this
5:50 agentic era are going to be part
5:53 researcher, they're going to be part
5:55 storyteller, they're going to be part
5:57 media operator, and they're going to be
5:59 a part community builder. They're
6:02 basically going to know how to take one
6:05 insight and turn it into a tweet, uh a
6:07 short-form video,
6:10 uh a YouTube title, a newsletter angle,
6:12 a landing page headline, a founder
6:14 story, and a sales conversation. So, in
6:17 a sense, it's almost like a marketer is
6:19 becoming like a general a generalist
6:21 marketer. We're seeing this as a as a
6:22 greater trend, like people are becoming
6:25 generalists, they can do multiple things
6:26 because if the if their job is to manage
6:29 agents, uh they need to understand
6:30 different components of that. And that's
6:32 exactly what's happening in the
6:34 distribution marketing
6:36 uh world.
6:38 So, the first rep I would do if I'm
6:41 trying to learn distribution, uh is I
6:43 would do a distribution map. So, I'd
6:46 pick a niche I care about, uh like
6:49 dentists using AI or solo consultants,
6:52 maybe real estate agents, Shopify
6:54 operators, whatever. Could be a business
6:55 that you want to start or you you know
6:58 you're you've you're building already.
7:00 Then, write down the 20 places the
7:03 attention their attention
7:05 uh goes. Uh so, like the newsletters,
7:07 the creators they pay attention to,
7:09 maybe it's like Reddit threads that get
7:13 popular, Slack groups, podcasts, events,
7:15 search terms, the tools they already pay for.
7:16 for.
7:19 After that, write one painful sentence
7:20 they would actually say out loud,
7:22 something like
7:24 um I know I should follow up with leads
7:27 faster, but by the time I sit down and
7:29 do it, half of them are cold.
7:31 That sentence is where distribution
7:33 starts because you're getting you're
7:34 basically like
7:37 transporting or or yourself into their shoes.
7:38 shoes.
7:41 Um then you should do the second rep, so
7:43 the the evolution of this. Write 20
7:46 hooks for the same idea. Make some
7:50 curiosity hooks, some fear hooks, some
7:53 status hooks, some money hooks, some I
7:56 wish I knew this earlier hooks. If you
7:58 want to become great at distribution,
8:00 you don't you're not you don't want to
8:02 ask yourself, "How do I promote this?"
8:04 after the product is already done,
8:05 right? So, you want to start asking
8:09 yourself, "What existing desire am I
8:12 pointing this at?" before you built. And
8:14 I think that that shift alone, that
8:18 mindset is going to change the quality
8:20 of your ideas, and that's what makes
8:21 someone who's really good at, you know,
8:24 distribution in this era. So, you know, TLDR
8:26 TLDR
8:28 on on distribution
8:30 is you want to put yourself in their shoes,
8:31 shoes,
8:34 and you want to be this like part
8:37 storyteller, part um
8:40 uh researcher, part media operator, and
8:43 and really just have a lot of
8:45 shots on net in this in this world
8:47 because, you know, some some are going
8:49 to win and some aren't going to win.
8:51 Um and on this on the Startup Ideas
8:53 podcast, on this channel, um
8:54 um
8:55 you know, I'll share more of these tools
8:58 as as I'm learning in real time. So, you
9:00 know, feel free to like, comment, and
9:02 subscribe to to get more of this in your feed.
9:03 feed.
9:05 The third most valuable skill is
9:08 robotics engineers who can basically
9:11 build hardware, wire in AI, and source
9:14 manufacturing. Now, I know that most
9:16 people, probably a very small percentage
9:20 of people, actually has any uh
9:22 experience in robotics engineers. But,
9:25 the reason why I put this in here,
9:28 um before I explain exactly how to do
9:31 it, is because, you know,
9:34 software was an incredible business uh
9:36 for last 20 years, building SaaS. And
9:38 there's still opportunities in in
9:40 building SaaS, building consumer mobile
9:42 apps, enterprise apps,
9:45 but the mode is is moving to hardware.
9:47 And I'll explain more
9:49 uh I'll explain more about this.
9:53 So, um and by the way, I think a lot of
9:55 people are sleeping on this. Um
9:57 basically, the last decade, the internet
10:01 rewarded people who moved pixels around.
10:03 Uh I was one of those people. Um but the
10:05 next decade is going to reward people
10:08 who can move atoms around, too. That was
10:11 like my big insight. Um robotics used to
10:14 be this PhD-feeling thing, expensive
10:17 parts, custom hardwares, weird tooling,
10:19 long timelines. You had to go to school,
10:21 get a PhD in robotics to go and build
10:23 something like that. But now, the world
10:25 we live in is is a lot different. You
10:28 have this open-source robot learning
10:30 projects, you have cheap cameras, you
10:32 have low-cost arms, you have better
10:35 simulation, you have multimodal uh motor
10:37 models, and you have community-sharing
10:39 data data sets.
10:41 So, you have, you know, companies like
10:44 Hugging Face, who has Lo Robot, which is
10:46 basically trying to make robot learning
10:48 more accessible. And even on Hugging
10:51 Face, which is like a uh a d- almo-
10:52 almost like the way I think about it is
10:54 like a database of all these open-source
10:56 projects that you can go and download
10:57 those repos,
11:00 you can find something some some some
11:01 really interesting open-source
11:04 technology and inject it into a robot,
11:06 uh and who knows, that could be the next
11:09 big thing the next big thing.
11:12 Um there's low-cost arm projects like
11:15 SO-100 and the SO-101 ecosystem. And
11:17 there's also smaller uh
11:20 vision-language-action models like the
11:22 uh small VLA that are pushing towards
11:26 robot policies you can train and run
11:27 without needing some
11:29 uh giant industrial setup.
11:32 So, the interesting thing here is it
11:35 goes beyond the AI layer. Uh It's the
11:37 person who can make the whole loop work,
11:40 right? So, it can you get a cheap arm on
11:43 your desk? Can you mount the camera? Can
11:45 you collect demonstrations? Can you
11:47 train or fine-tune the model? Can you
11:50 make the robot repeat one useful task?
11:52 Can you look at a supplier listing in
11:54 China and understand if this thing's
11:57 actually manufacturable? So, the person
11:58 that can do all three, the building the hardware,
11:59 hardware,
12:01 wiring in the AI, sourcing
12:03 manufacturing, understanding that, come
12:06 on, that is just some skill set to to
12:07 know. And this is something that I'm
12:10 learning in real time
12:12 because I think it's so important.
12:15 If I wanted to learn this skill,
12:16 my first
12:19 rep would be extremely concrete. So, I
12:20 would buy
12:24 or assemble a low-cost robot arm. You
12:26 can add a cheap camera to that.
12:28 And then I would teach it one boring
12:32 task like sorting three objects,
12:34 pressing a button, or moving it from one
12:36 tray to another.
12:38 Then I would document every failure. The
12:41 camera angle was was bad maybe or the
12:42 lighting changed.
12:45 Maybe the gripper slipped or the data
12:47 set was too small.
12:49 The model looked smart in one setup and
12:51 then fell apart when the object moved
12:53 like 2 in.
12:54 That's kind of the point. This is how
12:56 you learn a skill.
12:59 Robotics specifically teaches humility
13:02 humility pretty quickly
13:04 and that's the humility that becomes
13:05 your expertise.
13:07 On the sourcing side, this is really
13:09 important. I would learn the basics of
13:11 working with suppliers. A lot of us
13:13 listening to this podcast, you know,
13:15 we're we're software people. We like
13:17 digital products.
13:18 But it's also important to learn
13:20 physical products, too, cuz of all the
13:23 opportunities that are coming. So, go on
13:26 Alibaba or similar uh, marketplace. I'm
13:28 not affiliated with Alibaba, um,
13:29 um,
13:31 and study how components are sold. Ask
13:34 for a sample before you talk about bulk.
13:36 Um, you can ask for motor specs, you can
13:39 ask for controller board details. Uh,
13:41 you can ask for CAD files if they exist.
13:44 Uh, replacement parts, lead times,
13:47 minimum order quantity, shipping terms,
13:49 and a short video of the part doing the
13:52 exact thing that you need. So, you're
13:53 you're going to be learning a new
13:56 language, um, and the language is, you
13:58 know, can this actually be made,
14:01 shipped, repaired, uh, and used by a
14:03 normal person?
14:05 Um, this skill is so rare, uh, because
14:07 it sits between worlds. Uh, software
14:10 people often avoid hardware,
14:12 um, and hardware people sometimes avoid
14:15 distribution and NAI. So, the person who
14:18 could connect open-source AI models,
14:21 physical prototyping, and manufacturing
14:23 has a shot at building things that feel
14:25 like science fiction, but sell like
14:27 practical tools, and I think there's
14:29 just so much opportunity here.
14:32 The fourth undeniable skill to learn
14:35 right now is curate- curators. So,
14:37 understanding how to curate, who are
14:40 good at yapping, and can do short-form
14:42 video in their sleep. I'm sure you've
14:45 seen these people, uh,
14:47 you know, yapping, uh,
14:49 on Instagram, and they're taking over
14:52 your algorithm. And I mean yapping in in
14:53 the best possible way. You know, the
14:56 internet is drowning in information, and
14:58 the person who can make sense, uh, of it
15:00 in public, um,
15:02 is, you know, very valuable. And
15:05 curation has evolved past past like,
15:06 here's five links in a newsletter,
15:10 right? Curation is like, uh, here's, you
15:12 know, five products that I think in this
15:14 niche that you'd really like, and
15:16 explained in a really storytelling,
15:17 really cool way.
15:20 So, the curator of the agentic era
15:22 watches the timeline and says, this
15:24 matters because {dot} {dot} {dot}." Like
15:26 they understand that. They can see a new
15:29 model demo or a weird startup launch or
15:31 a robotics clip, a policy change, a news
15:34 item, a pricing update, a story about
15:37 XYZ, and they can translate it for that
15:40 particular niche. What should you learn?
15:42 What should you ignore? What should you
15:44 try this weekend? What is hype? What is
15:46 actually useful? So, to be amazing at
15:47 content, this is sort of the big
15:50 insight, uh, which might sound obvious,
15:53 but the the the insight is like, you
15:55 know, you don't need to be super smart to
15:57 to
15:59 get millions of followers in social
16:01 media or or just doesn't even be
16:04 millions, get 50,000, 100,000 followers
16:06 in a niche and build an incredible
16:07 business around that, um,
16:09 um,
16:12 by creating net new content. You can
16:14 just look at what's happening in your
16:17 niche and curate really interesting
16:20 things in short form in an authentic
16:23 way where you're just yapping to your
16:25 phone and the algorithms right now are
16:28 prioritizing yapping. Why they
16:30 prioritizing yapping?
16:33 They are prioritizing it because they
16:34 are seeing AI slop move into the
16:36 timeline and people are getting tired of
16:38 that. People don't want that, right? So,
16:40 there's nothing more raw than authentic
16:42 and being like, "Hey, my name is Greg
16:45 Eisenberg and I suffered from XYZ, um,
16:47 until I found these like five really
16:49 interesting products or this story or I
16:51 met this person that helped me and let
16:53 me tell you about it."
16:56 That is the type of content that the
16:58 algorithms in the timeline are
17:00 promoting. So,
17:02 before I give you the rep, actually, I
17:04 would say like the yapping matters so
17:06 much, like learning how to yap is just
17:09 is is just such a good skill, um,
17:10 um,
17:13 and the people who are great at it,
17:15 frankly, uh, make you feel like you're
17:17 getting in this like group chat of a
17:19 research report because it's fast, it's
17:21 opinionated, it's useful, and a little
17:24 entertaining. Um you know those people
17:25 in your niche that you see them yapping
17:27 and you're like, "Oh my god, they're
17:29 Like I feel like I know this person,
17:31 right?" So, let's say you wanted to
17:34 actually, you know, become a curator
17:36 {slash} yapper or learn about it, I
17:39 would do uh a 7-day curation sprint. So,
17:42 for 1 week, pick a lane. Um maybe it's
17:45 AI agents for real estate, maybe it's
17:47 robotics for small businesses.
17:49 You know, whatever niche it is. But
17:52 every day, find three things and make
17:54 one short video using the same structure.
17:55 structure.
17:58 I saw this. Most people will think it
18:01 means this. I think it actually means
18:03 this. Here's the move.
18:05 That structure forces you to have a
18:07 take, which is the difference between
18:09 curation and forwarding links.
18:12 The thing with a lot of yappers and
18:14 curators that do what they do really
18:16 well is they have a take. They're either
18:18 anti-something or pro-something. So,
18:21 have a take, right? Um the key thing
18:23 here is you're going to want to build uh
18:25 you know, some people call a swipe file,
18:27 some people call it a taste file.
18:30 Basically, a document of examples you
18:33 love. Great hooks, great analogies,
18:36 great titles, weird use cases, comments
18:38 that reveal what people are like
18:41 genuinely confused about. Curators are
18:43 obviously only as good as their taste
18:46 inputs. So, if your inputs are generic,
18:48 your outputs are going to be generic. Um
18:50 if your inputs are weird and specific
18:53 and high signal, people are start are
18:54 going to start coming to you and
18:56 trusting you
18:58 because you consistently find the thing
19:00 before they do. So, it's worth you're
19:02 worth the follow.
19:04 Um so, that's an incredible skill set to
19:08 learn. Um the next skill set, uh the
19:10 fifth skill set,
19:13 is what I call the building distributor.
19:14 So, it's the person who can ship both
19:18 the product and get in front of people.
19:20 So, this might be the most important
19:21 skill set if you're a founder, if you
19:24 want to build a business, um
19:24 um
19:26 because for years,
19:28 uh there was this clean split.
19:31 Uh one person would build and one person
19:32 would sell.
19:34 You know, you'd have your Wozniak, who
19:35 was like the technical person, and then
19:37 you'd have your Steve Jobs, who was like
19:39 the marketer salesperson.
19:41 You know, one person writes code, one
19:43 person writes copy, one person makes the
19:45 thing, one person gets the attention.
19:48 And AI in this agentic world is com-
19:50 compressing the split.
19:52 So, one person now could prototype the
19:54 product, make the landing page, you
19:56 know, write the launch thread, you know,
19:58 make a video about it, record the demo,
20:00 DM the first 100 users, edit short-form
20:03 clips, iterate based on feedback,
20:04 pretty, you know, pretty much everything.
20:04 everything.
20:07 Um so, that person has the leverage
20:09 because they don't have to wait for the
20:11 handoff. They can complete the loop
20:13 themselves, so nothing really gets lo-
20:17 uh lost. So, the loop is the whole game,
20:18 because if you built something small and
20:21 you put it in front of people, uh watch
20:22 where they get confused, you know,
20:24 change the product, change the story,
20:27 try again. Most people only do half.
20:29 They build it in private forever or they
20:32 talk about it in public forever and they
20:33 actually never ship the thing. The
20:37 builder distribu- distributor uh learns
20:39 by cycling between both.
20:41 I think um
20:43 this is so cool because I think that,
20:44 you know, when people talk about the
20:47 one-person, one-billion-dollar startup,
20:50 you know, Sam Altman talks about it, um
20:52 you know, I think that person is going
20:54 to be the builder distributor. And when
20:56 you start seeing people like Peter who
20:58 who founded Open Claw, you can tell that
21:01 he's not only incredible at building,
21:03 like he built an incredible thing with
21:06 Open Claw technically, but he also is,
21:07 you know, if you go to his acts and you
21:09 just see how he speaks, he's an
21:12 incredible marketer, too. And and he's
21:14 incredible at, you know, doing customer
21:16 support and so many things just by
21:17 little things, you know, I don't know
21:18 him personally, but just little things
21:20 that I've picked up on.
21:23 This is one of those people, right? Um
21:26 so, let's say you want to be um
21:27 good at this skill. Well, you know,
21:29 what's what's a one quick thing that you
21:32 can start by honing in on your skill. So,
21:33 So,
21:36 I would do what's called a 48-hour loop.
21:40 So, pick one tiny problem you personally
21:42 understand, and then build the smallest
21:44 version with AI.
21:46 Um it can be ugly. Uh it can be a script,
21:47 script,
21:50 a form, a simple web app, an automation,
21:53 uh you know, anything. Then create 10
21:55 pieces of distribution for it before you
21:57 even feel ready. So, that could be one
22:00 demo video, three short clips,
22:03 you know, maybe three posts, two DMs to
22:04 people who have the problem, and then a
22:07 landing page. You're basically training
22:09 yourself to stop step separating the
22:11 product from the market. So, what's
22:13 powerful is that AI makes the building
22:15 part faster. We all know this.
22:18 So, the marketing part uh learning can
22:21 start way earlier. Uh you don't need to
22:22 spend now 6 months wondering if people
22:25 want it. You spend a weekend building
22:27 enough to earn a real reaction, and then
22:29 the builder distributor is dangerous
22:31 because now all of a sudden you can turn
22:33 attention into product feedback and
22:37 product feedback into better attention.
22:38 So, it's this beautiful loop that you
22:40 can end up uh be building. Um and the
22:43 only way to become an in in incredible
22:46 builder distributor is you got to spend
22:48 more time building, and then spend more
22:50 time launching and distributing, and
22:53 then building that loop. Um I talk a lot
22:54 about on this channel this concept of
22:57 the ACP funnel. It's how it's the future
22:59 of building businesses. You know, you
23:02 build audience at the top, then you
23:04 convert that to community, and then you
23:06 build a product there, and that's also a
23:08 loop that
23:10 uh you know, the builder distributor uh
23:12 is excellent at.
23:15 The last skill that I think is just so
23:19 valuable in this era is IRL community
23:21 builders. So, this one feels almost old
23:23 school, especially, you know, because we
23:25 talked about like AI robotics and
23:28 open-source technology and local AI, but
23:29 that's kind of like that's kind of why I
23:33 like it. Um as more work moves to agents
23:36 and chats and tools and feeds, real
23:39 rooms actually become more valuable.
23:41 People still want to meet other
23:43 ambitious people. Uh they still want to
23:45 meet people like them uh who are into
23:48 the same things that they might be into. Um
23:48 Um
23:50 and they also still want trust, and they
23:53 want energy, and they want to be around
23:56 others who who teach them things or who
23:58 are they just entertained by. So, AI
24:02 makes content abundant. Um it makes
24:03 software abundant. It makes advice
24:06 abundant. Um so, where does scarcity
24:08 move towards? Well, scarcity moves
24:11 towards belonging, trust, and contexts.
24:13 Who do you actually know?
24:16 Who would answer your text? Who would
24:17 help you hire?
24:19 Who would you intro- who would introduce
24:21 you to a customer? Who would tell you
24:24 the honest version of what's happening
24:27 in their market? The IRL community
24:30 builder knows how to create that. They
24:31 know how to pick the right room, set the
24:33 right topic, invite the right mix of
24:35 people, and create a ritual that people
24:37 want to come back to.
24:38 A great community is actually more like
24:41 a habit than an event.
24:44 Same time, same kind of people,
24:46 same promise, better conversations each time.
24:47 time.
24:49 The first uh rep, if I was trying to get
24:53 good at IRL events, which
24:55 by the way, I should I should note like
24:58 there's hundreds of millions or
24:59 you know, billions of dollars up for
25:01 grabs for people who can create
25:03 incredible events. And I'm not just like
25:05 pulling numbers well, you know, just
25:08 like randomly like you can look at huge
25:10 event companies
25:13 that just absolutely crush it selling
25:15 events. Like you know, if you even look
25:17 at tech like look at
25:20 SaaStr. I think Jason Lemkin created an
25:24 event all around SAS. And you know, I
25:25 think he's probably shared some of his
25:27 numbers like they're massive. If you
25:28 look at
25:30 South by Southwest, you know, these are
25:33 huge events.
25:35 You know, these are there's just a lot
25:36 of opportunity. I think that and I think like
25:37 like
25:39 in general
25:41 people don't want massive events
25:43 anymore. They actually want these kind
25:45 of smaller more bespoke events and
25:47 that's where the opportunity lies.
25:50 Okay, so let's say you wanted to get
25:53 good at being the IRL community builder
25:55 person. You know, what's one little
25:57 thing that you can do to learn the
25:59 skill? Well, why not why not host you
26:03 know, six seven eight people around one
26:04 sharp question?
26:05 So you don't start with this massive
26:08 event. You start with like a dinner or a
26:11 walk a hike a breakfast.
26:14 And the question could be you know, if
26:16 it could be something like
26:17 what skill are you learning because of
26:19 AI? Let's say if you wanted to build a
26:21 community around AI it would be
26:23 something like that. Or what are you
26:25 automating in your company right now or
26:27 what do you think everything in tech is
26:29 is missing? And then you invite people
26:31 who can actually answer it. Then after
26:33 the event you send a short recap with
26:36 the best quotes, inside jokes maybe
26:39 ideas and what one follow-up everyone
26:40 should do.
26:42 The recap is important because it turns
26:44 the room into a network. So that's the
26:46 whole goal with the whole IRL community
26:49 builder. How do you turn these rooms
26:50 into a network? Because that's what
26:53 creates memory. And it gives people a
26:55 reason to forward it. It makes the next
26:59 invite easier. So over time, the room
27:01 becomes a media asset, a recruiting
27:04 asset, a deal flow asset, and honestly,
27:07 a life asset. So I think this skill
27:09 pairs beautifully with others. The agent
27:11 person who can build tools for the
27:14 community, the marketer can grow it, the
27:17 curator can turn the best conversations
27:19 into content, the builder distributor
27:21 can launch products from it, the
27:23 robotics person can bring the weird
27:25 demos, and that's when it starts getting
27:27 really interesting. So,
27:29 um there you have it. Those are the six
27:31 skills that I think matter most in the
27:35 agentic uh era. So, TLDR, what are the
27:37 six skills? One, people who can set up
27:40 agents properly, manage them, and run
27:42 local models. Two, marketers who know
27:45 how to build distribution. Three,
27:46 robotic engineers who can build
27:49 hardware, wire in AI, and source
27:51 manufacturing. Four, curators who are
27:53 good at yapping and can do short-form
27:56 video in their in their sleep.
28:00 Five, the builder distributor, the one
28:03 person who can both ship the product and
28:06 get in front of people. And six, the IRL
28:09 community people uh the the IRL
28:11 community builder who's bringing people
28:13 into these rooms and starting networks
28:15 from an IRL perspective.
28:18 The bigger point of all this is
28:21 that the future favors the person who
28:24 can combine all these capabilities.
28:26 Um there's obviously too many tools to
28:28 know all of them, and the advantage goes
28:30 to the people who know how the pieces
28:33 fit together. Can you make agents uh
28:36 useful? Can you get attention? Can you
28:38 build physical things? Can you explain
28:39 what matters? Can you ship and
28:41 distribute? Can you bring together in
28:44 real life? That is that skill stack, and
28:47 this is the six skill stacks that, you
28:49 know, pick one and you know you don't
28:51 have to be amazing at all six but pick
28:53 one and get dangerous. You know pick two
28:55 and you have some leverage.
28:57 Um but pick three and you become you
28:59 know the kind of person that everyone
29:01 wants on the team in the room or
29:03 building the company.
29:07 Um we're in for a crazy next 5 10 15
29:09 years about what's going to happen
29:11 uh in the job market in the economy. No
29:14 one really really knows. But the one
29:16 thing we do know is that knowing these
29:20 skills or or skills in general it's
29:22 going to make you it is your defense. It
29:24 is your shield. You know I believe that.
29:27 So I wanted to make this episode because
29:29 um I think a lot of people like know
29:31 that they should be doing something but
29:32 they're not sure what they should be
29:35 doing. So I wanted to put just this into
29:37 one place that's really simple
29:39 easy to understand the six most
29:41 important skills. You know share this
29:43 with a friend who you know might might
29:46 and this might be helpful to. Uh hope it
29:48 hope it has been helpful to you. And
29:50 I'll see you at the next one. Thank you
