TEC Talks Podcast: Jordan Andersen
Ep. 07

TEC Talks Podcast: Jordan Andersen

Phoenix, Arizona

Episode description

Today, Chuck speaks with Jordan Andersen, co-director of Hermit Tech, whose team members are responsible for a few recent viral articles which can be read on one of their co-director Nikhil’s blog here.

Topics include the typical TEC Talks format with a few essential questions about what it’s like to be in Jordan’s shoes, and then Chuck and Jordan have a fun discussion about how AI can often get overhyped, overpromised, and oversold.

Download transcript (.srt)
0:00

You're listening to Tech Talks, a podcast by the Technology Education Collaborative.

0:04

Tech is an Arizona nonprofit that empowers people with useful information

0:07

about the technology they use every day.

0:23

Today on our podcast, we have Jordan Anderson coming in from Australia.

0:28

Is that right?

0:29

Yeah, yeah.

0:30

Awesome.

0:30

Thanks so much for coming on to our show today.

0:33

So today we are going to go through our usual format where we ask Jordan a

0:37

series of questions like we normally do on Tech Talks, and then after that,

0:41

we're going to have an interesting conversation about different mindsets we

0:45

can use when we are writing software.

0:47

And I am very excited about that conversation because just a few minutes

0:50

ago, Jordan and I were having a little casual chitchat and we already

0:53

found a lot of common ground.

0:55

Without further ado, I'm going to give Jordan a chance to talk about

0:58

himself here for a few minutes.

0:59

I'll give him these questions and we will learn more about him.

1:02

But first off, thank you, Jordan, for coming onto the show today.

1:05

Yeah, thanks so much, Chuck.

1:06

I definitely share your excitement.

1:08

Really looking forward to this chat.

1:10

Awesome.

1:10

Yes.

1:11

The first question that we got here for you is what is your title or position?

1:15

I'm a director at Hermit Technology and Hermit Tech is a tech consultancy

1:20

firm and our MO is that everything we do is with radical ethics.

1:26

Uh, and so what that means, um, is that, you know, we won't bull anyone

1:30

that we chat with or honest, and we could get more into kind of what

1:33

that looks like and examples, but yeah, I'm on a team, there are six of us.

1:37

We have expertise ranging from data engineering, right through

1:40

the data science and analytics.

1:42

That's kind of my role.

1:44

And my game is to go and help people who need, you know, data and database

1:48

solutions to all their fun stuff.

1:49

Thanks for explaining that.

1:50

I really love that ideology, that mindset, just keeping it very true.

1:54

And to the point, no nonsense, right?

1:57

A little bit more on to our second question here.

1:59

What do you do on a daily basis?

2:00

Just like any consultancy, you know, we go out and look for leads, uh, so

2:04

that, so that we can find our clients and do some work, really what that

2:07

looks like is just meeting people, talk to them and finding out what

2:11

they're paying points are.

2:12

And it's, it's funny.

2:13

We've, uh, we've talked a lot about, uh, how we're going to run the company.

2:18

Uh, we have a completely flat structure.

2:20

All six of us are directors.

2:22

So we trust one another with decisions and, you know, you don't have to check

2:26

with anyone because we all believe that we're going to act with each other's

2:30

best interests and the company's best interests within our ethos.

2:34

So, yeah, it isn't like we go around cold calling with all of that, you know, on,

2:38

on a daily basis, I, uh, I focus on writing about my experiences.

2:44

Chuck and I got connected through, uh, one of my co-directors, uh, Nick

2:47

Suresh, uh, one of his blogs recently went viral and was all over Hacker News.

2:53

Uh, and that's just kind of the stuff that we do is that, you know, we write

2:56

about our experiences and share our, our honest opinions and just don't, you know,

3:00

fluff around and don't hide, uh, you know, how we feel and criticisms.

3:05

Um, but really kind of that's, that's the way that we get the message out.

3:09

Uh, I guess you can call it marketing, uh, about what we do.

3:12

Um, so that's kind of a big chunk we're doing today.

3:14

I'm currently contracting for a government agency.

3:17

And so, you know, my nine to five right now is actually filled up with providing

3:22

kind of solutions architecture services.

3:25

They've kind of scaled out of, uh, requirements, not out of choice with

3:29

their data and some of their products they're offering to, uh, different local

3:33

governments and so, uh, my role there right now, most of my day is, uh, setting

3:39

up cloud pipelines and cleaning up some of their, uh, data science projects

3:44

that they have a few of their models.

3:46

And I'm doing a bunch of, uh, data modeling to, uh, to move them out of

3:50

kind of a, I guess, individual data mart type approach with their database

3:54

to, uh, to a dim fact model.

3:56

So yeah, very technical in the kind of, uh, what I do with during the day.

4:00

Um, and it's, you know, it's all part and parcel with keeping my skills sharp.

4:04

Uh, but also, you know, those are kind of the services that, that Hermitech offers.

4:08

Our next question, what is your favorite thing about your job?

4:11

Yeah, the, this sounds cheesy, but like, like helping people.

4:14

Uh, and so going in and, uh, you know, the most exciting thing that I've

4:18

had in the tech industry is going in and understanding an organization's

4:22

challenges, problems, pain points, and then just like the massive amount of

4:27

dopamine that I get of, of helping them with that and solving it, you know,

4:31

through, through being a nerd and sitting down, writing code for a black screen.

4:36

Yeah, it just, I get, I get such a drive out of making things better so that an

4:42

organization can, can do the thing, you know what I mean?

4:45

Uh, so, you know, I come from, uh, recently I worked in healthcare in,

4:50

in the data team, they're developing different products, uh, mostly on a

4:54

database, my big role was making data more available to emergency department

5:00

teams so that they could get a better understanding of what they're doing

5:04

and how they're doing it.

5:05

And so I had a great opportunity to work directly with emergency department

5:09

nurses, doctors, and data teams.

5:11

And it was, it was just such a cool role that I, that I sat in for that period

5:15

of time, because like, I just, you know, I stood on the floor when things go

5:19

crazy in the middle of the day in the emergency department and, uh, just so I

5:23

can understand, you know, what they're, what they're actually dealing with and

5:26

why they need the data and what those, those, you know, visualizations really

5:31

need to look like so that it helps them do their job.

5:34

Uh, so yeah, that, you know, again, in a very cheesy way, I just love helping

5:38

people with the super technical side of things that we do in tech.

5:41

I'm kind of the same vein, you know, I write software because I love it.

5:45

And because I just, I get a dopamine rush from solving problems that make my life

5:49

better in turn, you know what I mean?

5:50

And if I can do that for other people too, that's a win.

5:52

Totally.

5:53

And I think that's a big thing about, you know, what we do in, in software

5:56

engineering, data engineering, the whole bit, right?

5:59

Like, like we get that big rush of dopamine and serotonin and, you know,

6:03

in a way that it's funny that that's the selfish side of the work that we do.

6:07

Right.

6:08

And, but yeah, no, it's, um, it's really cool when you can do that.

6:11

And you're not just another monkey, you know, fixing a bug that someone created

6:16

because it was, you know, just poorly designed, it's, uh, yeah, really pushing

6:20

things forward is, is such a fun way to get paid, right?

6:25

Absolutely.

6:25

Yeah.

6:25

And you know, the job isn't always fun, right?

6:28

It's not always the dopamine rushes.

6:30

And in fact, there there's, there's days where probably things are rough and

6:33

that's kind of a good segue into the next question, which is what is your

6:36

least favorite thing about your job?

6:38

If you, if you have any least favorite thing, or if you don't really have any

6:41

complaints, what are things you think could be better?

6:43

Oh, Jack, don't worry.

6:45

I got some complaints and I touched on it, you know, it's, um, one of the, I

6:50

mean, just like any other developer, it sucks to have to debug someone else's

6:55

code that's, you know, either was poorly designed, rushed to put together, you

7:02

know, whatever the reasons were, but, you know, going through somebody's

7:04

uncommented, just elephant graveyard of, of lines and lines and stuff like that.

7:10

I think, you know, all of us share that disdain.

7:13

I'd actually probably have to point to the thing that I dislike the most is

7:17

kind of why we started Hermit Tech.

7:19

And it's just the systematic waste of time, money, and other resources, and

7:27

that other resources, the most precious ones, you know, just the souls of people

7:32

in tech shops, right, I'm sure that, you know, all of us have had that experience

7:37

where you work somewhere where a manager is breathing down their neck or going on

7:42

and on about burn velocity and sprints and, you know, how many story points we're

7:48

getting through, and, you know, that becomes a focus instead of doing good

7:52

technical work with high fidelity and strong data integrity, right, when, when

7:58

the ladder takes a backseat, because we just got to get it out in the next two

8:02

weeks sprint, that type of philosophy and pressure, right, and the frustration that

8:07

comes along with it.

8:09

Like I said, that that's really where Hermit Tech, I guess was the impetus to

8:13

actually get us all together and doing the thing that we do is that we've

8:17

experienced all that. And again, a bunch of us have experienced the great sides of

8:22

being a developer. But it's just the frustration of watching these

8:25

organizations getting sold on the worst ideas ever, you know, one of the big

8:30

companies would come in working with KPMG Deloitte or any of the other big

8:36

firms is that there's this excitement of a project and strategy. And then when the

8:40

rubber hits the road, you get, you know, a bunch of kids who are 18 months out of

8:44

graduating college, that you know, your organization is forking over thousands of

8:49

dollars a day for this thing that just doesn't have any governance or control.

8:55

And, and so then as in-house talent, you're stuck with this just absolute

9:00

heaped pile of something, right? And so that that's absolutely the most

9:08

frustrating part, because then, you know, again, the in-house talent is, is

9:12

demanded that they make this thing fixed, fitting a, you know, a square piece into a

9:17

round hole. It's just that type of this interaction and transaction throughout the

9:23

tech industry, and not just tech industry, like across industries, but with tech and

9:28

digital assets. That's, you know, really why we came together to be a bit of a

9:34

positive beacon to stop that. So I mentioned that, you know, we're, our

9:38

ethos is that, you know, we're radically ethical. And, you know, a big thing that we

9:42

do is that, you know, often we advise clients that they shouldn't do the thing,

9:48

you know, they contact us like, Oh, you know, we have this idea, this project, for

9:52

example, you know, the Red Hot thing, the past three years ago, we want to do a cloud

9:56

migration, we got to get to the cloud, we got to get up there, right? We got to, we

9:59

got to be serverless. And, you know, the first question we always ask is like, Well,

10:04

do you really need to, would it be more effective for your organization, if you

10:08

just, you know, reorganize your data, did it in Postgres, posted it locally, because

10:13

you kind of already have the infrastructure sitting in a basement somewhere? Like, would

10:19

that not just be a better solution to do the thing that your organization wants to do?

10:23

And so, yeah, that's, to tie back to your question. Yeah, that's the most frustrating

10:30

thing. And, and the six of us, we want to do things better, you know, and do better

10:35

by people who use technology, and data to really achieve why they started this

10:42

business that they do, be it in healthcare, or transportation, or, or, you know,

10:48

agriculture. So yeah, that's the long winded answer of what makes me frustrated.

10:55

I definitely hear a lot of sentiments that I share in your response to that. One

11:00

exercise I always take people through is what I call right sizing, where I say to

11:04

them, Hey, let's start with the most bare bones, cheapest single entity possible,

11:09

right? A single server, a computer, right? Maybe even a Raspberry Pi, if you wanted

11:13

to get really crazy with this. Let's start with a really small thing that's cheap

11:17

and easy and answer the question, why won't this thing solve our, our business

11:21

needs? Right? What constraints are we going to run into first, such that we

11:25

actually need to go to the cloud, or, you know, you know what I mean, when I say

11:29

cloud, something more nebulous than just simply one server in the cloud, something

11:32

like an AWS distributed architecture or whatever is what I'm kind of getting at

11:35

when I say the cloud here. But you kind of get the gist, right? So if when we do

11:39

a right sizing exercise, we identify the areas in which we maybe won't have

11:42

enough enough disk space, maybe won't have enough RAM, maybe we won't have

11:46

enough uptime or high availability. But because we're thinking about this from

11:50

the things one server can no longer do for us, we're actually able to get down

11:54

and figure out what the customer really, really needs. Instead of them just

11:58

assuming they have to go to the cloud, they have to spend $5,000 a month on a

12:01

single Redis cluster, they have to do all these other crazy things that just

12:04

really aren't necessary to get the job done. Because they're not operating at

12:07

the scale that they think that they are going to need to operate at.

12:09

Yeah, totally. I mean, the thing that you mentioned there is that, you know,

12:13

that that right sizing is a wedge, right to get in, so that you can really get

12:19

into that. Well, what is this company doing? What do you guys actually want to

12:23

do? Why did you reach out to us? And I think, you know, having that philosophy

12:28

is so much more effective, because we can all just say like, Oh, you want to go

12:32

to the cloud? No problem. I'm going to set up your, you know, DBT will stop it.

12:37

No, we'll use Matillion. And then we'll get you up in AWS and use DBT. Oh, you

12:42

definitely need Snowflake, right? And you just just terminology bomb them with all

12:47

of the fads of today with with products. But no, you're totally right, Chuck, like

12:52

that, that that thing of like, hey, what will actually help your organization? And

12:59

why are you bringing us in as consultants as an external group, right? Like, what

13:04

are we actually trying to do here, team? And, and let's work together towards that.

13:07

And it's such a more fulfilling thing to do. And, you know, I've got a big

13:12

philosophy that it that it helps the tech industry as a whole to do things

13:18

better, right? Because like, I think we're getting, you know, talking

13:21

different C suite people, you know, we're moving more towards a distrust in tech in

13:28

general, not just consultants, but like, ah, you know, those software engineers are

13:32

taking so long. And, you know, that's been a bit of the rhetoric, always, but I

13:36

think it's, it's starting to slide and get a bit worse in, you know, the mid to

13:40

big size corporations, particularly you find in Australia, the government,

13:44

right? Like, there's a lot of skepticism. And so, you know, we have to not not sell

13:51

harder, but like, you know, there's a lot more conversation and you got to say the

13:57

right words to gain trust. And it's just been a weird shift over the past, you

14:03

know, bit of time.

14:05

So the next question that we've got queued up here for you is what is one

14:08

practical thing you'd want someone who is considering pursuing your job or your

14:11

industry, whatever it is you do, what would you want them to know? And I know

14:16

that we've kind of touched on this and maybe a little bit of a ways like kind

14:19

of what feeds into your philosophies and your mindset and all that stuff to

14:23

someone who is hearing this, what would you tell them if they wanted to get

14:26

started? Let's say that there may be a college graduate or maybe just still in

14:29

college, maybe even in high school. What would you say to them such that they

14:32

could be more prepared for getting into this career path?

14:35

Learn to use Git, version control, that's that like, and I start there,

14:43

because it's, it's technical, therefore, you know, us nerds, we can, I could do

14:48

that, I can go out and read Stack Overflow and read a book and like,

14:51

seriously, pick up Pro Git and just like, read the first five chapters and just

14:55

start a little project. Like, honestly, like, just just do that. Learn Git

15:00

version control properly. That's like the technical side of what people can do on

15:04

a more, I guess, broader principle approach. I'm borrowing this from, from

15:10

Nick, my co director, like, just take people out to coffee, like, talk to

15:15

people, just have a conversation and learn from other people. And, and, you

15:20

know, that type of approach, you just, you open up and unlock your ability to

15:25

learn so much more effectively, and not just about tech stuff, right, like,

15:29

about what challenges people have in the world. And I'm feeling, sounding

15:34

hippie-ish, but like, when you start with that, and you get used to learning

15:39

from people around you, in your industry and out of, like, you look down

15:44

the road, and it just sets you up, you know, in different roles that, you

15:49

know, all of a sudden, you're having a conversation with someone who does

15:51

hiring, or you get a referral, or, you know, when it starts to look like

15:55

that, you find your way in front of something, you know, a CEO, and then

15:59

you can have different conversations and have a bigger impact in the

16:02

industry. I mean, from a consulting point of view, like, you asked, you

16:07

know, what does my day to day looks like? Yeah, like, seriously, a lot of

16:10

it is, is asking interesting people, if I can buy them a coffee, you know,

16:14

from there, you start, if you want to look at a business and marketing

16:17

point of view, like, that's the way that you can, quote unquote, you

16:20

know, find leads, or, you know, start building your business, but it's

16:24

really just like, talking to people, what about their problems, you know,

16:28

what are you interested in? So and I say that, again, typically in the

16:32

tech industry, you know, we, we just want to sit and write in front of a

16:36

computer, right, we want to write code, we don't want to talk to people.

16:38

But you know, it's get out of that comfort zone. And you might, you might

16:42

find a passion in what someone else is doing. And then, like we spoke

16:46

about, find a way that you can help someone else. And you know, heaven

16:49

forbid, get paid for that. So yeah, those are kind of the two things,

16:53

learn version control, get really good at Git. And then just just talk to

16:58

people in and out of the industry about tech stuff, because it's

17:01

fascinating.

17:02

Regarding Git, I also agree that it's one of the most foundational things

17:05

to learn nowadays. It's a little weird at first, right. I mean, I also

17:08

was kind of confused with it over, you know, 1015 years ago, I don't

17:11

remember exactly when I started using Git, I think the thing that made it

17:14

click for me was remotes, and how a repository is basically just something

17:19

that can exist across different remotes. And once you learn that,

17:22

like, oh, there's GitHub, and then there's GitLab. And then there's,

17:25

you know, get tea and all these different source control hosts. But

17:28

underneath it all, it's just a Git thing that you're pointing your

17:31

repository to. And that's just the remote, something, I don't know, I

17:34

just made it made it click for me. And I really hope that other people

17:37

can can put in the time to, and most importantly, the patience to learn

17:41

it, because it's really not so bad once you get used to it. But I have

17:44

met quite a few people over the years who are still kind of struggling

17:46

with with the basics on it. And I do my best to try and explain it to

17:49

them, because I acknowledge it as a little unusual at first.

17:53

Yeah, totally. And like, look, just the I mentioned before, like, learning

17:58

a new thing, any new thing, just get a project, get a pet project that

18:01

you want to tinker around with, and, and do that. Or, you know, if, if

18:06

anyone's listening, and they're, you know, let's say, a few years into

18:09

the industry, and you're thinking like, I don't have time, you know, I

18:12

don't, I don't want to be from my computer at night or on weekends, I

18:16

just do it at work, right? Like, convince your IT department to allow

18:20

you to install Git and just whatever you're working on now, just start

18:26

version controlling it, right? That's just the best way to learn a new

18:30

skill. Because I totally agree with you that, you know, I've worked with

18:34

people who've been in the industry for 15 years. And, you know, in a

18:38

way, I empathize with them, you know, if you got kids or grandkids, and

18:43

you're looking at it like, I don't want to learn a new thing, I know

18:45

how to write SQL, and I get paid, so why would I do anything different?

18:49

And so, you know, there's, there's going to be a few barriers to that

18:52

toward the end. But yeah, like it, like, if you want to learn, just

18:55

start doing it.

18:56

Those are the main questions. The we've got one final question here

18:58

for you. What would you like to share with those who are listening

19:02

currently? What are you working on? What passion projects? What are you

19:06

involved in? Anything you want to share? I know, we've discussed a

19:08

little bit about your consultancy, your firm, and some of the ethos

19:11

behind that. But is there anything else you'd like to share?

19:14

In terms of the practical stuff? Chuck and I, we spoke before we

19:18

started recording about Hetzner and, you know, exploring different

19:23

platforms, different tools that are not, you know, the big three that

19:27

are not Azure, AWS, and GCP. And so we're playing around with a few

19:33

things there. We're looking to develop our own CRM that's, again,

19:39

hosted in a bit more of a financially feasible way, be

19:44

really light with it, and put together a product that we use

19:48

internally, you know, borrowing kind of from from base camp

19:52

approach, right, that, you know, we have pain points. And so we're

19:55

all in tech, we've got, you know, back end and front end

19:58

experience. So why don't we just make something that works for us

20:01

and see where that goes. So that's kind of on a product side.

20:04

And then we're also big on developing processes. And so, you

20:09

know, I hate the corporate term, you know, ways of working, but

20:12

really, like, you know, I'm up in in the Gold Coast area in, in

20:16

Australia, and the rest of the team is down in Melbourne. And

20:19

that's, I gotta check my geography, but like, 2000

20:23

kilometers, so but yeah, so you know, we're separated. And then

20:25

in Melbourne, all five of them, they're also living in different

20:29

areas. And so, you know, it's really like, how do we be

20:32

productive with the mostly async dev cycles, and every now and

20:37

again, we get together in pair program. And we found that, you

20:40

know, pair programming in two, maybe three, where you have an

20:44

observer being the third. But yeah, no more than that, if we

20:48

actually want to tackle tackle something technical and deep,

20:51

we've realized that we got to, you know, talk about it, break

20:53

it, break the task down into chunks, and then split off into

20:56

into teams and pairs and tackle it. So we were formalizing, you

21:01

know, how does Hermit Tech work? Because, you know, it works

21:05

really well for us that we've, you know, developed a few other

21:08

products and, and solved problems for clients being, you

21:13

know, super async. So, you know, can we put together some sort of

21:17

a cookbook that other people might benefit from? So yeah,

21:21

that's that. Those are kind of the three big things that, that

21:25

we talk about that we work on.

21:26

Awesome. Okay. Well, that wraps up this portion of our

21:31

discussion here. Thank you, Jordan, for going through our

21:34

questions. Jordan and I are going to have a second part of

21:37

this conversation where we dive a little bit into a little bit

21:41

more of the mindset and philosophies of good software

21:44

development principles, at least as we see it, based on an

21:48

article that went popular that we mentioned earlier called I

21:51

will drop kick you if you use that spreadsheet. It's a little

21:54

more aggressive than I made it sound, that's for sure. And it's

21:56

a great read. So now we'll go ahead and move on with the next

21:59

portion. I got to pull this article up. Do you have it up?

22:02

So it depends on which one you're talking about the I will

22:05

drop kick you if you use that spreadsheet or use spreadsheet

22:09

again, that was an older one that Nick wrote. But it's still

22:13

got very a lot of positive feedback. The most recent one

22:17

that dropped at the end of June was I will compile drive you if

22:21

you mention AI one more time or something along those lines.

22:25

Yeah, it's so good. And like, we laugh about it. But yeah, so

22:32

no, those are those are kind of the two. So a quick rundown,

22:36

like the spreadsheet one was more around like data

22:38

engineering, like, stop shoving everything into, you know, five

22:42

megabyte size Excel files and like do things properly. And,

22:46

and what does that look like? And a lot of like what we spoke

22:49

about in the first segment that like, stop adding to tech debt,

22:52

just because your stupid sprint burndown chart looks better,

22:56

right? Like, let's do things properly. So that's kind of that

22:59

one. And then the I will drive you if you mentioned AI, like

23:03

that was God, it was so good. It's just like the culmination

23:06

of conversations we've had over the last year of just CEOs

23:11

getting absolutely sold on bullshit, right? And just crap

23:15

products that, you know, just because you use AI, it doesn't

23:20

mean that you're doing anything useful other, you know, you're

23:23

just wasting money. So it's a big conversation on like, like,

23:27

yeah, more directed towards a C-suite, like, what are we

23:30

actually doing? Like, and when somebody says these types of

23:33

things around AI, like, when you do like a fucking pip install,

23:37

like chat, open AI, like deploy chat GPT as a Gen AI, like,

23:43

you're not doing anything, right? Like, it's just but like

23:46

a CEO can put that on their on their CV and then, and then, you

23:50

know, fade off into the distance and gather a bonus. And so it's

23:53

the criticism of that type of stuff. It's like, what are we

23:56

actually doing?

23:57

Yeah, I got to be honest, it's it's absolutely all over the

24:00

place. Everyone thinks they can just plug this thing in, and

24:03

suddenly their application can do anything. I even had this

24:05

this thought for a little while at first, I wanted to create

24:08

something that would summarize your Git commits, right. And so

24:10

I thought, well, hey, this thing's a large language model,

24:12

right? So I should be able to pass in a diff patch of all the

24:16

the commits I did on a certain, you know, basis, and it should

24:19

be able to understand this. And so I passed it through like

24:21

three or four different production, large language

24:25

models, like Claude, GPT, four, and I think a few variants of

24:30

Claude. And I got mixed results from all of them. I even tried

24:33

to be very specific, like only respond. Exactly. Like only

24:36

respond with like a rating of one to 10 on how productive this

24:39

commit was. And like, if I'm over here, someone who is kind

24:43

of in the industry and is writing software thinking this

24:45

stuff, it's no wonder everyone else is also thinking this

24:49

stuff, because it seems like and it's advertised as this thing

24:52

that can do anything and everything. Right. And why are

24:55

we buying into that? Well, because it's new, because it's

24:57

novel, because it's very cool to see something being able to

25:00

distill the sum of human knowledge into a little prompt

25:03

that you can interact with. It's, it's undoubtedly undoubtedly

25:06

cool. But at the same time, I think people think, or are

25:09

rather being told that it can solve all your problems. What

25:13

are your thoughts on that? Is that kind of your perspective

25:14

as well?

25:15

Oh, yeah, 100%. Right. Like, exactly. Like you just

25:18

described, like this, this gen AI thing, like, ah, we're

25:22

gonna get AI and it's gonna solve our problems. But no, no,

25:25

you're totally right, man. And I totally agree. Like, oh, God,

25:29

Nick and I, we were at a conference last year. And it was

25:33

a tech conference. And, you know, it was really cool

25:37

networking opportunities. And so during one of the events, it was

25:42

actually Nick, he got chatting with this group of women from an

25:45

e commerce company. And I like somebody somewhere had sold them

25:51

on the this idea. And they like carried it, right? Because they

25:55

were, they were great at what they do. We talked about

25:57

marketing, when we got when we sat down with them, we're

26:00

talking about their marketing, and they're just so good at

26:03

selling stuff online, and like, what they need to consider and

26:06

taking care of their customer, but they were so good at that

26:09

stuff. But you don't have to be a software data engineer data

26:13

science to do that, right. And so somebody had sold them on

26:18

this idea that, you know, AI is going to solve all of their

26:22

technical problems. So the first the first interaction, one of

26:27

these women had met, had met Nick, and Nick was like, yeah,

26:30

you know, data science by trade, and this is hermitag. And so we

26:33

do and, and she was like, Oh, great, like, we're, we are going

26:39

to get AI to set up and do our analytics, it is going to solve

26:47

our analytics problems. And we're just going to get AI to do

26:51

it, and it's going to be great. And like, you totally empathize

26:55

with this, this woman who's so good in, you know, kind of the

26:59

tech industry, because they are e commerce, but you empathize

27:02

with her so much be like, Oh, who lied to you? Who brought up

27:07

the smoke and mirrors? Right. And so it was, it was, yeah, it

27:11

was fun. That was on like, day one. And so like, he met, Nick

27:14

met, I guess, one of their like, high high level managers, maybe

27:19

she was in the C suite. And then like, eventually, like the two

27:22

of us got introduced to their team of three. And Nick was

27:25

like, look, like, we just need to sit down to lunch. And like,

27:28

we gotta have a chat. And so like, we, we right there very

27:33

easily could have just sold them on like, yeah, definitely, we

27:38

will set up a gen AI product package, you know, use some fancy

27:43

terms that, you know, they didn't know that all they needed

27:46

to do is import, you know, open AI. But so like, we, like, we

27:51

could have just done the snake oil and sold them on whatever.

27:55

And then we would have made money off that. Um, but like,

28:00

I've been saying, like, hermit tech, we're, we're radically

28:03

ethical, like, we just, we're not gonna lie. And so Nick and I

28:06

sat down to lunch with with this team. And we just had a

28:10

conversation, we're like, look, this is what kind of data is

28:13

generally what data science is, you know, like, this is what

28:16

supervised model training looks like you do this unsupervised

28:19

model training thing. And you know, that's, those are just

28:23

kind of two models. And, and so like, we're just describing

28:27

like, you need to know something about your thing, for

28:31

a data science model to be successful, like, otherwise,

28:36

you're just going to get stuff, right? Like, you can do all the

28:39

design, the architecture, you can push that into it, you're

28:41

going to get an output, but it's just going to be, you know, a

28:45

mound of dirt, right? Like, it's going to give you numbers, but

28:49

do they mean anything? And so we just kind of put a quick

28:52

tutorial and just talk them through. And, and, um, you know,

28:56

you and I spoke on our last conversation about really

29:00

understanding a company and organizations challenges the

29:04

pain points. And so we just asked him, like, what are you

29:07

actually doing? And like black and white, just like a lot of

29:11

other organizations, that they had analytics set up, they had

29:14

some dashboards data being pushed into, like a power bi or

29:18

tableau or whatever it was, but like, they just they weren't

29:22

confident that having all that set up is was advising them how

29:27

to do sales better, right? And like, and like that black and

29:29

white, that's what ecommerce is, right? Like, you know, you sell

29:32

more of your product, you become more profitable. And they just

29:37

they weren't confident that their current setup is doing

29:39

that. And so like, very quickly, like, it's not an AI solution

29:44

you need, right? Like, you don't need to go through all that,

29:47

like, what that is, is just better data engineering. And,

29:51

and, you know, they were, you know, small to medium sized

29:54

startup. And, and, uh, so, so they, I gathered, they didn't

29:58

have a big tech presence in their staff to push those

30:03

agendas. And that's fine, right? It's just the reality. And so

30:07

when they wanted to shift more into that, of course, like

30:10

everything that's on Twitter and all the other platforms and in

30:13

the news, like AI is going to solve everything. And so it was

30:17

just logical that they came to that conclusion. And, and, you

30:21

know, that was a moment of, of, as our, our tech firm that I

30:26

feel as though that was a success, right? We didn't make a

30:28

penny off that organization, but that wasn't the point that we

30:33

really felt like we helped those people. And then the kind of, I

30:38

guess, highest ranking manager, it sounds like she's in armed

30:43

forces, but no, like they're, they're, um, you know, they're

30:46

manager of the, of the three people we spoke to, she's great.

30:49

She's like, awesome, I'm going to go online. I'm going to learn

30:52

a little bit more about the analytical side, um, see what we

30:56

can do with what we already have in house. And it was just like a

30:59

really positive experience and interaction that, you know, we

31:04

kept in touch with them and ensure that she, you know, went

31:07

on to learn a little bit more about how to do analytics with

31:09

what they already have. And so like circling back that, geez,

31:13

like, yeah, you know, people get sold on this idea that AI is

31:18

going to solve all the world's problems and it's going to steal

31:21

our jobs. Like, you know, those of us who know any programming

31:25

language know that's just not true, but unfortunately the

31:29

people who hold the purse strings and who are in charge of

31:31

layoffs and who are in charge of hiring or make the decisions to

31:35

go out and get one of those really big firms that will come

31:39

in and promise the world, and then send you someone who's

31:42

fresh out of college to do their gen AI agenda and their

31:47

technical strategy. Like that, that's the big problem, right?

31:51

Those people that make those type of decisions in

31:53

organizations just, just aren't aware of, of the reality of it.

31:58

And I think that is a big power of, of that article that, that

32:03

my colleague Nick Suresh, um, put out the, you know, I will

32:07

pile drive you if you, if you mentioned AI one more time, is

32:11

that like, we just hope that that message, the reality of

32:15

things starts to reach the desks and ears and eyes of people who

32:20

are making decisions in our industry, because I think that's

32:22

where we're going to start, you know, getting a bit of a change

32:25

away from this red hot thing that is AI, um, into something

32:31

that's actually going to help an organization, you know, achieve

32:35

their mission statement and move towards their goals.

32:37

Awesome thoughts. And I I'm going to ask a question here

32:41

that is a little devil's advocate sort of, um, I want us

32:45

to put ourselves in the shoes of these folks that kind of have to

32:48

say these things in a certain regard. And to expand upon that

32:51

a little more, if you are a CEO of a company, at least in the

32:54

U S and you're publicly traded, you are legally responsible for

32:58

ensuring that money comes into the company at all costs, right?

33:02

You're responsible for the growth in accordance with the

33:04

shareholders needs. I personally struggle with this a

33:07

lot, so I'm not expecting any or any perfect answer rather. How

33:09

do we reconcile the legal demand for a CEO to, to bring in as

33:14

much money as possible with the simple reality that the stuff

33:18

isn't as magical as it is being made to sound. I mean, isn't,

33:20

isn't a CEO sort of obligated in a certain sense to lie basically

33:24

is, is kind of how the system is set up. I mean, shouldn't they

33:28

be trying to extract the money at just the right thin, fine

33:31

line, such that consumers are happy enough and everyone is

33:35

happy enough, even though they are indeed being lied to. I

33:39

mean, I, I don't like it, but that's kind of how it's built.

33:41

Would you agree with that?

33:42

Well, you bring up a really good point of line that you had in

33:45

there, you know, that like, isn't it the responsibility to

33:48

CEO to lie to everyone and like, at the surface it's, and not

33:53

just at the surface, you know, I've been in organizations that

33:56

like people who just bought into that, right? Like, like the

34:00

movie, don't look up, like it's a lot of just, don't talk about

34:05

the thing, right? We don't talk about Bruno here. He's kind of

34:08

thing. Um, I, I see it more, you know, people are compelled to

34:14

lie. Um, and to oversimplify it, you kind of, you know, how, how

34:20

these CEOs in two different camps, right? You have the ones

34:23

who are aware and they know, and, and again, you know, what I

34:29

agree with you, what you said is that I feel they are compelled

34:32

to, to, um, shape their message that maybe bends the truth or

34:37

maybe, you know, full of lie. And, and it's a hard place to

34:40

be in, you know, like I got a lot of empathy for CEOs of those

34:44

kind of publicly traded mid-sized companies, a lot of

34:47

pressure. And you're right. Like, right. Like if they don't

34:49

make money, like people's livelihood, it's not even just

34:53

within their own company, but people, you know, bought shares.

34:56

Um, there's a lot of, you know, paying the mortgage and putting

35:00

food on the table, uh, type things at risk for a lot of

35:03

people. And so like, I got a lot of empathy. Those are, those

35:05

things are hard. Um, so like we'll park that. Yeah. It's

35:08

really hard, right? Yeah. And so like we'll park that that's kind

35:11

of like, let's say one oversimplification of, of our, our

35:15

CEO. Um, but then the other one is just like the scorpion,

35:20

right? Like the, the snake oil selling, um, you know, uh,

35:24

smiling teeth type, take CEO that will gladly lie, you know,

35:29

and, and ride out there two to five years as CEO as company,

35:33

um, bring in the AI, make it happen, tell everyone, go and

35:37

press conferences and then, and then jump on their, on their

35:41

boat and sail into the sunset before the house of cards comes

35:46

tumbling down. Right. And, and like, I've seen it before I

35:51

joined Hermit Tech. And then definitely, you know, a big, a

35:54

big offering we have is, um, uh, to be PG about it. We unscrew

36:00

the, uh, migration, right? Like we, we, we go in and we fix the

36:06

mound of stuff, a mound of garbage that's been left by

36:09

previous C-suite and previous consultants. And, you know, we

36:13

go in and we're like, all right, well, like we can dismantle this

36:15

and, and fix it for your organization. So you don't go

36:18

under, but yeah, like there's, there's definitely that second

36:21

archetype of just like, great, do it, do the Gen I, Gen AI, you

36:26

know, import open AI, make the thing happen. We won't tell

36:30

anyone that it takes a competent engineer two weeks to set

36:32

everything up, including documentation, but we're going

36:35

to roll it out. We're going to bring up the red carpet. It's

36:37

going to look great. We're going to cut the ribbons. Um, I need

36:40

to take the bonus and, you know, um, take off it's, it's, you

36:45

know, that type of thing. Um, again, we spoke in our last

36:48

conversation about how Hermit Tech, we do, uh, radical things,

36:54

radically ethically, like we just not interested in working

36:57

with organizations that have that latter flavor. Right. Um,

37:02

and so to, to circle back to your devil's advocate question,

37:05

like, yeah, definitely it's hard, but I think that, you

37:07

know, you can, in those two situations, you can make money

37:10

for your stakeholders again, to oversimplify in one of two

37:14

ways. One, you can sell the snake oil and you can, you know,

37:17

get a big consultancy firm in that will send juniors to dump

37:22

whatever they can do and walk away or the former CEO who's

37:27

aware and does have to play the game. You know, you can,

37:31

hopefully that CEO has a better understanding of the needs of

37:35

their organization. And that's the person we like sitting down

37:39

with to really understand their pain points and, and get a full

37:43

appreciation of what the business is and how we can help

37:47

their bottom line by, you know, automating stuff to remove

37:51

manual work so that those same staff members, they don't get

37:55

laid off their time and brain powers freed up to do more

38:00

effective things. Like a client that we were talking to, they

38:05

they'd highlighted that they do so much manual transfer, like

38:09

copying rows and columns from this spreadsheet into that one

38:13

and then push it into that software. Then that software

38:17

churns and spit something out into another CSV and they go in

38:20

and select specific rows and then they move it into the next

38:23

one. And so like all of that, yeah, let's, we can automate

38:27

that, you know, Perpetek will go in and automate that so that

38:30

those people that do the, you know, control C control V,

38:33

instead of doing that, they can be refocused on their analytical

38:37

work and do the creative stuff so that that company can serve

38:42

their clients more effectively with what they actually do,

38:45

right? Like they don't like their mo isn't to copy and

38:49

paste. Their mo is to provide deep insights that are valuable

38:54

to their clients from the data that they receive. And so that,

38:59

again, going back to your question, like, well, how do you

39:01

make money? How are you responsible to stakeholders and

39:04

being publicly traded? Like, you can go that former route. It's

39:08

not easy. And there are definitely pressures that the C

39:11

suite and, you know, co founders go through that are

39:15

like enormous. And part of why I'm not interested in going that

39:19

leadership path, you know, but like, there is a way there. And

39:24

we've worked with awesome companies that see that and, and

39:28

have been successful of improving how things work inside

39:32

by doing it, you know, properly.

39:35

Love it. Yeah. So just to kind of summarize, right, that the

39:38

approach is controlled by scope, right? You make the

39:41

choice of whether you're going to choose to work with a company

39:44

that is acting in a manner that they're clearly intending to

39:47

deceive or not. And you seek out those who prefer to operate in

39:52

reality, right? And what I mean by that is the approach that you

39:56

and hermit tech take is that you're going to connect real

39:59

principles, what actually matters, what can be

40:01

accomplished with what the business is trying to do, right?

40:05

And you're trying to dissuade them from hearing all this other

40:08

nonsense. That is the snake oil peddler. That is the, the

40:10

corporate exec that is lying to them just to boost their

40:13

shareholder value, right? So you are grounded in reality, and

40:16

that is your approach. And one would really, really, truly hope

40:19

that a grounded in reality approach would scale in time, no

40:24

matter what company you're at, right? Because eventually the

40:27

rubber is going to meet the road. If you are, say, I don't

40:29

know, Tesla, and you're lying about how good your, your cars

40:32

are, eventually one would hope that that truth would catch up

40:35

to you. Do you know what I mean? So that's kind of the, the

40:38

thing that I'm hearing from, from you guys is that you want

40:41

to really keep the scope focused so that you're within the

40:43

bounds of reality. Does that kind of catch the essence of

40:47

how you guys look at things?

40:48

I think that, um, people can choose the reality. And so

40:53

truly getting like some people's reality is consciously rung

40:57

consciously is to peddle that snake oil. But I think I think

41:01

the big thing is, you know, having the opportunity to choose

41:04

reality and choose your direction requires a good,

41:08

thorough, deep understanding of what your organization, what

41:11

your company does. Right. And I think that, you know, that work

41:15

and, you know, you can say, like, oh, you got to stay in

41:18

touch, you got blah, blah, blah. Like, you really need to get an

41:22

understanding of where, you know, down to your core, down

41:26

to the core of the company, where does this shit want to

41:28

sail? Where should it sail? And then within that, what do your

41:32

crew members want out of this? You know, and I think like, you

41:36

know, a good, a good CEO is going to have a real good feel

41:40

on all of those levels. And I'm like, we're talking about mid

41:43

size organizations, like, that is impossible when you start

41:47

pushing over 1000 plus employees, right? Like that

41:50

becomes very hard for the leader to get in touch with everyone on

41:53

the floor. But yeah, that's probably more the thing, you

41:57

know, again, being in touch with what your organization wants to

42:00

do, then you can start choosing your, your reality.

42:06

Jordan, is there anything else that you wanted to add in? Like,

42:09

do you want to share any, any interesting conversations with

42:11

people that you've had, or perhaps Nick has had since these

42:15

articles went relatively popular here on the internet or anything

42:18

else you'd like to share with us before we close up this

42:20

conversation?

42:21

Yeah, definitely. Since Nick, one of my co directors, since his

42:27

writing has gone so far, especially most recent ones,

42:31

we've realized how many people are experiencing the same things

42:37

that is in Nick's writings and reflection of that things that

42:41

we talked about at Hermit Tech and, and our previous

42:43

experiences, like it's wild, how many people have reached out and

42:49

said, thank you for writing these things. Now, I can point

42:54

people to a blog, instead of me, you know, telling them how

43:01

things should be, right. And, and so it's just that referral

43:04

to, to gain support for doing things in the right way. It's

43:10

been wild how many people are in a similar position. And, and

43:15

that's where we feel that the way that we're doing things at

43:19

Hermit Tech can be successful as our own company, you know, for

43:23

the six of us to move forward and make a living out of this

43:25

thing, but also that through people out there that we can

43:29

help that, that we align with in those things. So, so it gets

43:34

like, you know, for people listening that, if you're

43:37

frustrated at work with, you know, those burned down charts

43:42

and getting things done by the next sprint, or, you know, if,

43:46

if you're the manager, and, you know, there is pressure to make

43:50

sure your burned down chart looks good and, and all those

43:54

things, but like the way that you're doing things, you feel

43:57

like there's too much, just spinning your wheels. There are

44:01

a lot of people out there in all different levels from the

44:04

C-suite down to, you know, your, your software engineers, never

44:08

in between, that, that share those frustrations. And, and,

44:11

you know, I think the best way forward is to connect with one

44:16

another. A great way is, is through reading blogs and making

44:21

comments and, and sharing your opinions on these things, or

44:24

writing your own blogs, putting your own thoughts out there,

44:27

because there's so many people that we're just, because of the

44:31

tech age, you know, all of us are sitting in our own home

44:33

office or in a cafe or sitting on our phone on the subway,

44:37

having these same thoughts, and we're so physically

44:40

disconnected, like we can connect with one another. And

44:43

this is how, you know, Chuckie and I are having this

44:45

conversation. So I guess, I guess like a parting message or

44:49

remarks is that like, get out there proverbially on the

44:53

internet, read stuff, write your own stuff, publish it, put it

44:57

on, even if you want to do it on LinkedIn, whatever, like, get

45:00

your thoughts out there. If you feel like they're gonna piss

45:04

someone off, that's okay. Because if it's frustrating to

45:08

you, there are enough people in this world, that somebody else

45:12

probably has a similar opinion. And you know, if you're moving

45:15

towards a better way of working in tech, that's less

45:18

frustrating and less soul sucking. Like, I think it's a

45:22

good thing to be mentioning or explaining how frustrated you

45:27

are about these different things, because I think there

45:29

become to an understanding. And then we can start making a plan

45:32

of doing things better in our tech industry.

45:35

Well said, did you guys get any hate mail for anything that you

45:38

guys have posted recently? Like the, the AI post, I'm guessing

45:41

had some interesting responses, I imagine?

45:46

Just like, go on Hacker News, have a look for the blog, I'm

45:51

trying to find the best way to find it. Anyway, like, check if

45:54

you linked to the blog, like have a look, the comment section

45:57

is just awesome on fire. And even at the top of his blog, he's

46:02

got posts of like reviews, and there's a link to all the

46:06

reviews, and he just grabs like a bunch of the hate mail and he

46:09

posts it and says, really, it's so good. It's amazing. So yes,

46:16

if you want to if you want to good chuckle, or if you hate all

46:19

of these ideas, and you want to be around people that have the

46:22

same deal as you like, don't worry, like Nick has posted a

46:26

link to them on his blog. So cool.

46:31

Well, thank you so much, Jordan, for your time. It's really been

46:33

a pleasure talking to you. It's really been awesome hearing

46:35

about hermit tech and how you guys do things really absolutely

46:39

I'm privileged to have been able to talk to you today. So thank

46:42

you so much for your time.

46:46

Usually this podcast is recorded at the Advanced Cyber Systems

46:49

Lab located at the Washington Street campus of Gateway

46:52

Community College. The AC SL is a tech hub open to the general

46:56

public where you can 3d print useful items, test your tech

46:58

skills out on enterprise grade servers, record a podcast and

47:02

learn to use Linux all free of charge. If you don't know how

47:04

to do any of that or what it means, that's okay. Stop in

47:07

anyway. And one of the friendly folks here will be more than

47:10

happy to teach you how