Latest / Bitcoin for Financial Services Podcast / Why 'Learn to Code' Aged Poorly: AI, Vibe Coding, and the Future of Financial Software

Transcript

Why 'Learn to Code' Aged Poorly: AI, Vibe Coding, and the Future of Financial Software · Bitcoin for Financial Services Podcast

0:00
I'm the grim reaper for these big tech companies because if you give a guy like me
0:04
access to these powerful coding models, suddenly I am literally a thousand times more productive
0:11
than I once was, which means that I can actually build a product which used to take Microsoft or
0:16
Apple $10 million to make. Now it costs me 50 grand to make. Bitcoin is a currency on the internet.
0:24
If you don't trust any other currency, this is where you go. It's going to outperform everything.
0:28
Bitcoin could eventually be worthless.
0:29
Every day is a good day to buy Bitcoin.
0:32
Bitcoin reached its highest price to date.
0:34
Bitcoin.
0:34
Hitting a new high.
0:35
It really has reached escape velocity.
0:37
There is no second best.
0:38
Who cares about Bitcoin?
0:40
It's the only secure database that's ever been invented.
0:42
Antisocial.
0:43
Bitcoin is the real deal.
0:45
Stupid.
0:46
Everything is bullish for Bitcoin.
0:48
Immoral.
0:49
It's going up forever, Laura.
0:50
Fix the money.
0:52
Fix the world.
0:53
Thank you.
1:23
Welcome to Bitcoin for Financial Services Media, the podcast where sound money meets professional services.
1:33
We explore how to build the financial services industry within the principles, values, and ethos of Bitcoin.
1:39
A big part of that is our flagship event coming up in Denver from October 20th through 22nd.
1:45
Legacy finance is being rebuilt in real time and Bitcoin is at the center of it.
1:50
The Bitcoin Financial Services Summit brings together the people actually building the infrastructure, RIAs, lenders, CPAs, miners, treasury companies, and Bitcoin native operators all in one room.
2:02
If you want to understand where Bitcoin is financed is going next, get your tickets now at bffs.media.com.
2:11
Looking forward to seeing you there.
2:12
We're joined by an extra special guest, Kendall Wye, co-founder of Bitment.
2:20
Let's get started.
2:26
Welcome back to another episode of the Bitcoin for Financial Services podcast, the media
2:32
company combining financial services and Bitcoin.
2:35
And today we've got a great guest on.
2:37
We've got Kendall Y, my co-founder in Bitment, the infamous.
2:44
I feel like everyone I talk to is like, who is this Kendall guy?
2:48
Can I maybe meet him one day?
2:50
This just happened on a conversation on a call yesterday.
2:53
And so now people are going to get to know him a little bit.
2:57
And Wyatt, you're on as well.
2:59
Good morning.
2:59
We're doing an early Friday morning pod.
3:01
And I've got my Bitment mug for the occasion.
3:04
Good morning, gentlemen.
3:07
Good morning. Pleasure to have you, Kendall. It's been a pleasure to get to know you some,
3:11
but I too had to wait a year between knowing of Kendall prior to meeting Kendall. So it's
3:18
a long line. You better get in while you can because there's no shortcuts.
3:26
I'm like the wizard behind the curtain. Jordan's role in my life is to protect me from this
3:31
that is true that is true we had we had to call an audible um you know guests always like people
3:42
get busy things fall through and so called kindle yesterday and said let's do let's rip a podcast
3:47
we haven't done in forever i mean kindle go maybe do that background first did we do a podcast
3:55
together. Could we do a podcast together again? What was that all about? We did do a podcast. So
4:03
Jordan and I met about five years ago, 2021. We had a mutual friend who connected us because
4:11
he knew that we were both of the same, cut from the same cloth, basically Bitcoiners,
4:16
entrepreneurs, that type of stuff. And Jordan bugged me for a couple of years. He's like,
4:22
we got to do something we got to do something i'm like i don't know man keep in mind i'm a
4:26
software developer so i'm working like my comfy software development job getting paid like a
4:31
crazy amount of money and i'm like i think i'm all right uh but then i do remember we were in
4:39
naples actually jordan we went to i forget what we were doing during the day we did some sort of
4:45
outing thing during the day and then we had dinner at a bar and then you're like all right kendall
4:49
Here's my idea, a podcast.
4:51
And I was like, all right, Jordan, let's do it.
4:56
So then we recorded, I don't remember how long it went.
4:58
It went for like a year, right?
5:00
We did it for like a year or something like that.
5:02
And then, and then I had to bounce.
5:04
So I was like, sorry, I gotta go.
5:05
I gotta go do something else.
5:07
So you took your CIA your CIA job and you like look I can speak publicly about global markets anymore Yeah Yeah The media space you know I don know
5:19
Some things are just better left unsaid to the public, I think.
5:23
I don't know.
5:26
Probably in 2021 and 22 and 23, basically under that regime, that probably was true.
5:32
But now it seems like people are just saying whatever they want and they don't care.
5:36
So we got you on.
5:39
So, okay, Kendall, give us your quick summary, your background, where you've been at, worked at, and what you're up to now.
5:48
Okay, so I'm the computer guy.
5:51
I'm the bits and bytes guys.
5:53
I write all the code.
5:56
So interestingly enough, you guys will love this.
5:59
When I was like, I was probably 14 years old.
6:04
So I'm 31 now.
6:05
So this is close to two decades ago, 2010.
6:09
I was an OG YouTuber.
6:11
I made YouTube videos all the way back in 2010 and I made them about like
6:16
Apple technology. I was like an Apple fan boy back there and back then.
6:20
So I bought like the first generation iPad and I was like blown away by this
6:23
thing. So, um,
6:25
I should have stuck with it cause I'd be making crazy money on YouTube right
6:28
now. If I stuck with it.
6:31
Anyway,
6:31
something happens in my life where I get bored with things and then move on.
6:35
but so anyway I got this long journey in the technology space I was like in high school I
6:42
worked as an Apple technician like in the back screwing you know screwing MacBooks and iPhones
6:47
all those things and then you know for the past decade or so I've been a software engineer
6:55
so I cut my teeth at KFC corporate so I'm from Kentucky so I went to the local the local brand
7:02
giant Kentucky fried chicken and, um, software developer there for four or five years.
7:09
And then, um, I always kind of knew that I wasn't cut out for the corporate world.
7:14
So I knew that I wanted to emerge into the entrepreneur space, but I'm also, I'm also
7:19
pretty cautious and conservative in my, in my movements.
7:23
So I'm never really super gung ho about just going all in type of thing.
7:27
So really for the past five years, um, I've been working in like the crypto fintech space
7:32
and you know obviously now george and and i are working together i am all in at this point with
7:37
with bitman so yeah that's that's the arc this episode is brought to you by satoshi paccioli
7:46
a full service accounting and tax firm built for the bitcoin economy if you run a bitcoin business
7:50
mine bitcoin or invest in it you need advisors who actually understand how this industry works
7:55
satoshi paccioli handles bookkeeping financial reporting tax compliance and strategic planning
8:00
for Bitcoin-focused companies and individuals.
8:03
They speak the language of Bitcoiners
8:04
and help you stay compliant
8:05
while building for the long term.
8:07
To learn more or book a consultation,
8:09
visit satoshipacioli.com.
8:12
Are you allowed to talk about the companies
8:15
that you worked with?
8:16
Like, obviously, you're a block.
8:18
That's fine to talk about.
8:19
And then you got to work with Nick Carter
8:21
a little bit at Coinmetrics, right?
8:25
Yeah, yeah.
8:25
So my first fintech or my first crypto company
8:29
was Coinmetrics.
8:30
which was a data analytics, crypto data analytics company based out of Boston.
8:37
So interesting business model, honestly, you just run, you know,
8:41
crypto nodes and you collect all the data and you sell it to Wall Street firms
8:45
and all these, all these institutions. It's a great business model, honestly.
8:50
And that was cool. I mean, I did that for 18 months.
8:54
That's Nick Carter's company.
8:55
So Nick Carter is the co-founder and chairman of the board.
8:58
That company actually recently sold to Talos.
9:02
And Talos is still a private company,
9:04
but will probably be public at some point in the next few years.
9:08
And then I got, I kind of got poached by Block.
9:11
So Jack Dorsey's company.
9:13
And I was, I was really invigorated by the project at the time.
9:17
It was the TBD project, which is like the, it was,
9:21
I was a believer back then that project kind of fell through.
9:25
So anyway, I worked at it.
9:26
Well, so it was called TBDEX and what it was, was like a decentralized exchange.
9:33
So this was back in like 2022, 2023, where like, what are the DEXs?
9:41
Like SushiSwap or like these crypto DEXs, where it's like on-chain decentralized exchanges were really ramping up.
9:48
the thing you have to understand about financial markets is probably the single most important
9:53
mechanical system is the order book so it's the buys the bids and the asks in the order book you've
10:00
probably seen that chart like on kraken where there's like a burn down of like green on one
10:04
side and red on the other side and they're basically matching all of these trades that's
10:09
probably the single most important system in entire financial world um in terms of the actual machine
10:14
And so like these, these DEXs were like these crypto bros.
10:18
We trying to do it on chain It not really the way to do it because it just doesn scale Like you can really scale a dex that way so but you know jack dorsey company obviously very bitcoin focused and so they didn
10:32
like all the coins and i was i was you know of the same of the same mind so really the proper way to
10:37
build a dex is actually off chain but you sort of reconcile things on chain and so really it was just
10:44
a it's just like a messaging protocol and there's this whole open source initiative around decentralized
10:49
centralized identity, which we can go into. You basically marry identity with this concept of a
10:55
messaging protocol and then like reconciling things on chain. That was the idea. And in theory,
11:03
that is actually, I think the proper way to scale like a global financial system for like the matter
11:09
of trading. But it was way before its time, like probably that might be the case in a few decades
11:15
type of thing. So that didn't really work out. And then for the past year or so, I worked at a
11:22
startup like in the stablecoin space. Sort of like a, I mean, it's like a payment slash capital
11:32
markets product. Stablecoins really had their whole time in the sun over the past 18 months,
11:38
which I think we could talk about that if you want. But I think that's kind of not as,
11:42
exciting these days as it once was.
11:48
I did that for a while and now I'm full-time BitMit.
11:52
Here we are.
11:54
Dude, that's awesome. I did not know that. A couple of questions
11:58
on some of your previous work experience.
12:02
Did you get to meet the colonel?
12:06
The KFC experience is really interesting. It's a great company to work for, honestly.
12:10
I mean, they don't pay great because they're not a tech company, but pretty good culture.
12:18
And you do get to do lots of fun things like that.
12:21
One thing that they make you do is every single employee has to spend two days a year actually working in the restaurants.
12:29
And so you have to go in the restaurant and be an actual employee of one of the corporate franchise restaurants.
12:35
And they always give me the hardest jobs.
12:37
I don't understand why.
12:38
they give me the job of being the the cook like the actual bride cook in the back and i was like
12:44
it is exhausting back there it's like you're sweating like lunch rush comes in it's like
12:52
it's tough it's not easy yeah that's that's awesome uh i think that's such a cool policy
12:58
uh and on a more more serious now you know one thing we haven't talked about in a while jordan
13:03
But I'm always interested how people get into the Bitcoin space because, you know, relatively, it's still a pretty small ecosystem.
13:15
Like, there's not a ton of big companies that employ a lot of people.
13:18
You've worked for a couple of those.
13:20
You know, so you went to CoinMetrics block after, like, cutting your teeth, you know, having the experience at CoinMetrics.
13:26
How did you get to CoinMetrics in the first place?
13:29
Were you, like, creating content?
13:30
Were you reaching out to people?
13:33
Yeah, that's a, that's an interesting question actually. Cause so when I left, I left KFC and I did actually kind of go gung ho in the entrepreneur space. So I, um, I, I kind of took like a six month period where I tried to build a startup and it was very unserious, honestly, like it didn't, it didn't work out. And like, um, it was kind of, I always knew it was going to be like a learning experience type of thing.
13:59
um but then so then during that period i was like really interested in like the crypto and
14:04
financial stuff and i've always been this type of guy where this is where i'm at currently this is
14:08
the second time i've done this where i will build a savings account and then i will you know once i
14:15
feel like that savings account is at a good enough number i will bet on myself and so i'll just live
14:20
off of savings type of thing um so i did that back then i'm kind of doing that again right now
14:26
with it meant to some extent um but so the first time when i when i did that i was really into the
14:34
crypto space and the financial space the startup i was trying to build was not in the crypto and
14:38
financial space and um and then i i was like well i kind of need another job again and so then i
14:45
just started you know scoping out the landscape i probably want to work in this space um and the
14:51
coin metrics came up on my radar and i was like really really fascinated with the work of nick
14:57
carter like i think nick carter is maybe one of the most to me the most important individuals in
15:04
the entire crypto you know fintech new new age uh new age financial system interesting we'll circle
15:11
back to that yeah his work is just very highbrow i know he's a little contentious in the bitcoin
15:16
So yeah, we can go into that if you want.
15:18
But anyway, so he co-founded Coinmetrics and I was like,
15:20
let me just reach out to these guys.
15:22
And so I just literally cold emailed them.
15:24
I was like, hey, this is me.
15:26
Just introduction.
15:27
I'm really interested in working on the crypto space.
15:30
You know, are you guys hiring type of thing?
15:32
And they got back to me and they like yeah actually we have a position open and the rest is history so cold emails absolutely work establishing real human connections absolutely
15:43
works your wealth deserves intention basilic financial is a bitcoin focused wealth management
15:51
firm built to help our clients find balance between money and meaning we integrate sound
15:55
money strategies across four distinct personalized services financial planning individual wealth
16:00
management retirement services for smbs and bitcoin financial services your prosperity your
16:05
purpose intentionally guided at basilic financial visit our website at basilic.io to schedule free
16:11
consultation truth um in that vein just because i have to mention now that was a phenomenal tf
16:21
bitcoin financial services summit october 20 22nd denver colorado make real human connections but
16:27
But to Nick Carter, I don't know him.
16:31
I have vaguely followed his work primarily through Twitter.
16:37
You obviously work for him.
16:41
Is Bitcoin Twitter shading my opinion on him?
16:45
Do you guys still follow his work?
16:47
Do you guys keep in touch or anything?
16:49
Is he misunderstood?
16:51
What are some of the important things you think he's done or been working on?
16:54
so i don't want to make it sound like me and nick carter are friends because we're not like
16:58
buddy buddies texting each other all the time that would be cool i'd be down for that nick hit me up
17:02
but um so i mean on one level you always have to understand that everybody's always talking
17:12
their own books and so you just have to be you just have to recognize that and nick carter's job
17:17
is that of a venture capitalist in the crypto space investing in blockchain companies so on
17:24
some level, he's always going to have some bias like that. That's, that's unavoidable. At the same
17:28
time, I think I try to see through the bias and see like, what does the individual actually believe
17:33
in? Like what's the system of, of morals and values, which the individual believes in. And I
17:39
think that Nick Carter's values actually are consistent with my belief of the long-term Bitcoin
17:45
values, which is that, you know, it's like, it's obviously it's ultimately, I think the systems of
17:54
values go, go back towards protecting the individual and from the collective, so to speak.
18:00
But the pathway to get there is, is not as simple as just be a really super hardcore libertarian.
18:06
The pathway is fraught with, uh, with complexities along the way. So I think that Nick Carter,
18:11
anytime you're you elevate i think you get elevated to the level of influence which nick
18:17
carter is at you um you're just in a really difficult position and you don't you can't
18:22
satisfy everybody you just have to try to stay like as authentic as you can um and i think nick
18:27
does a pretty good job at that also by the way so i comment this is why i don't like to do media
18:32
things because i don't want to be in that position but uh but at the end of the day like
18:39
the reason why I really appreciate Nick Carter's work is because I'm just the type of guy that I'm
18:46
nerdy enough to where like, I want to know the sophistication. Like I want to know like the sort
18:51
of obscure, uh, historical accounts of like things that people don't even ever talk about
18:57
everything. And I want to be like, I'm just, I'm just a ruthlessly curious individual.
19:02
And Nick's work almost always captures my, uh, almost always captures my attention like that.
19:08
So I know there's a lot of spats in the, in social media and stuff like that.
19:13
Those are fun and everything.
19:14
But at the end of the day, like I just, I just really appreciate Nick's work.
19:24
Jordan, Jordan, you're on mute.
19:35
Yeah.
19:36
Feel free to mute me and unmute me.
19:38
So we can cut that.
19:40
Kendall, given our audience of, you know, financial advisors and like CPAs, I'm curious to hear you kind of talk through really the last, I guess, 18 months since we've been building Bitment.
19:53
Like, maybe tell that story because it's very, you know, we've been building for the CPA space and now we've expanded that beyond to more financial advisors, attorneys.
20:04
And of course, AI has kind of blown up in the past, I don't know, two years, but has really ramped up in the last, I guess, seven months now.
20:13
I kind of think of the year end 25 going into 26 is like a real landmark in terms of AI being much more powerful.
20:22
So maybe give our audience that landscape, especially if there's someone who is out there.
20:27
They're an accountant, they're a financial advisor,
20:31
they're interested in Bitcoin, they're interested in AI,
20:33
but they don't really know how to use AI best.
20:38
They're still getting to know Bitcoin.
20:40
Maybe talk to that person and with the experience you've had
20:43
in the last 18 months of the products you've built.
20:48
Yeah, what's been some fun
20:50
insights that you've come across you think might be relevant for for our audience i was just
20:55
speaking with my friend yesterday about the the concepts of double entry bookkeeping and how
21:01
powerful the concept yeah i know right this is this is the type of things we get into
21:06
and it is the double entry bookkeeping is this like super powerful um it is like it's a it's
21:18
like an OG type of technology. It's just a, it's a standard way of doing things.
21:22
That's, that's Pacioli, right? Jordan. Yeah, that's true. Yeah. Pacioli created double entry
21:28
bookkeeping. So, um, I'm obviously not an accountant and I didn't know it. I didn't even
21:36
know what double entry bookkeeping was before we started working on Bitment, you know, 18 months
21:41
or whatever um and it's this it's like a classic trope where uh the software engineer comes in and
21:49
they're like oh yeah this is going to be easy like we just you know copy paste uh quickbooks
21:54
so i have absolutely been humbled by like the level of complexity in the accounting world i i
22:02
can obviously see now how it's like you know you have to go to college and get a you know a degree
22:07
and everything that's a different conversation obviously but my point is just that it's a it's a
22:11
sophisticated way of it's a sophisticated uh profession so that's been that's been humbling
22:18
for sure i think we've absolutely nailed that for sure um the matter of bitcoin and public blockchains
22:28
is just really interesting for that concept because reconciliation is easy to do it's all out there in
22:34
in the open. A large part of the Bitcoin story, I do still actually believe Jack Dorsey's framing
22:46
of the Bitcoin story is still massively undervalued, which is that, look, guys, don't
22:51
overcomplicate this. It's just digital cash. It's just digital money. The fact that you can just
22:57
treat it like digital native money, meaning it exists native to the computing systems, which
23:04
you and I are, or we're all three, you know, operating on right now as we speak, that is
23:09
just crazy in and of itself that we can, it exists solely in that realm.
23:13
And so therefore it's native, you know, it's native to that realm and it can be utilized
23:16
by things like AI.
23:19
So, um, obviously our background is, is in the financial space.
23:25
Um, that's like, you know, your, your background Jordan is in the, uh, CPA space, the tax space.
23:31
I have a background in the crypto fintech space, like we said. So it makes sense that we're focusing on that world.
23:38
To your point with AI, basically what happened about the turn of last year going into 2026 was the AI models became sufficiently powerful that in certain circumstances, they really did start to automate a lot of labor.
23:59
and that that changed things dramatically if you're trying to build a tech company
24:04
um and it's going to take we can go into it if you want but it's going to take five or ten years
24:10
for all those things to actually diffuse but the world of software engineering in particular
24:15
is like ground zero for this and um like my job has been radically changed over the past 12 months
24:22
and i didn't have any say in it right it just happened to me and you gotta you gotta dance
24:26
with the wins that you're given.
24:29
And, you know, so obviously we're focused on the financial industry,
24:34
but at the same time, Bitment is a technology company.
24:38
And so we are playing the technology game.
24:41
And so you have to actually approach it from like,
24:44
well, what's the landscape of current technology companies?
24:49
And I think the landscape over the next five to 10 years
24:52
is of course all about how does ai um how is ai utilized in these systems and the question is
25:00
mostly a question of uh economic productivity like at the end of the day the purpose of these systems
25:07
is to increase leverage for the productive individual and then we can go into like how
25:16
How does that fit in? But I think prior to call it December 2025, a tech company was built one way.
25:24
I think post-December 2025, a tech company is built an entirely different way.
25:31
Yeah, certainly. I think you have some really interesting thoughts in particular about what that means and how you guys are building Bitcoin or Bitmen.
25:44
So I want to circle back to that in a second.
25:47
I totally agree with you on the timeline.
25:50
I think January of this year is the first time I've ever gotten a terminal.
25:54
I'm not a tech bro.
25:56
And I was intimidated at first I was like wow Claude You like well this is like the Matrix There like green things going all around No literally it like I can read anything I don know a single
26:08
prompt line command.
26:10
But one of the things
26:13
that I think is interesting,
26:14
you know, just kind of like set this up into
26:16
the future is like I'm an enjoyer
26:19
of irony. And this
26:21
irony isn't lost on, you know,
26:23
probably 10, 15
26:25
years ago, whatever it was like.
26:27
all the talk you know it's like what are we going to do with steel workers oh just learn to code bro
26:33
and then like the first thing ai actually like first industry probably it disrupts the most to
26:39
your point it's like software engineering right and so i think that's kind of like a crazy
26:44
course of events learn to code okay what just kidding no literally i'm just like okay the first
26:50
job to play so like programmers so uh i was just curious on like your thoughts on that and like
26:55
being in the industry and having to like endure that?
27:00
So, I mean, it's always, these things are always like a double, you know, double-edged because
27:06
it's both, it's both the case that tech jobs for sure are the first to be disrupted.
27:14
And most individuals are attracted to the tech industry because they have
27:19
a predisposition towards constant change.
27:24
And so they're kind of adapted to it already.
27:27
It's kind of the part for the course for them.
27:29
That said, you know, it's not easy.
27:32
It's obviously not easy.
27:33
Um, um, you, I, it's, it's just an, it's a brave new world we're going into where you
27:43
have to be very adaptive and you got to be willing to learn new things.
27:47
And that is kind of a large part of, I think, everybody's job going forward is just continuous, continual learning.
27:54
But I can tell you, like, concretely, like, you know, prior to December 2025, I was, you know, I was using AI.
28:04
I was actually using cursor a lot, but I would still I still had a heavy hand in the review process and making, you know, reading every single line of code still.
28:14
after december 2025 i hardly ever write code i hardly ever read code and yet i still manage a
28:22
code base i so in this sense i i quite literally think of my ai agents as employees of mine like i
28:31
i got out basically i got elevated to the position of manager which i never even asked to be to to to
28:38
for that to happen and it's funny that have that people all throughout my career were trying to
28:42
elevate to me position of manager there's a couple times where it happened and i was basically like
28:45
i don't want to do that so but then it's just like forced upon me right but um that that is the
28:53
landscape now you know people have different desires and different preferences and everybody
28:59
has to figure out what what they want to do with life it's still fun for me so i'm having fun in a
29:04
different way um but but it's certainly different and the reason why it it comes it came to the
29:12
software development world first this is my my gambit is that software developers have been
29:20
trying to automate their job for the last 20 30 years so there's already a tremendous amount of
29:26
infrastructure in place there's two things there's infrastructure in place for software
29:31
which software developers were already building just for themselves to automate.
29:35
And then they can just now they can just layer in AI to do the automation.
29:39
The second thing is that so much of software is already universally standardized.
29:46
And so there's strict standardization across the entire software realm.
29:50
Like mostly a vast majority of industrial software, like probably greater than 90% runs on a Linux based system.
29:58
So it's some sort of Linux operating system.
30:01
mac os is actually a linux-based operating system windows is the one odd ball out actually windows
30:07
is a is a m dos or something like that ms dos um so you have like an extremely robust set of
30:15
standardization in the software space which is like here's a little bit of alpha for you which
30:20
is that like the reason why cloud code is so powerful in the terminal is because it can
30:26
directly leverage the Linux tooling.
30:30
And so it's interfacing directly with that standardization.
30:33
Whereas if you use Claude in the application, you have an abstraction
30:38
layer between you and the standardization.
30:40
And so it's not as familiar with the underlying standardization.
30:44
So anyway, this is the reason why software was the first to be
30:48
automated like this.
30:50
other industries, other domains have far less standardization.
30:56
And so they are far,
30:58
they're just gonna take a lot longer to automate.
31:02
And this is where it becomes like,
31:04
what I refer to as human systems.
31:08
Like the reason why for example the financial space is so not standard The reason why you hear all these words these obscure words thrown around left and right by people in the financial space is because they are very human systems
31:27
It's the historical account of the financial system is just humans talking to each other.
31:32
and because humans are talking to each other and all this you know diverse uh set of backgrounds
31:37
it's there's no one way of doing it right and so for that reason it's um it's going to be
31:45
it's still going to diffuse and these things are going to like these systems are going to automate
31:50
all of these white collar jobs it's just that um the there's a there's a long process time
31:57
for the for the infrastructure and the standardization to come into play
32:01
Do you mind to go a little bit deeper into the whole like vibe coding?
32:10
Because I think there's a lot of advisors probably listening to this that are like, yeah, I played around with the tools and it's blown my mind.
32:16
Right. I'm going to do X, Y and Z in my firm.
32:18
and I know I could speak to this a little bit, but I'd love to get your perspective on it,
32:23
on just, and especially in this space where you're dealing with the sensitive documents that
32:30
financial advisors, tax people, attorneys deal with. Can you just talk us through
32:36
why vibe coding is very interesting up until a point, and then it's a complete waste of time
32:44
because ultimately there's data security.
32:48
And if you let some of that data leak out to a bad actor,
32:54
obviously that has huge implications
32:55
for the letters behind people's names.
32:58
So just kind of curious to get your take on that
33:01
as you're talking to an audience of people
33:03
that are probably playing around with these tools,
33:07
feeling like they're more productive,
33:08
but ultimately are probably actually working more now
33:12
because they still have to get their work done
33:13
manually because it's not embedded.
33:16
Like the AI is not embedded into their workflows.
33:20
Well, and then Kendall, just building on that too,
33:22
maybe if you could reference like some specific tools,
33:24
you know, when Clawbot came out,
33:25
you know, it was like all the rage
33:27
and you felt like an idiot if you weren't using one,
33:29
but then like a couple of weeks go by
33:31
and people's code bases are getting deleted.
33:33
Emails are getting sent, you know,
33:34
so you like contextualize this with like
33:36
maybe some of the specific developments.
33:40
Yeah, so there's a lot that can be said here.
33:42
A lot of it's pretty abstract and philosophical, but I would say at scale, like five or 10 years from now, the job of a software developer is largely that of basically like a cybersecurity engineer.
33:54
And so the industrial value of an individual who says they are a software engineer is largely that to make sure that the systems are secure, cryptographically, you know, secured and the networks are secure and all those things.
34:10
And the processes are not going to be auto deleted by rogue AI, which, by the way, that whole open AI thing with hugging face, that's a psyop, 100% psyop.
34:21
Oh, interesting.
34:22
yeah it's but anyway we don't have to go to that um um but so but the thing is we're not actually
34:31
even there yet like you'll hear that that's that's a narrative right now that you'll hear
34:35
but it is a narrative which is it like you can't do this on your own because you can't trust that
34:41
you're not gonna you know delete your systems or expose client data yes that's true but that's
34:46
honestly not the biggest thing right now here's how i describe it
34:50
building a software application is like building a skyscraper. So a skyscraper is this enormous
34:59
piece of civil engineering with, you know, it's on a grand scale, both literally from our
35:05
perspective and that it's very tall, but then also on the detailed level. So you literally have
35:10
architectural blueprints down to like individual screws or washers on individual screws in a door.
35:17
right so you have this tremendously large project on the on the macro scale and every detail also
35:26
has to be in place prior to ai coding um it was as if it was the way you know um skyscrapers have
35:37
always been built which is that you have humans walking around you have welders that are welding
35:42
to things. They have to use these large cranes to pick up the heavy stuff. After AI coding,
35:50
it's as if the law of gravity no longer exists. And so it's as if an individual can,
35:56
they themselves just lift up a giant steel I-beam and put it into place themselves.
36:02
And you can see how that would obviously speed a lot of things up. Obviously, if I don't have to
36:08
rely on a slow moving crane, then I can just literally pick up a giant I-beam and stick
36:14
it into place with my own two hands.
36:16
It's going to speed things up dramatically, but it's actually not going to speed things
36:20
up a ton to be honest because you still have all these little nuts and bolts You still have all these little washers everywhere Everything has to be put into place So there all these details And that what you pay an engineer for
36:33
All that level of comprehensive detail to make sure everything is in place.
36:38
And it sounds, the job of software is to make things anthropomorphized such that they're
36:47
easy to use.
36:49
And these software systems are sycophantic.
36:51
And so they create the appearance that they're very easy.
36:54
But if you actually go through the legwork of building it,
36:58
you will realize that it's building a skyscraper and there's like little
37:02
washers and nuts and bolts everywhere. And it's not fun. Honestly,
37:06
you had to like, it's very laborious.
37:08
Even with these systems is very laborious because you got to go through every
37:12
little thing. It's like, so anyway, that's the job of a software engineer.
37:16
And so the point of like the vibe coding stuff is still fantastic.
37:21
because it actually does have a very good use case,
37:24
which is that if you're a non-technical person
37:27
and you're trying to convey an idea to a technical person,
37:31
a great way to do it is to just vibe code the thing.
37:35
That will convey,
37:36
that acts like a great communication layer to the engineer.
37:39
Now the engineer is going to take that
37:41
and they're going to totally throw out all the source material,
37:43
but they're going to take the inspiration
37:45
and they're going to actually build the real thing.
37:48
So yeah, that's how I think this works.
37:51
Yeah, and that's been our life for the last 18 months. And I think this is ultimately what a lot of service providers are going to have to do over the course of the next five to 10 years.
38:05
Because prior to this, everyone in my world and Wyatt's world, we just were forced to use off-the-shelf, mass-marketed software. And it was like, okay, just pick on Asana. Asana is this project tracking tool.
38:20
and it could slot in anywhere.
38:23
But then you can't really customize it
38:25
to exactly your use case.
38:28
You can customize it to a degree,
38:29
but you can't connect it easily
38:31
and you could be a Zapier,
38:33
but then you have to manage those connections.
38:36
And so you're constantly having to just use
38:38
what's given to you as a non-technical person,
38:41
click around, set it up to the best of your ability,
38:44
probably not using it to the fullest extent
38:46
because you don't have time as a practitioner
38:48
to really dive deep.
38:49
Whereas I think now we're moving into this world where guys like you, Kendall, are working with guys like us actually together to go and say, what is the custom solution that you actually need? And let's go build that.
39:06
That's this forward deployed engineer model if you want to go deeper into that. But I'm just curious on your thoughts on the world we're leaving versus the world we're going into specifically for the financial services space and just the prospect of customization and automation so that we become essentially what you are today, where you go to a client, you listen, you understand what they need.
39:34
then your agents are doing the work on the back end.
39:37
I think that all of us in the financial services
39:40
are going to experience that in the next five to 10 years
39:42
where we're that human in the loop
39:45
or the person they shake hands with
39:46
or jump on a call with and talk through things.
39:49
But ultimately building a financial plan
39:52
or building a tax return or a tax plan
39:54
in our scenario for Wyatt and I,
39:56
that's actually not going to be work
39:58
that we're cognitively doing anymore.
40:00
We'll be reviewing that work
40:01
and then we'll be communicating with the client.
40:03
So just curious to get your thoughts on all of that.
40:07
Okay.
40:07
So you'll hear this word diffusion, right?
40:12
In the AI space, which is that the, in my own explanation of the matter, it's like diffusion
40:19
is the concept of AI is super valuable for particular use cases right now, not so much
40:25
for other use cases.
40:26
And over time, that productivity will diffuse into the broader economy.
40:30
So things which are seemingly not exposed to AI, the classic example being a plumber, will at some point be, you know, AI will do the job of a plumber.
40:41
That will absolutely happen, by the way.
40:44
So that's the concept of diffusion.
40:47
Another word for diffusion is decentralization, right?
40:52
Because historically, these big tech companies are monopolies and they are monopolies.
40:58
They're allowed to be monopolies, basically.
41:00
there's a giant centralization of power by the tech companies.
41:06
So you have, you know, the Googles, the Microsofts, the Apples.
41:11
What AI does, AI properly regulated,
41:15
and it does have to be regulated to do this,
41:18
actually acts as a decentralization mechanism
41:22
to where these giant big tech firms actually lose their monopoly powers
41:28
and they lose it through the mechanical process
41:30
through which they lose it
41:31
is because of guys like me,
41:33
which by the way,
41:34
Apple, if you want to buy me out for 10 million,
41:36
you can do that.
41:39
There is a price.
41:39
seems stiff but okay
41:42
not true we have to take them down kindle
41:48
i'm just kidding i'm conveying a point which is that
41:54
i am i and like that's that's a huge that's an enormous difference
41:59
so um so this is the process of of diffusion decentralization
42:05
On the other hand, there is still like these things are always there's always a tension. There's always like a Scylla and Charybdis. I just saw the Odyssey and it was great. And the other side of the coin is that the AI systems work best through this mechanism of standardization.
42:24
The matter of standardization is on some philosophical level, a type of centralization, because you're creating a single process, a single protocol for every single use case.
42:38
So you have both these, you have these dynamics, you know, butting up against each other where it's like you have this diffusion process, which will actually decentralize technology against the standardization process, which will actually centralize technology.
42:54
And so in the case of, I'm going to bring this full circle, in the case of the financial services world, it looks something like this, in my opinion.
43:06
Right now, a lot of financial operators or operators of a lot of different industries, they rely on these big tech products.
43:13
They rely on the Asanas of the world or XYZ.
43:17
XYZ. And that big tech company is some company far away, potentially in a political jurisdiction,
43:26
which you disagree with. You can never even talk to them. You can't call them up.
43:31
They are maybe using AI models, which are censoring your thing. They're not sovereign
43:35
or something like that. They're monetizing your data. They're very far away. There's like
43:40
basically an abstraction. It's not human. It's like the system you're dealing with the system.
43:44
in these human systems such as the financial system government politics things like that
43:51
uh that doesn't work you need a real human on the other line you need to you need to be able
43:57
to talk to the technology you need to be able to interface with with the real human nature with all
44:04
of its warts and all right that's that's the feature not the bug so you you have this process
44:11
of diffusion against this process of standardization and then in the in the in the case i think how that
44:17
translates to places like the financial world is that you do get more proprietary what i call boutique
44:24
software experiences which is what you're referring to jordan when you say like we can build sort of
44:29
custom applications for people you get this you get this um this process where because a guy like
44:37
me is able to build, you know, really powerful things. Now I can come do it for relatively cheap
44:43
relative to what it used to be. And I can build, um, a financial firm, a piece of custom software,
44:50
which is super valuable to their exact use case and it's boutique. It feels like their own and
44:58
they can call me and they can talk to me. Um, and that, that is, I think that's the dominant force.
45:05
insofar as, for example, open source AI is allowed to grow.
45:13
That's the regulatory component to this.
45:15
If open source AI is actually shut down,
45:20
then that dynamic, it's no longer the dominant force.
45:23
The dominant force is still the centralization of the big tech companies.
45:28
So, yeah, that's how it works, I think.
45:31
Yeah, I mean, that was awesome.
45:32
There's like three different questions I have of that.
45:36
I'll lay them out for you and then you can pick which one you want to dive into.
45:40
Because one of the things I'll attribute to you guys, I think is like really smart.
45:45
And if you frame this in like the context of BitMent and your guys' development, you know, there's this vertical versus horizontal and operating system level, you know, AI integration.
45:55
That's one.
45:56
number two would be then
45:58
so that customization
46:01
is that
46:03
like vibe coded?
46:04
Am I vibe coding that Am I empowered as a practitioner to build this customized tool for a client or something Or do I actually need an engineer to your previous point like identifying the bolts
46:17
and washers or is like the vibe coding good enough for the customization?
46:21
And then the third would just be like words of wisdom to practitioners like Jordan and
46:27
I, because I don't disagree with you, Jordan, like, you know, five, 10 years, like I totally
46:30
see like the agents doing the work and like we're reviewing.
46:34
But then to your point, Kendall, will a software development experience, you know, this, this change first and you yourself like are not reviewing like the code anymore.
46:44
Like what's, what's actually like that balance.
46:49
Yeah.
46:50
Okay.
46:50
So.
46:51
And you can, you can, you can pick any of those.
46:53
I want to flow into the conversation.
46:55
Just that one little snippet you had there.
46:57
I was like, okay, now I got three questions.
47:00
Oh, I know these things are, these things are very complex.
47:02
The thing is, is like solving actual industrial value is actually really difficult.
47:09
It sounds easy.
47:10
And sometimes you can get lucky if you find like a specific narrow vertical, like a specific use case where it's like all we need is like a little program.
47:20
We need a little script.
47:21
And suddenly my day is like 50% easier.
47:24
You can get lucky like that.
47:26
In my experience, a lot of that actually does not exist.
47:31
And instead, it's like a lot of these.
47:37
So on some level, the artificial intelligence is like a type of bureaucracy.
47:48
And so bureaucracy is sort of the antipathy to growth.
47:53
If you want to actually grow, you need to limit bureaucracy on some level.
47:59
and so ai is good at you know creating bureaucracy and so it's easy to like vibe code something
48:08
the problem is and it may solve a problem right away the problem is how does it fit into the
48:13
broader whole how does it fit into the comprehensive conceptual uh of your of your industrial use case
48:20
and on the whole although it may be great in one particular instance on the whole more likely than
48:27
not, it probably actually will slow you down. And this is the matter of bureaucracy, right? It's
48:34
like, this sounds like a good idea. We just need a committee. We just need a committee to figure it
48:39
out. And boom, that's the solution, right? If you're a hammer, everything looks like a nail.
48:48
But in reality, you have to think about the broader whole. So this is where I think the tech
48:56
company, like what it takes to build tech company starting December 2020, 2025 changed.
49:03
There was this longstanding orthodoxy in Silicon Valley startup land where you just find a specific
49:10
niche, you find a specific niche and you go after it and you solve one particular problem.
49:17
That makes sense in a world of optimization where you're optimizing the economy under the constraint
49:24
that software engineering is expensive and difficult.
49:28
If that constraint goes away,
49:30
then that orthodoxy no longer makes sense.
49:33
And that constraint went away
49:34
through these AI coding tools.
49:38
So in that world,
49:40
I think what you have to do is,
49:42
like our strategic gambit at Bitment
49:45
is that you have to,
49:47
if you want to actually automate industrial use cases,
49:50
like if you want to get to a world
49:53
where you as a financial provider are not doing any of the cognitive labor behind the scenes,
49:58
you're just speaking with clients, you're just acting as this intermediary between
50:01
clients on one side and your agents on the other. If you want to get to that world,
50:07
this goes back to the matter of the infrastructure and standardization that I was getting to earlier,
50:11
you have to actually build the broad horizontal operating system to enable that as the
50:18
comprehensive hole. Like if you just try and build a single vertical of automating your bookkeeping,
50:25
yes, you can do that. My argument would be net net. It probably actually negatively impacts the
50:32
whole So if you if you do want to if you if you do want to actually solve these use cases you have you have to solve it at the operating system level If you a financial professional working with Bitcoin you know the pain of messy CSVs broken cost basis and accounting tools that were never built for this asset
50:53
Bitment fixes that.
50:54
Bitment is the Bitcoin native accounting platform built for CPAs, bookkeepers, advisors, miners, and serious Bitcoin investors.
51:02
You get clean Bitcoin bank statements, reliable cost basis tracking, full reporting, and fiat integration through Plaid.
51:10
No crypto talk, no workarounds, just proper accounting for the hardest money on earth.
51:15
Create a free account or book a demo at bitment.co today.
51:23
Okay, so Wyatt always does this because I'm going to have to jump in like four minutes.
51:27
So I want to make sure I give Wyatt the last word
51:30
and not try to end it too quickly.
51:34
Thanks.
51:35
Thank you.
51:36
It's nice to see some development on your part of the evolution of our...
51:42
I'm learning.
51:43
By the way, it's also like 5.30 for Wyatt right now in the morning.
51:47
So it's like very early.
51:49
Well, and I would argue I've asked the best questions too.
51:52
You have asked the best questions.
51:55
I talk to Kendall so much.
51:56
So I'm like, I don't know what to ask him anymore.
52:00
Right.
52:02
No, well, I just think that that's like so fascinating about the like vertical and horizontal, you know, aspect of how you laid that out.
52:13
And I think that's like some some pretty profound wisdom.
52:17
And then like, I've just like felt this anecdotally.
52:20
It's like, wow, I feel so empowered by AI.
52:23
I spent some time building a little internal app that I've used a couple times,
52:28
but I really felt like I needed it this one time,
52:31
and this was going to save me so much future headaches or whatever,
52:35
and it just hasn't really amounted to that.
52:38
So I've been onboarding to Bitmap+.
52:40
I'm just blown away by what you guys have built
52:43
And just like having like that physical, you know, visual, like across different, you know, applications that I've used, you know, in the past integrated into like an AI native operating system.
52:56
It's just like, holy smokes, this is crazy.
53:00
I guess my final question building on that, it does leave me confused though, because I'm like empowered by vibe coding, clearly not capable or have the time or willingness to build.
53:13
like an operating system. So now I kind of feel like I'm in no man's land a little bit. So I need
53:18
like, should I vibe coach? Should I go on the operating system? Like, you know, what actually
53:23
is it? So it's both, it's like both, right? It's like you vibe code as a mechanism to provide
53:28
yourself the knowledge of like, what do I actually need solved? It helps you form up
53:34
your product design, your product experience, like not non-technical people live in the real world
53:42
and they actually deal with like real humans and they speak with real, real humans.
53:47
And, and that's, that's a messy process.
53:49
Software is just an encoding.
53:51
It's just a represent, it's a representation or representation of reality.
53:56
So there's a, there is a literal process there of going from reality to, from a, from a presentation
54:02
to a representation.
54:04
So the process of vibe coding is actually that process of building muscles internally
54:09
for a non-technical person to go into that world.
54:12
But then, yeah, I mean, actually providing industrial use case, it's still a difficult process.
54:21
I mean, I obviously have a bias.
54:22
I mean, I want people to pay for bitment and buy for bitment.
54:26
There are certainly going to be use cases where you can just vibe code a single application and it works really well for you.
54:34
And that's good.
54:36
But my gambit is that if you – here's the way to think about it, actually.
54:39
like employees employees are you know human labor is not cheap it's like you know you got to pay
54:46
somebody i don't know 100 grand a year to do a job that's not cheap right that's like expensive
54:52
of course relative to like your claude bill uh that's like claude is basically worth zero
54:59
compared to that so claude is effectively free what i would just say is that it's
55:05
It's expensive for good reason.
55:08
Like it cost for good reason because there a lot of things there that are outside of the realm that like you just can it it difficult to articulate So this process of vibe coding and building a product
55:20
I think is an entryway into like that whole skyscraper thing of like realizing
55:25
that, Oh, actually we have, I didn't even realize it,
55:27
but we have a skyscraper that we've already built. We just have to,
55:30
we just need to encode it in software. So, yeah.
55:33
Yeah. And I mean,
55:34
I would just end it saying that ultimately, I think everyone's going to have to move to
55:40
this world where you have a fractional CTO that you work with.
55:45
And I think that's ultimately what we're building at Bitment now is it's this boutique experience.
55:52
You know, maybe we can assign five to 10 clients per software engineer.
55:58
And so you have the CTO and you might still use Vibe Coding as that communication mechanism
56:04
to say, hey, we were brainstorming internally and we thought, what if we did it this way
56:10
within the OS?
56:12
Or what if we added this new feature set?
56:14
And you vibe code what you think it should look like from a front-end UX standpoint.
56:20
And you might use Cloud Code yourself to do that.
56:23
You hand that off to your outsourced CTO and then give them a week, two weeks, a month,
56:29
whatever the size of that project is.
56:32
and then you come back and it's like you're a kid on Christmas.
56:35
And that's been my experience literally for the last 18 months
56:37
is like I'm a kid on Christmas Day, you know,
56:41
many times when I hop on with Kindle where I said,
56:45
hey, we need to make these fixes to the bookkeeping product
56:48
or hey, what if we added this feature?
56:51
And then give them a week and it's like, all right,
56:53
Kindle presents it to me.
56:54
It's like, holy crap, that's insane.
56:56
I could have never done that.
56:58
And so I think that's the ultimate mixture,
56:59
at least in the current state we're in, is you can dream things up as a practitioner,
57:07
but you should not be the one that's implementing it because either you're going to expose your data
57:12
or you're just going to lose your mind because it doesn't work exactly how you want it to.
57:17
So let someone else fine tune it. And instead of paying a software engineer, you know, 250 or more
57:22
a year, which is what the going rate generally has been for a long time. Well, now you can get
57:28
that person for a lot cheaper. And it's the same way people have outsourced the CFO role
57:33
instead of hiring a full-time person. So man, Kendall, I think we'll have to bring you back on.
57:39
We'll have to just maybe keep you as a friend of the pod because there's a lot to chew through.
57:46
I did too well. I should have held things back.
57:52
All right, man. Well, very cool. Any last thoughts from either of you?
57:56
this is cool this is fun yeah this is great right man really appreciate you coming on that was
58:01
awesome absolutely cool all right thanks y'all for listening and we'll see you next week
58:06
we appreciate you joining us if you want to help build the future of bitcoin financial services
58:15
hit subscribe leave a review sign up for our newsletter and share this with someone in your
58:21
network who's integrating bitcoin into their practice if you work in finance and want to
58:25
understand how Bitcoin is changing wealth management, lending, custody, retirement planning,
58:30
and capital formation, you need to be at the Bitcoin Financial Services Summit October 20th
58:36
through 22nd in Denver at the space. We're bringing together some of the smartest operators
58:40
in the industry for practical conversations focused on what's actually working today.
58:46
Register now at bffs.media.com. The more professionals we reach, the stronger that
58:53
our ecosystem becomes and our clients are better served.
58:57
Stay sharp.
58:58
Stay sovereign.
58:59
Keep leading from the front.
59:23
your trusted friend or your sworn ally
59:28
know it's the harshest mistress of all
59:36
and life is just a shame