The Lindy Effect

Spurred by a reading from Nassim Nicholas Taleb, Jorge and Harry have a rambling conversation about building resilient businesses during times of disruptive technological change.

• Antifragile: Things That Gain from Disorder by Nassim Nicholas Taleb
• Peak Priorities
• unfinishe_
• The Great Mental Models by Shane Parrish
• Xenix
• Cursor
• Lovable

Episode Transcript

Harry (00:00):
People aren't asking that question, "Hey, is this gonna
be a temporary thing or not?"They're going ahead and building
out...
they're talking to their LLMand they're generating some code
and they're, "Look, it works.
It's so cool!" And what they'vejust done is they've built a
perishable thing, but they'rethinking about it as a
potentially permanent thing.

Narrator (00:22):
You're listening to Traction Heroes.
Digging In to Get Results withHarry Max and Jorge Arango.

Jorge (00:32):
Hey Harry.
It's good to see you.

Harry (00:35):
Oh, fantastic to see you again, Jorge.
Thanks for making time today.

Jorge (00:39):
I always have time for you, my friend.
What has changed since we lasttalked?
Because we talked about changelast time.

Harry (00:47):
What has changed?
Let's see.
It's raining.
We had some of the mostbeautiful weather of the year
the last few days and I see theraindrops coming down, so the
simple pragmatics of weather,those things are changing.
We got our new websitelaunched.
I'm very excited to say.
We got a new assessment justabout completed for determining

(01:14):
your competency level assomebody who might be interested
in learning how to prioritizebetter.
So there's so many thingshappening.
I'm super excited about it.

Jorge (01:23):
When you say we, are you talking about your consulting
business or...?

Harry (01:27):
My apologies.
Yes, Peak Priorities, mylittle...
the shingles hanging down andthe little nascent consulting
company, which we do speaking,coaching, executive coaching,
consulting, and certain specialprojects.
I love it.
I'm just incredibly blessed.

Jorge (01:48):
Fantastic.
I'm gonna check it out.
Last time I was there, you hadjust opened up the newsletter,
so it sounds like there's morestuff there.

Harry (01:56):
Oh.
And yeah, and I published anarticle recently, which you may
have received on CollapsingSliding Windows.
Yeah, that new strategicplanning construct.

Jorge (02:05):
Yeah, if folks get value from this podcast, they really
would get value from yournewsletter.
So highly recommended.

Harry (02:12):
Thanks.
Did you bring anything to readtoday?

Jorge (02:15):
I did, I brought a reading and how about I just
barge into it?

Harry (02:21):
Okay.
Sounds good.

Jorge (02:25):
"When you see a young and an old human, you can be
confident that the younger willsurvive the elder.
With something non-perishable,say a technology, that is not
the case.

We have two possibilities: either both are expected to have (02:38):
undefined
the same additional lifeexpectancy (the case in which
the probability distribution iscalled exponential), or the old
is expected to have a longerexpectancy than the young, in
proportion to their relativeage.

(03:00):
In that situation, if the oldis eighty and the young is ten,
the elder is expected to liveeight times as long as the
younger one.
Now conditional on somethingbelonging to either category, I
propose, (building on theso-called Lindy effect in the

(03:21):
version later developed by thegreat Benoît Mandelbrot):" this
is now me giving you like the,air quotes thing, right?
this next part is in italicsbecause this is the thing the

author is proposing (03:35):
"For the perishable, every additional day
in its life translates into ashorter additional life
expectancy.
For the non-perishable, everyadditional day may imply a
longer life expectancy.
So the longer a technologylives, the longer it can be

(04:00):
expected to live."

Harry (04:03):
That is so interesting.
It's so obvious after I heardit, and of course never would've
been able to put it into thoseterms.
Where in the world did youstumble on this?

Jorge (04:17):
This is from Nassim Nicholas Taleb's Antifragile.

Harry (04:21):
Oh yeah.

Jorge (04:24):
And this is an idea that didn't originate with Taleb, but
it's the first time I cameacross it, which is the Lindy
effect.
And I read this passage hopingthat we'd talk about this idea
because for me, like you'resaying that it makes a lot of
sense.
For me, when I first read it,it felt really counterintuitive.

(04:48):
And then the more I thoughtabout it and the more I became
aware of it, the more I waslike, "That's an example of the
Lindy effect.
And that's an example of theLindy effect.
Yes, this is a thing!" Right?
Like the more something and hemakes a distinction between
perishable and non-perishablethings, right?
the more somethingnon-perishable has been around,

(05:09):
the more likely it is to bearound.

Harry (05:13):
Yeah.

Jorge (05:14):
Which is a really interesting idea, particularly
in our time of great changewhere there's all this new stuff
popping up.

Harry (05:23):
I'm just reminded of just how brilliant he is as a
thinker and as an author.
And I did not read Antifragile.
Of course, I understand theconcept behind the book.
But, that's a pretty strongendorsement to go pick up a copy
of that book and read it coverto cover.
I also love how you seem to befinding these handles like the

(05:48):
Lindy effect or Chesterton'sfence or something that you can
put in my backpack and carrywith me and use once I
understand it.

Jorge (05:57):
This is the mental models thing that Shane Parrish talks
about, right?
Like the idea that there arethese ideas that become part of
your mental toolbox, way oflooking at the world.
We've talked about framing inthe past, right?
Like the frames you bring tothings.
And I think that a lot of whatwe do in these conversations is
discuss these things, right?

(06:19):
And again, for me, it startedas a kind of counterintuitive
thing where I was like, "Is thatright?
That can't be right." And Ithink that my initial reaction
to it my initial hesitancy toadopt that one came from my
involvement in technology, justbecause working in this space,

(06:41):
it feels like there's so muchnew stuff coming out and things
become obsolete so quickly.
And yet, there are some thingsaround that when you start
looking around, it's like, "Hey,that thing's been around for a
while." Right?
Like the operating system inour computers, like the Mac

(07:02):
operating system, that's aversion of Unix.
And that thing is about as oldas you and I are, right?

Harry (07:13):
Yeah, it's not a spring chicken.

Jorge (07:16):
No.
So, even in the world oftechnology where things change
so quickly, there are thesethings that do manage to stand
the test of time.
And his assertion there, theway that I understood it anyway,
is that if you're looking tomake a decision between two

(07:39):
kinds of systems, it behoovesyou to look at the one that has
some track record.
I'm smiling 'cause that thewords came out of my mouth and
it sounded like, this is a kindof obvious thing to say.
But again, I feel like I needto keep reminding myself in this

(08:01):
time of so much change that,"Hey, there are things that have
stood the test of time."

Harry (08:08):
The other thing too is like...
your comment about software, Igrew up on Unix systems, right?
Literally, that's what I had,was using Unix on a PDP 11 in
college and then my first jobwas with a company called the
Santa Cruz Operation that putlicensed AT&T System 5 Unix and
put it on PCs.
It was called Xenix, and thatwas before Linux.

(08:32):
But I was thinking howimportant this is right now
because we are entering the ageof disposable software, and the
types of things you can buildusing a Cursor, Lovable, or any
of these sort of codedevelopment accelerators is
likely to produce something thatis remarkably clever and very

(08:57):
useful, but not necessarilyterribly durable.
And there's a very good chancethat you can express the idea
quickly and turn it intosomething that has utility.
Will it be around very long?
Probably not.
It hasn't necessarily been wellarchitected.
It hasn't gone through theannealing process of people

(09:25):
thinking about how it should notjust function, but how it
should work under the covers andhang together so that it's
scalable, extensible, usable,durable, flexible, and so on and
so forth.
So it just seems like a verytimely topic.

Jorge (09:45):
That's why I wanted to bring it to our conversation,
because I've been thinking aboutthis myself as I build out
capabilities, workflows, right?
Obviously, I'm experimenting alot with AI, and in my work with
Greg, we're helping businessesuse this to make their work

(10:11):
better.
And so, I'm working with thesethings day in and day out.

Harry (10:16):
And this is your company, unfinishe?

Jorge (10:19):
Yeah.
And one of the things that isalways on the back of my mind
is, all of these big players inthat space are relatively new
organizations that haven't beenaround for that long, right?
Yes, they're in partnershipwith the Microsofts and the

(10:39):
Oracles of the world, which havebeen around for a long time.
But OpenAI and Anthropic andthese organizations, they're
fairly new, right?
They don't have the trackrecord.
And you're building out allthese workflows that integrate
these tools that arecloud-based, meaning you don't

(11:02):
have control over thetechnology, they're not part of
your infrastructure.
There's somebody else'sinfrastructure that you're
using, temporarily.
And, it creates thissituation....
and maybe this is not just acharacteristic of the AI

(11:24):
companies, it comes with thenature of cloud-based software.
It creates this situation whereyour organization's
infrastructure is being builtout in a kind of ephemeral
model.

(11:45):
It like it could go awaytomorrow, right?

Harry (11:51):
Yeah, there's nothing very solid about it and it's
turning the whole process intoan organic maintenance event
rather than as a, design, build,deploy, maintain event.
And, I have computer serversunder my desk that have software

(12:17):
that was designed and built in2001.
And I can connect the twoservers and flip the switch and
turn it on and connect it to amonitor, assuming I can find a
monitor converter.
And all that stuff runs today.
It doesn't have to go out tothe internet.
And it all works still,shockingly, But if I were to
build something today, theprobability that I could go wire

(12:41):
them up and flip a switch andturn it on in ten or fifteen or
twenty years is gotta be closeto zero.

Jorge (12:50):
That's why I wanted to emphasize the distinction that
Taleb makes between perishableand durable things.
Because as the world has movedto information you and I, with
our background in informationarchitecture, I think that we're
especially prone to see theworld through this lens, that

(13:14):
information has become centralto business.
It has been for a long time,but it's become even more so,
right?
And there are businesses thatare nothing but information,
right?
Like, their infrastructureconsists of information
technology, basically.
And it's very different tobuild a business around physical

(13:36):
plant than it is to build abusiness around information
systems.
And one of the ways in which itis different is that because of
the nature of how informationtechnology advances and the pace
at which it advances, it doesfeel like it's more of a
perishable good than a durablegood.
Right?

Harry (13:55):
I love the idea of being more intentional about
classifying things as eitherintentionally perishable or
intentionally durable.
And in whichever category youso choose, that in theory,
should determine an approach tohow you're going to relate to

(14:18):
the design, architecture, build,dependencies, deployment, and
maintenance of that over time.
It feels more analog right now.
It's like people aren't, atleast in my experience, people
aren't asking that question,"Hey, is this gonna be a
temporary thing or not?" They'regoing ahead and building out...

(14:41):
they're talking to their LLMand they're generating some code
and they're, "Look, it works.
It's so cool!" And what they'vejust done is they've built a
perishable thing, but they'rethinking about it as a
potentially permanent thing.
And intentionally transitioningit from a disposable piece of
functionality to a morepersistent, durable piece of

(15:03):
technology then has to becomepart of our engagement process.

Jorge (15:09):
So, here's an idea, and I'm gonna say this because this
can get pretty abstract anddetached from how this might
help someone gain traction, andrepeating what we said in the
last episode, I'm gonna try toslap the snow chains on this
one.

Harry (15:26):
Yeah, okay.

Jorge (15:27):
So maybe one way to break this impasse, this idea that,
"Oh my gosh, is everything goingto be perishable here because
we're working with these leadingedge technologies?" Maybe you
could make a distinction betweenthe business you're building

(15:52):
and the means through which thebusiness operates.
Maybe the latter is perishableand the former is durable.
I'll cite one of my favorites,the Walt Disney Company, right?
They started that in the 1920s,so the technology that Walt and

(16:13):
his brother were using backthen is very different from the
technology that the DisneyCompany uses today.
Walt isn't around.
His brother isn't aroundanymore.
The company is a going concern.
The technology....
if they had stuck with thetechnology that they had a
hundred years ago, they wouldn'tbe around today.

(16:35):
But you can't last a hundredyears if you don't have the
vision that the thing thatyou're making is a going
concern.
I think it's very important forpeople to differentiate the,
durability of the means versusthe durability of the ends, If
the end is to build somethingthat lasts, you can employ

(16:58):
perishable means to do that.
In fact, today, for somebusinesses if you're in
information technology, youmight.
want to invest in things thatare perishable.
But you have to do it inservice to durability.

Harry (17:15):
I think you're onto something here, and where my
brain went with it was there hasto be a better set of questions
that we would ask, because thatsimple, whether it's on one
side, are we building a businessor on the other, are we
building the means?
Doesn't go far enough.
Because you have to ask thequestion, how long does the
investment, either theinvestment in money or the

(17:36):
investment in resources, theinvestment in effort, how long
does that need to persist andhow much continual investment in
keeping it up and running is.
There's gotta be some kind ofreturn on effort or return on
investment model that allows youto step back and answer the
question, where is it on thisline?

(17:59):
Because if it's all the wayover to the business side, then
one would say, in theory, youshould never talk to an LLM
about the code the applicationthat you want unless you are
very intentionally prototypingsomething that you want to
demonstrate and you areintending to re-architect it,
throw it away, and buildsomething up that's gonna have

(18:19):
some persistence to it overtime.
Versus the other side where,the means is to perhaps express
some innovation and we have thisbusiness and we're trying to
climb these s-curves ofinnovation, we're trying to
explore and experiment.
And in order to do that, maybewe say, the investment in time,

(18:42):
energy, and money doesn't needto last very long, but we have
to convey the idea quickly.
And so, the question of howlong does this thing need to
last or how long will it last, Ithink those questions become
really critical.
Look, I co-founded a companycalled Virtual Vineyards, it

(19:03):
became wine.com, and we wererunning on our prototype...
our prototype became productioncode.
It wasn't that prototype wasintended to be production code.
Amazon did the same thing,right?
Their initial system wasrunning as production code for
quite a long time before theywere able to peel it all back

(19:24):
and rebuild the systems to makethem, scalable, extensible,
performant, and all that goodstuff.
So, job one is getting clearabout, is this about building
the business versus are youbuilding the means to express or
support that business.
I think job two isunderstanding how much payback

(19:47):
over what period of time do youneed.

Jorge (19:49):
Again, I feel like now more than ever, we are at risk
of losing sight of the fact thatdurability matters.
And yes, it's a very excitingtime to be working.
There's a lot of new stuffhappening.
But very little of it, if any,has stood the test of time.

(20:16):
And we need to be mindful ofwhat we're doing when we're
working in this space.
That's my take on it.

Harry (20:25):
Yeah, and I think from a traction point of view,
architects and senior-levelpeople are gonna have
perspectives on this stuff thatmay be quite valuable here.
We may have to rely onexperience and judgment in order
to know where to draw the line,and just because somebody is

(20:46):
incredibly smart and facile withthe new tools and brilliant at
putting something together thatdoes something magical and
amazing, it doesn't necessarilymean that thing goes into
production right away withouthaving passed a series of
conversations that would allowus to know whether or not we're

(21:07):
making prudent decisions.

Jorge (21:10):
Yeah.
I think the key there, you usedthe word architect.
The key here is, when making adecision about a component of a
system, a vendor, a product,time is an important variable to

(21:30):
consider, right?
It's a determinant.
And when you're architecting asystem, if you know that there
are going to be components inthe system that have had less
time to prove themselves, thenyou can architect the system as
a whole to accommodate possiblechanges in those parts, right?

(21:50):
In aircraft engineering,redundancy is an important,
principle, right?
So it's something like that.
Consider which parts of thesystem might be at greater risk
of not being around just becauseof the fact that they're new.
It's not because they're bad,it's just because of the fact
that they're untested, unproven,and design redundancies that

(22:14):
allow the system the ur-thing tobe durable, to stand the test
of time.

Harry (22:23):
I think your pointer to risk here is probably a really
worthy topic.
Understanding what the risksare and the implications of
those risks and how they mightcome into conflict with the
goals that we have is a criticalmissing piece of information, I
think, in a lot of thesediscussions.

Jorge (22:42):
In our last conversation, we talked about how being aware
that there are stages to changehelps you manage change.
So knowing that there aredifferent facets to risk can
help you mitigate risk.

Harry (22:56):
Yeah, I think that's right.

Jorge (22:57):
That's the gist.

Harry (22:59):
It's a good place to stop, I think.
And at some point in the futurewe should go head on, talk
about risk management.

Jorge (23:08):
Absolutely.
When this podcast has beenaround for another twenty-five
years, and people will know, "Ohmy gosh, yeah, this has really
stood the test of time!"

Harry (23:19):
Yeah.
Sounds like every episode weget out there makes it more and
more likely that it'll be outthere twenty-five years from
now.

Jorge (23:24):
That's exactly right, except for the fact that both
you and I are perishable.

Harry (23:29):
And it's all running on cloud technology.

Jorge (23:33):
Alright, well awesome sharing with you again, Harry.

Harry (23:36):
Thank you so much.
Great reading.

Narrator (23:38):
Thank you for listening to Traction Heroes
with Harry Max and Jorge Arango.
Check out the show notes attractionheroes.com and if you
enjoyed the show, please leaveus a rating in Apple's podcasts
app.
Thanks.