David Morgan-Gumm (00:00)
Hey everyone, and welcome back to SQL Squared, where we chat about all things data development and the cool tech shaping our world. We love bringing on awesome folks to share their stories and insights. And speaking of awesome folks, today we've got a really special guest in the house and someone very special to me, Mike Hodson. Mike is a strong tech leader and has a real passion for data. Thank you so much for joining me, Mike.

Mike Hodson (00:07)
Yes.

What a great intro. Thank you, David. When I have to hire you to introduce me everywhere. Pleasure to see you again as always. And thank you very much for having me on. I'm looking forward to the chat.

David Morgan-Gumm (00:23)
Hahaha.

Yeah.

It's been a while since we saw each other last in person ⁓ playing snooker in Liverpool.

Mike Hodson (00:39)
It has that frame is still going on. I believe that the mammoth team game of snooker that we have. Yeah, it's. Yeah, but yeah, it's good to see you again. Hopefully this will probably be quicker than that frame of snooker. Turn that to be.

David Morgan-Gumm (00:41)
Hahaha

Yeah, yeah, That's really fun.

Yeah. Yeah. Yeah.

Yeah. Yeah. Maybe we'll kind of get it cut down within an hour at least. All right. Today we're going to dive into Mike's journey, how he fell in love with data and we'll get into some really practical stuff like Scrum and Agile. We'll also talk about how data is being used in recruitment. The latest with AI in data and development. And of course what Mike's most excited about.

Mike Hodson (00:58)
You

David Morgan-Gumm (01:23)
about the future of data as we always go into. It's gonna be a great chat. So thanks again for joining me on the podcast, Mike. Do you wanna go into a bit about yourself, your career journey and your passion for data, I suppose?

Mike Hodson (01:40)
Yeah, if I go for career journey then you obviously know some of this. Yeah, so I'll explain it to you as if I've never met you before. So yeah, career journey, that's true, that is true. So yeah, career journey, if I start from end of education, official education, maths degree.

David Morgan-Gumm (01:44)
Yeah, go for it. Start off with the career journey.

Well, maybe most of the audience won't have met you, so go for it.

Yeah, go for it.

Mike Hodson (02:09)
maths, stats and operational research was the degree I got because I just enjoyed as I'll touch on probably all that kind of stuff up to that point in my life numbers, data, stats, all that stuff so just enjoyed it had a bit of a natural talent for it so just kept doing it basically until I ran out of road and had to get a job and left uni with a maths degree like I think a lot of people other than going and being a maths teacher

David Morgan-Gumm (02:30)
Ha ha.

Mike Hodson (02:37)
Wasn't really sure what to do with it. And I didn't want to be a math teacher just because, ⁓ I suppose similar to other people's experiences of math teachers in general, nothing against them, but I just didn't want to, didn't want to do that and didn't feel it was me. So got a job as a data manager, not too far away from where I lived. And it looks interesting, something new to learn. So sequel. ⁓ but you could use some of the.

David Morgan-Gumm (02:55)
Mm-hmm.

Mike Hodson (03:06)
set theory, things like that from maths that crossed over. It was still stats, numbers, interesting stuff to me. So I thought, right, I'll try that for a bit. I'll try it for about six months and then, you know, maybe use it as a springboard to go elsewhere. I remember being six months in and thinking, right, I'm six months in now. It's time for the springboard. Something like 18 years later, I'm still there and still, still enjoying it. It's never boring.

David Morgan-Gumm (03:29)
Yeah.

Mike Hodson (03:34)
There's it's probably the job. the data is always changing AI, et cetera. There's always something new around the corner and recruitment is always changing. So the two together just make it an interesting place to work. it's gone.

David Morgan-Gumm (03:51)
does. I feel the same. I think what I've got a lot less time in the job than you have, ⁓ still it seems like yesterday when you started. ⁓ And my journey was very, very similar. I didn't really want to go into data, to be honest, when I left university. But yeah, I am.

Mike Hodson (03:55)
Thank ⁓

⁓ that surprises me.

David Morgan-Gumm (04:17)
fell into the role and then grew to love it and never went back. But yeah, kind of, yeah, it's, you get, lucky, suppose, falling into an industry that you really have a passion in and kind of things that you wanna, you know, you wanna help improve and stuff that, you know, technology that excites you and keeps your job interesting.

Mike Hodson (04:41)
I agree. I think we're lucky to be in this industry at this time, working with data. It's an exciting time, isn't it? Whichever way you look at it. I don't think anyone knows exactly what is going to be going on over the next few years, but it's going to be exciting to find out. you can't pretend it's not there.

David Morgan-Gumm (04:45)
Mm Yeah.

No? ⁓

It will be.

No. And I think things are going to, yeah, kind of skyrocket forward and data will always be really important. Cause we always need systems and processes that manage the information that runs our world. And that's kind of where the whole idea of SQL square came from, you know, looking forward and seeing that, you know, we're going to need people in these, in these roles and, you know, younger generations, get them interested and learning the.

what it means to be a developer in the data space. Cause it's very different to being a developer in the software space. you know, ⁓ working on websites or, you know, UI and UX functionality, which a lot of people fall into, especially after doing like computer science degrees and things like that.

Mike Hodson (05:34)
Okay.

Yeah, so a different role and a different kind of way of finding it, think, usually, isn't it? So the people I know that are in data haven't come in through that way, really. The way that you have, I suppose. You know, the computer sciences route. Yeah.

David Morgan-Gumm (06:00)
Yeah, you kind of just fall

into it. I suppose if you've got a love for numbers and statistics like yourself, it's a natural stepping stone.

Mike Hodson (06:02)
But it's a.

Yeah, I think I would have always ended up doing something in data. So when I was little, um, my dad was an accountant. So I was always sort of surrounded by numbers are important, whether it's financial or, or whatever. Played darts a lot when I was little as well. So I was in, learn how to do times tables and averages and all that stuff re really young. So, um, and then just carried it on. So it's, sort of got.

David Morgan-Gumm (06:31)
here.

Mike Hodson (06:38)
probably carried away and obsessed with, with numbers and what you can do with them. then now, like you said, data is just everywhere. Isn't it? If you just keep your eyes and your ears open, there's always data to be collected and there's always processes that can be optimized or improved. I think you're on a safe bet with a podcast on data. Future proof. ⁓

David Morgan-Gumm (06:40)
You

Mm-hmm.

Yeah, yeah, that's good. That's good. Keeps me in the job.

what about what about current day then? So what are you what you're doing at the moment? What's your job title? What your responsibilities?

Mike Hodson (07:11)
Yeah, good question. ⁓ so that came in at the current place as a data manager, which was really, was, it was a good learning experience, but it was, it was running around putting it, putting fires out really. So it was, we were, yeah, we were kind of very, very low resource IT team trying to do everything. So it was a great learning experience, but it was very

David Morgan-Gumm (07:28)
Yeah, fire fighting.

Mike Hodson (07:38)
very much in at the deep end. So it was learn the recruitment process and how that works and the industry, and also learn how all this other stuff works. We, we were doing the DBA role, the, I suppose the developer role and the infrastructure bit. there was, you know, nights of rebooting servers and all that stuff and wrestling with AD and taking computers apart.

David Morgan-Gumm (07:40)
Yeah.

Mike Hodson (08:08)
and all that stuff. So I've had quite a good introduction to it all from every angle whilst everything's on fire. So it kind of stands you in good stead for everything else that follows, I would say. You've got a good base.

David Morgan-Gumm (08:12)
Yeah. Yeah.

agree. It's important. I think, especially in the age of AI now, which, which is, you can, you can save pros and cons, especially for junior people getting into the space. it's a, it's a lot more protected. You're not, well, it depends, suppose, on what business, what businesses you go into, but, you know, for people starting off in their career, I can't imagine it's as

Mike Hodson (08:29)
you.

David Morgan-Gumm (08:47)
fast paced or as problematic, ⁓ you know, things necessary aren't as urgent as they would have been in the past, because they've got systems to manage it. ⁓ But I think, yeah, it kind of does, gives you a good foundation, sets you up strongly for, you know, future career development.

Mike Hodson (09:12)
It did, it did with me definitely.

I had that as the base level one kind of thing and then progressed to the SQL developer. So a bit more proactive and a different role, more coding and less, less things on fire, slightly less things on fire. then SQL team lead. So then line management responsibilities as well, where you get good enough at a role that you then put in charge of the people who do that role.

but you're not doing a different role. that, that thing, the hybrid thing, then Scrum Master is where I've ended up. So there was a gap in our project management. What we were finding was we were having really good days and weeks, and then getting to the end of the month and realizing we'd not actually added any value. We'd not really delivered anything. And we didn't really know why for a long time. And it was because there was a gap between what we were doing and what the business.

David Morgan-Gumm (09:41)
Yeah,

Yeah.

Mike Hodson (10:08)
needed us to do. So I just kind of got sucked into that gap, was interested in it, found Agile and thought, right, this is it. This is, can see how this is going to work and then fast forward a few more years and yeah, certified Scrum Master with the Scrum Alliance. So I've, I've gone very, leaned into it heavily.

David Morgan-Gumm (10:09)
Mm-hmm.

a lot of the basics and foundations about agile development and scrum from you. And, you know, I've taken it forward into future roles and implemented it just as you did really. Um, maybe with a, maybe with a few tweaks based on the team, cause it's always got to be, um, you know, geared towards the team that you work in and the people that are in it and things like that. can't just kind of take a blank template approach and slap it down. And you've got to make changes.

Mike Hodson (10:49)
Yeah.

David Morgan-Gumm (10:58)
But yeah.

Mike Hodson (10:58)
That's one of the things I like about it

as well is that it's pretty lightweight and adaptable. So I've not used the word agile there on purpose, but I agree. It's, ⁓ I like how it's not whack people over the head with a big textbook. You've got to do it like this. It's you can use common sense.

David Morgan-Gumm (11:03)
Yeah, flexible. ⁓

Yeah.

continuous improvement approach, the retrospectives, figuring out what you've done wrong, what you've done right, and just constantly, like a moving train along the track that you're constantly moving forward, building new carriages behind the train as you go on, might not be the best metaphor. yeah.

Mike Hodson (11:35)
It works for me. Yeah, I can see it. I

think ideally you'd be doing it with every area of your life, not just project. It just kind of makes sense to me. But unless you step back and look at what you're doing every now and again and how you can maybe do it better or make it more enjoyable for yourself, you just. You kind of like to go with your analogy, a train without a driver, you just just rattling along the tracks.

David Morgan-Gumm (11:46)
Yeah. Yeah.

Yeah, yeah, yeah,

yeah, with no control. well, I suppose just just for our listeners, could you briefly explain what Scrum is and why it's so beneficial in development?

Mike Hodson (12:04)
Mm.

Yes. So if I start with the beneficial bit, I think it leans into what humans are good at, I think. So it's always felt to me like you're going with the grain when you're scrum. And we've tried other ways of doing projects as well. So we, we did waterfall methods before.

David Morgan-Gumm (12:17)
Go for it.

Mike Hodson (12:40)
We found that jail and I found, found that more against the grain things that humans just aren't as good at in my experience, writing big, long documentation, reading big, documentation, interpreting other people's words correctly. You only need to see what happens on email chains to see that that can go wrong. Can it? Whereas with scrum, everything written down anywhere is basically just a sign point.

David Morgan-Gumm (12:59)
Yeah

Mike Hodson (13:09)
signpost for a conversation because humans interacting with each other is the best way of understanding what people mean and where you need to go. So until the robots take over, I'm still going to be saying humans working together is the best way of getting stuff done. Most fun, most straightforward, best solutions. So Scrum leans into all that stuff, I think.

David Morgan-Gumm (13:20)
Yeah.

Mike Hodson (13:37)
So that are the benefits, I would say.

David Morgan-Gumm (13:38)
That's

a great description of the benefits. I could not have said it better myself.

Mike Hodson (13:42)
in terms of what it is, so it's characterized by the quick version is you've got three roles in scrum. So it's straightforward. So you've got the product owner who is the voice of the business. So they're the ones responsible for the whole thing, actually adding value. If it's in the wrong, if you're chopping trees down, but it's in the wrong forest, they're the ones that are supposed to go. You're doing a really good job chopping the trees down, but it's in the wrong forest. It should be over there. that that's what they're doing.

David Morgan-Gumm (14:11)
Wrong type of wood.

Mike Hodson (14:11)
And then you've got, yeah, exactly. Yeah.

And if you're on say the dev team, so one of the other roles is the dev team. and they're the ones doing the work. in this analogy, I'll go with it now. These would be the ones chopping the wood and they wouldn't necessarily know if they're in the wrong forest. So they're the ones doing the work that the closest to it. And then you've got the scrum master is the other role. that's in, in my world, that's me. And that's the facilitator.

David Morgan-Gumm (14:26)
Yeah.

Mike Hodson (14:41)
So the one that keeps it all together and keeps it moving and keeps communication going. So, the scrum masters, one of the main roles is get rid of any impediments that anyone that's stopping anyone, whether it's the product owner or the team. If someone's got a, ⁓ an ax that's blunt scrum master would sort that out. Yeah, exactly. ⁓

David Morgan-Gumm (15:06)
shop at it. Yeah.

Mike Hodson (15:10)
The other thing to mention probably sprints are the other thing that characterizes scrum. So rather than you just keep working at something forever, or if you imagine a workload is like a big cliff face and you go, that's our work for the year. Rather than do that scrum has sprints. So instead of a marathon, you do lots of shorter sprints. So usually two weeks, but it doesn't have to be, we do two weeks for the two weeks. You chop off the cliff face.

David Morgan-Gumm (15:32)
Mm-hmm.

Mike Hodson (15:40)
small, a small amount of work that you can do in two weeks chopped up into little rocks. And then you just focus on them for two weeks and don't worry so much about the cliff face. And then after two weeks, you see what you've delivered, show it to the product owner. And then all together decide what you're to do next. And there's, I think there's loads of benefits to doing that. It helps people from getting burnt burnt out or bored because you're just doing a sprints are more fun than.

David Morgan-Gumm (16:04)
Mm-hmm.

Mike Hodson (16:09)
a one year marathon and it does things like limits risk to just the two weeks. If you've delivered completely the wrong thing and the product owner comes back from holiday and goes, that's terrible. It's only two weeks rather than you could have spent a year on the cliff face. So there's lots of benefits.

David Morgan-Gumm (16:25)
That's a lot

of people forget about with Agile, especially I think leaders. I've had multiple challenges in my current role. We don't have a Scrum Master, but we really try and push the development approach. So we kind of split the role between myself and two others. And we get a lot of complaints.

from leadership and management about delivery speed and sprints and kind of much they see getting done. It's like they don't want to see it until a month's time, if that makes sense. Like they want more to be done before they see it, but when we're working in the agile method, we want to make sure that we're minimizing risk and we're showing them as early on as possible.

Mike Hodson (17:09)
Yeah.

David Morgan-Gumm (17:21)
it's kind of to those that don't understand the process thought of as a little bit of a negative sometimes that you're not doing enough by the time they see it. And that's what it might be just a frame of reference sort of problem. But I think the point of minimizing risk within your development pipeline, especially when you're trying to run a high performing lean team.

is a very important point.

Mike Hodson (17:54)
⁓ I'm familiar with that problem that you just mentioned. one of stakeholders, product owners, not really wanting to see anything until it's done or until it's finished for them because they've got to do a bit of extra work, haven't they? If they need to stay, et cetera. So that, that is a, that is some kind of, that's a pushback that I am familiar with this thing of, think people feel like they would rather work in a waterfall way.

David Morgan-Gumm (17:56)
Yeah.

Mm.

Yeah.

Mm-hmm.

Mike Hodson (18:24)
where

they would, they would kind of give you the spec. It would be perfect. It'd be perfectly understood. You'd go away and design it. Well, you know, analyze it, design it, build it, test it, come back to them when it's completely done in say six months. And then they're assuming in the six months, the world's not changed. The environment's not changed. They're still in the same role. They still want the same thing. They feel exactly the same as they did six months ago. It's going to be delivered exactly to the, to the letter.

David Morgan-Gumm (18:35)
Mm-hmm.

Yeah.

Mike Hodson (18:53)
And then they're going to see it and go, yes, that's exactly what I had in my head. I don't think it's realistic. think you, the best way doing it is they stay along the way and you need opportunities presented to them to stay, which means they have to see it before it's completely finished. So you're right. It's, think it is a framing thing and a context thing that it's.

David Morgan-Gumm (18:56)
Yeah

I agree.

Mike Hodson (19:18)
They need, they still need to be involved just in a slightly different way than they're probably expecting. You do need time from stakeholders along the way to do agile because they're part of the team.

David Morgan-Gumm (19:27)
Mm-hmm.

And I suppose that it brings me onto some of the challenges when adopting agile, to be honest, because moving, you know, used to work together moving away from your team. didn't foresee really the of adopting proper agile development. I just kind of thought, you this is the way it should be done. You know, why would we not do it that way? And

Mike Hodson (19:52)
Mm-hmm.

David Morgan-Gumm (19:58)
when I moved to my new team, I didn't join as a senior or anything. I was just joined as a SQL developer. But I could see based on my experience with you and in your teams that they were doing things very inefficiently and didn't have all of the fundamental processes in place. Even though they were kind of trying to do Agile and Scrum, they really weren't doing it. So I tried to...

forcing a lot of change, a lot of improvement. And to be fair, over the four years since, it's totally transformed and we've got everything kind of built into the business and the business processes and stuff. ⁓ No, but it was challenge to begin. of, yeah, one of the questions I want to ask you is,

know of any of the biggest challenges teams face when they adopt an agile development process? And like, how do I suppose the developers, the seniors, the leaders within the tech teams overcome any of those challenges?

Mike Hodson (21:08)
Very good question. I feel for you having that experience of kind of feeling like you're banging your head against the wall. I think the winning hearts and minds thing is quite often the trickiest bit, isn't it? Rather than the technical bit, it's the people bit that can be harder. And I think it is hard to convince people that your thing that you're telling them.

David Morgan-Gumm (21:13)
Yeah.

Mike Hodson (21:37)
is going to be amazing is better than what they think is amazing. It's hard to change their belief without actually showing them. You can kind of tell them.

David Morgan-Gumm (21:43)
Mm hmm. Yes, what they used to, I

suppose, in some instances, if you especially if you're changing process within a team that's been together for five or more years, you know, where you're going in with the best of intentions, but it's you know, they used to a certain way of working.

Mike Hodson (22:02)
Yeah, if you just go in and go the way you're all doing it is wrong. I know a better way. It's just not going to go well. You need to build trust first, don't you? And then, and then say, how about doing it this way? So there's that, that whole thing, which you could probably do an hour on that. You know, the how to take people on a journey and build trust and then gently take them over to, how about if we did it this way? ⁓ and.

David Morgan-Gumm (22:06)
You make enemies. No, it's not.

Mm-hmm.

Mike Hodson (22:31)
Showing people that things work is always more effective than telling them. It's just getting them to the point where you can get enough leeway to try stuff and show them it working. It can be difficult. We've, I've experienced pushback from similar to what you said, people that just are used to work in a different way. So it can be a bit of ⁓ a shock to people when there aren't loads of big documents flying around.

David Morgan-Gumm (22:42)
Mm-hmm.

Mike Hodson (22:59)
that no one wants to read. It's rare, most people don't. But some people do enjoy writing big documents and reading them.

David Morgan-Gumm (23:01)
Yeah.

I hate it.

Mike Hodson (23:09)
So do I. So do I. Most people do. And I find that I think the people that lean towards that, so that enjoy writing those big documents, generally most of the other people don't enjoy that and therefore don't read them. So you're almost, it's almost like it's being written for the person writing it rather than the audience. So I, yeah, I think the pushback or the challenges are from

David Morgan-Gumm (23:10)
I it so much.

Mm-hmm.

Mike Hodson (23:38)
people used to work in other ways and they then think that their way is the only way. So I suppose similar to what you had really just in reverse.

David Morgan-Gumm (23:48)
I am going to sound like

a bit of a broken record, I think, to a few regular listeners, if there are any. ⁓ When it comes to digesting large documents, I recently wrote a blog post on this, and something that you could probably find very, useful is a tool called Notebook LM. It's a Google product ⁓ built to basically ⁓

take a load of large documentation, whatever you're doing, if it's about business process, technical stuff within your business, or if you're doing research for an exam and or you're a student at university, you can put in all of the context, all of the different material that you've got. you can basically ask it questions specifically about that material. So it doesn't, it will use the web, it's built on Gemini. So it will use the web for some bits.

if you've got additional questions, but it uses your material first and foremost to give you the answers to your questions. ⁓ And so I've tried it at work where I have basically taken an entire confluence space, uploaded it, and then I can just ask it questions about the business process and it gives it back.

research purposes, I've written a blog post on it because I've been doing a management qualification and I've been using it for this and it's been the entire, you know, revolutionized the way that I learn. ⁓ You can upload all of your material and you can get it to generate a podcast.

obviously blew my mind. Generator podcast takes about 10 minutes. The tool creates like a 30 minute podcast of a man and a woman talking about everything that you've put in.

Mike Hodson (25:44)
Wow.

David Morgan-Gumm (25:45)
found it, I found it insane.

Mike Hodson (25:46)
You'll have to try

an automated episode of this. Just to see.

David Morgan-Gumm (25:50)
I feel like people will hate me for doing that.

But yeah, I there's, definitely have, you know, some plans in the future with like shorts and things. I want to be able to upload some information, train it on my voice or whatever, and just kind of try and get a short out there. But yeah, it's been, it's been incredible. So if you want to learn more about Notebook LM, have a look at the blog post that I've written on the

on the SQL squared website and give it a try.

Mike Hodson (26:19)
I'll have a look. I'll check it out.

Yeah. I'm interested to know. ⁓ cause it sounds like you could probably affect some of that in AI. It not, you know, as in I'm thinking like chat GPT or something. You could probably upload all the information.

David Morgan-Gumm (26:36)
Yeah,

I think what it's what it's it's built So you can it within chat GPT, you can like upload a single document or upload a certain amount of document in context for that specific prompt. ⁓ But within notebook LM, you can upload a suite of information. You know, I uploaded, I think like 50 plus web pages and journals and things like that. All of different things. Different topics.

Mike Hodson (27:01)
Right.

David Morgan-Gumm (27:04)
it's just a chat interface like chat GPT and you can ask it questions about the documentation or ask it to summarize certain topics and areas of research, things like that. ⁓ And it supports all sorts of ⁓ documentation, whether it be web links so it can access web pages or just text that you've scraped, ⁓ books, journals, PowerPoints, things like that.

Mike Hodson (27:34)
Wow, that's got uses, hasn't it?

David Morgan-Gumm (27:37)
Yeah,

yeah, yeah, definitely. So yeah, give it, give it a try. But I've mentioned it on the podcast before, so I'll put it, I'll put a link to it in the description as well, and to the blog post. And yeah, if any of the listeners out there haven't used it, definitely, definitely give it a try. It's

Mike Hodson (27:39)
definitely have a look at that.

David Morgan-Gumm (27:53)
think we've been over like why working in an agile way is beneficial and a couple of challenges that people need to overcome. We both work in the recruitment industry. I think, yeah, you've worked in the recruitment industry your whole career and I have to this point as well. But obviously we're heavily in the data.

⁓ and the, ⁓ you know, what, what actually runs that sales journey and how we help people, how we help the business basically make money. it's, it's an area that I'm fascinated by, suppose, ⁓ you know, how, how it's being leveraged. so I kind want to know, you know, is there in, have you got any insights that you can provide to people?

on kind of how data is leveraged within the recruitment industry and kind of how it makes an actual difference to the way a business is run and the money that they make and the industries that they work in.

Mike Hodson (28:57)
So yeah, we've both been in it a while. And it's changed. It's changed a lot over the years. was very straightforward and simple in hindsight. How it used to be in terms of person put CV on job board goes into list. Consultant rings through candidates, CVs go to client, some result in interview, some result in placement. There's more going on.

David Morgan-Gumm (29:00)
Yeah.

Mm-hmm.

Mike Hodson (29:23)
It's, it's way more complicated, but I think like with all data, you need a human at the end to, take an action off the back of it for it to make a difference. So it's just really like, to your point pointing out areas where you could, there might be a bottleneck and it's like, if you just hire another person in this department, suddenly you're to get the throughput way, way better. So data can just point out things that.

You could make a tweak and it's going to be much better in terms of, yeah, flow through a process. So I'd say with recruitment, there's the, you're probably familiar with these kinds of concepts, but there's three things you need to be good at recruitment or to be successful in your process. You've got the, I'd call it probably activity. So as in.

David Morgan-Gumm (29:54)
Mm-hmm.

Mike Hodson (30:19)
If you imagine it was another analogy, going to the gym to try and get say bigger legs. If you just go once a month and do one rep, say like days once a month. Yeah. If you just go and do one rep, nothing else is going to happen is it doesn't matter what else you do.

David Morgan-Gumm (30:24)
Everyone hates like they.

You're describing my gym

journey right there.

Mike Hodson (30:38)
Yeah, unfortunately I've got experience with this as well. a painful example, but yeah, if you, if you only do say one rep of anything, it's not going to make a difference. So there's that, there's the activity level. You need people to be doing enough. Then there's the quality. So if you're doing, supposing you are going every day and you're doing a hundred leg presses, but they're all terrible and sloppy and not actually really doing what it, what it needs to do. You're not hitting the muscle in the right way.

David Morgan-Gumm (30:40)
Yeah, yeah.

Mike Hodson (31:05)
They're not going to give you the best result either. So the quality of what you're doing and the last one is the targeting. So again, you could be going and doing a hundred every day, but if it's say, if you're on the shoulder press instead of the light press, you're probably not going to get bigger legs. So, so that's, that's the gym lesson in recruitment. If you, if you swapped it for say calls or something like that, any, any activity that the recruiters need to do, then it's gotta be.

David Morgan-Gumm (31:25)
that analogy. That's so good.

Mm-hmm.

Mike Hodson (31:35)
Enough calls to make a difference. Quality calls. Cause if they're all terrible, it doesn't matter how many you do and the targeting, because if you're doing loads of calls and they're all amazing, it's you're ringing the wrong people or the wrong clients, then that's you're not going to get anywhere. So it needs those three things and data can help with all of them.

David Morgan-Gumm (31:49)
Mm-hmm.

Yeah.

Yeah. No, I agree. mean, I've worked with telephony data quite a lot. So it's a, it's a, it's good analogy that I understand. think, yeah, with, especially within a, within a sales focused business with the data that you get, you have to provide insight from that data. You know, the business has to gain some sort of insight for either improved efficiency or improved opportunities.

And then also, you know, telling the salespeople what they're doing well, what they're not doing well, what the effective processes that they're running and the activity that they're doing. ⁓ So, yeah, I don't know if that's the same for other sales based organizations outside of recruitment. I assume that a lot of the ⁓ a lot of it is shared. So don't know if there's any listeners out there that work.

I know. think of a sales-based organization. ⁓ But within those industries, if you've got any similar experiences, I'd love to hear from you. But yeah, mean, Insight, I think, is the main, ⁓ let's just say word of the day. Yeah, yeah. Yeah, yeah.

Mike Hodson (33:14)
It could be, it be a of the podcast.

Yeah, I agree. If you're not providing an insight and therefore like a call to action for someone that can do something about it, whatever it is, there's not really any point is that if you just look into data and not, not ending up with an insight, you kind of just look into data for the sake of it. But yeah, like you say, an insight could be team leader. Did you know that all your team are doing a high level of calls, but this one person isn't.

David Morgan-Gumm (33:25)
Yeah.

Mm. Yeah.

Mike Hodson (33:44)
They can then go and act on it, can't they? And the same with all the other stuff, the targeting or the quality as well. So you can get data on all that stuff easier and more readily available than ever now.

David Morgan-Gumm (33:46)
Yeah. Yeah.

Mm-hmm.

Yeah, there you can. There's lots of additional tools and processes, especially within the cloud, to make the most of your data and to more easily gather useful insights.

Mike Hodson (34:00)
Thank

Yeah, there's some companies that that's their whole thing, isn't it? Is in the whole recruitment process, this bit here, we've got that now. Don't worry. Just, just plug us in and we'll give you insights on your data in that area and it'll change your business and change your world.

David Morgan-Gumm (34:27)
Yeah.

Well, I suppose what in your experience over the last few years, what are some innovative ways that you've seen data used to optimize the recruitment process? It could be anything from candidate sourcing and CVs to actually making placements and sales.

Mike Hodson (34:50)
We've probably looked at similar things over the years, I imagine, me and you, but we go to the Recruitment Expo in London every year just to see what's going on. Unsurprisingly, it's getting more and more AI heavy over the years and therefore probably more exciting. there's stuff along the whole process that companies are now...

David Morgan-Gumm (35:09)
I could imagine.

Mike Hodson (35:19)
using data for insights and you just, you just have to plug them in. You just can't plug them all in is the problem. But yeah, the ones that spring to mind, there's things like there's. Woo. That do they're the, they do career tracking. It's done. If you've come across these, but career tracking. So plugs into your CRM and supposing I've placed a team of let's say.

David Morgan-Gumm (35:25)
Pick and choose.

Okay.

Mike Hodson (35:48)
SQL developers on the day at X this monitors, I assume via probably LinkedIn or similar when yeah, exactly. It must do. But when you move, so you move to why I would get alerted in my CRM and it would say, David's just moved to why maybe give him a ring. I give you a ring and say, hi, David. Are you growing a new team at the new place? And you can see how that's a

David Morgan-Gumm (35:49)
Mm-hmm.

Yeah, scrapes some sort of websites.

Mm.

Mike Hodson (36:17)
a warm call and

David Morgan-Gumm (36:18)
Yeah, opportunity analysis.

Mike Hodson (36:19)
Yeah, so that that's interesting. Like that didn't used to be there. Five years ago, as far as I can tell, there's stuff like there's quite a few from what I've seen, like quill that do the, we can take all the kind of donkey work away from the consultant for you. So just leave them doing the selling the bit that they're good at that we need the human to do anything that we don't need the human to do. We'll take off you. So that thing is like.

AI note taking, scheduling, all that stuff. So that the bit that recruiters probably aren't that good at and don't like doing.

David Morgan-Gumm (36:58)
Mm-hmm. Now I suppose, yeah, notes is a good thing to cover because at my company, we've recently introduced a new telephony system called Dialpad.

have, you know, automated note taking, you know, AI analysis of your calls and transcriptions, and then taking notes and putting them into the CRM against the person that you've called, for example. And so that takes away a lot of the effort of

track of the calls, the notes, action points off the back of them, things like that. But we have found some issues with it. Maybe it's not up to the standard that we want. Or again, context is a big thing. When you've got a siloed call, there's no context of everything else that's going on around it. I know context is becoming a huge thing within the AI and automation world.

don't know if you've used or heard of MCPs, context protocols. I've been doing a lot of research and work into them. I've even developed my own MCP server for SQL Server over the last few months in C-sharp. Yeah, so an MCP, it stands for model context protocol. So when you're...

Mike Hodson (38:02)
No.

So an MCP for the EMG.

David Morgan-Gumm (38:24)
running a custom model or one that you've, you know, that you pay for, say, if say it's copilot within, it's copilot for business or copilot for enterprise within Azure. Or you've got a model running through AI Foundry. You can spin up a server or Docker instance, for example, of what's called an MCP server. And that acts as a bridge or a gateway between your AI agent

and something else. So for example, I've done successful proof of concepts with SQL Server where ⁓ you can spin up ⁓ a model in AI Foundry that you can say put on your internal CRM sites, ask it questions and interface with your SQL Server environment in natural language. So you can say, who is the highest permanent biller of 2020 to four, for example. And there's a lot of there's a big

application that sits in a Docker container behind it that takes that information and translates it into, okay, what tables do I need to look at? What information do I need to look at? And it can learn from all of those processes as well.

Mike Hodson (39:36)
And could you just then give that front end to a CEO and not needed to do the bit in between?

David Morgan-Gumm (39:40)
Yeah, exactly. Yeah.

Yeah. I think there's a lot of work. Obviously, what I've just done is a proof of concept. it was literally like, or can I get access to it? Can I make it work? I think there's a lot of work needed to make it accurate and make it reliable for your business and a lot and your data and your data structures need to be in an AI ready state, which I can comfortably say ours are not at the moment.

Mike Hodson (40:08)
Yeah, prob-

David Morgan-Gumm (40:09)
There's a lot

of work that needs to go in behind it. It's not just magic.

Mike Hodson (40:12)
No, that's it. Unless they're in an AI ready state, you, kind of open to a mistake popping up, aren't you?

David Morgan-Gumm (40:19)
Yeah,

yeah, exactly. And if that process provides mistakes, especially to senior leadership, it can quickly become untrusted and have all of your work become invalid because they don't want to use it anymore. Yeah, it does.

Mike Hodson (40:36)
Yeah, yeah again, it's it means human check doesn't it really?

before unless you're to make someone look silly or Unless it's in a safe place Where you could you could accept a mistake, you know one in one in 20 or whatever it is

David Morgan-Gumm (40:49)
Yeah.

Yeah, exactly.

But I mean, if you it depends on, you know, those 20 questions that you may ask the AI, they're not all going to be of similar importance or urgency. And you know, you could get the point where you've got an important question that you're asking that has a big impact on the next move that you make. And it could be that one that returns ⁓ far from true.

Mike Hodson (41:06)
No.

David Morgan-Gumm (41:21)
answer or figure that you know you haven't catered So yeah, it's simple. It's not simple. It's it was simple enough to get set up, but it would be extremely hard to to master and get right. But there's things like one of a lot of developers out there will already know of these MCPs, but ⁓ GitHub have an MCP, for example. So I've

This massively improved my workflow efficiency. ⁓ The agent within Visual Studio Code. you can on your local Docker environments, you start set the Docker hub on your machine in, you know, pull the GitHub MCP from the GitHub repo, deploy it to the ⁓ Docker instance in your machine and your agent that runs on your machine.

with through Visual Studio code now has access to whatever GitHub repository you make available. And so it can do everything from creating branches, making commits, assigning work to people, creating issues, adding comments to issues, reviewing code that people have put on PRs, creating PRs, merging everything, the whole process end to end. And so my...

management workflow become much better and much more streamlined. know, documentation on issues that need to be there, PRs with full descriptions of exactly what's been implemented and why. Yeah, it's, you know, much better than I would be able to write myself and I don't have to write it in the first place and spend the time on it. So it's good.

Mike Hodson (43:07)
And do you still need to do the human chap? you still need to, can you trust it?

David Morgan-Gumm (43:12)
you still need to do the human check. One of the things that you can do, I mean, it's a bit outside of data, but say you're working on a, you know, in C sharp or Python or Node or whatever. If you have a unit test environment, you can get it create the unit tests as well to confirm that it works. So there's still a testing element to it.

can talk to you a lot more about it later because I'm working on my own project that I'm not going to announce at the moment because I want it to be something that I do in the future, which is kind of fully being built by these agent coding processes. Yeah, so I've done a lot of work and research on it over the last few months.

Mike Hodson (43:54)
I thought you may have something going on. Just, I'd be surprised if you didn't. Yeah. Yeah.

David Morgan-Gumm (43:55)
Yeah.

No, I need to stay ahead of the game.

one last thing, ⁓ that I want to talk about within the recruitment space. ⁓ obviously we all know that data security is a major thing, ⁓ within, ⁓ any organization, ⁓ especially people that sell to people and personal data and things like that. ⁓ are there any like ethical considerations or challenges that you've come across when managing data within recruitment and kind of how

How have you overcome those in the past?

Mike Hodson (44:33)
Well, as you say, it's people, isn't it? With recruitment. So it's personal data. So it's different than, you know, if you were, if you were selling something else. So yeah, there's the considerations and the challenges of probably GDPR related. So as in making it compliant and that the legal side of it. And then the other side of it is making it fair for the people, the actual people.

humans involved so that they know what you're collecting on them and why and what you're doing with it and that they're okay with it and they understand that there's something in it for them. So that whole thing really plus you're the insights you get from your data are only as good as what's going in. So if your data is a mess like you've alluded to, you're going to get a mess of insights. It might be nonsense.

So something that springs to mind is it's probably about five or six years ago now, but you Amazon had a, ⁓ an AI internal tool doing the recruiting process or at least part of it. and they just eventually they've been off because, because of its bias. So what they, what they've done was they trained it on 10 years worth of CVs and then just gone right off you go, go and go and do new.

David Morgan-Gumm (45:43)
no, I didn't know that. No.

Yeah.

Mike Hodson (45:57)
CVs and the other, it had a bias towards males because the 10 years of CVs that had been trained on were mostly male. Yeah. So in the end, they just, they had to, they had to bin it. there's, there's, that's probably an example of where you've got to be careful with this stuff. You've got to go and be careful what you point AI at and what data you're looking at. If you're looking at messy data, you're to get messy insights.

David Morgan-Gumm (46:05)
Yeah, I'd got a lot more male applicants.

Mm.

Yes, it's one of the considerations with third party tools as well, right? I you said you go to the recruitment conference and there's a lot of, it's become very AI heavy. ⁓ I kind of wonder what challenges that they've all overcome to be able to sell that to the recruitment industry and to get involved with it and kind of what they've done differently to overcome those challenges. ⁓ It's not something I have an answer to, but.

Mike Hodson (46:48)
Mmm.

No, I'm not

necessarily sure it's their problem or interest either. You know, if they, they sell like products and then you make bad decisions off the back of it because of your data being a mess. I don't know whether it's on them or not. It's probably on you.

David Morgan-Gumm (46:58)
Yeah, maybe not.

Probably, probably, yeah.

Mike Hodson (47:12)
I know I've got a firsthand story from someone who's told me that they tried plugging in a third party to actually can't remember the name of it now, but it was a, it was a sales based one and it was give us your top five salespeople. then we'll learn what a top salesperson looks like at your company. And then we'll run that model against everyone else at the company who's in sales and kind of tell you who your next stars are going to be.

And yeah, so it sounds worth a look, doesn't it, on the face of it? But what they found was it just took the bias that had already been there, but kind of invisible, and made it worse and visible. So they didn't use it in the end because they didn't want that. They wanted it to be more diverse and inclusive. So they didn't go with it. But just an example of the third party was doing its job. It was doing what it...

David Morgan-Gumm (47:41)
well.

Yeah.

made it worse.

Yeah.

Yeah, I suppose. Yeah.

Mike Hodson (48:10)
sent on the tin.

David Morgan-Gumm (48:12)
then, Mike, we've covered some incredibly insightful and important topics so far. I wanna do a bit of a lighter tone to the next section, a quick fire round, something new that I've put in. And I just kind of want you to give me your immediate and kind of brief thoughts on these prompts, like you're my own personal AI.

And yeah, I just kind of want to know your thoughts just straight off the top of your head. Cloud computing is a very big part of this podcast. I'm not sure whether or not you guys have moved fully into the cloud yet, or whether or not you still kind of hybrid. But what's your cloud provider of choice?

Mike Hodson (48:58)
Azure, in a word. My reasoning would be because I know people that could help me with it.

David Morgan-Gumm (48:59)
Great answer.

yeah. Yeah, Azure is really big within the data community as to a lot of meetups, network with lot of people and pretty much everybody is mostly Azure.

Mike Hodson (49:05)
Yeah.

Yeah, the support's there, isn't it? If you walk into a meetup and go, I've got an issue with Azure, people will come running. So, A, sympathise, tell you about theirs. Yeah.

David Morgan-Gumm (49:25)
I want to help.

next, what's the most underrated soft skill as a data developer?

Mike Hodson (49:35)
Probably all of them, isn't it? But if I had to pick one, communication, I'd say. Because, yeah, again, if you've got a great insight, but you can't communicate it, or you can't understand what the business is asking you to do, you've kind of got no chance. So yeah, communication.

David Morgan-Gumm (49:40)
Okay.

That's very important. I mean, data analysts, for example, they've got to have really good communication skills to, like said, to be able to provide that analysis, to be able to provide that insight to the business in a way that they understand and, know, kind of be a translator between the data and the business stakeholders.

Mike Hodson (50:12)
Yeah.

David Morgan-Gumm (50:13)
if you weren't in tech, what would your dream job be?

Mike Hodson (50:17)
The first thing that springs to mind is this behind me. probably playing that or similar in one way or another. Yeah, something in music, I would say.

David Morgan-Gumm (50:27)
Amazing. I actually find I'm always surprised by how many people within the tech space like music, like the music industry. think, think like nearly about 50 % of my team at the moment are also musicians, like they do. They're in bands or ⁓ they produce music, you know, as a hobby. ⁓ I just, I just found that to be a bit of a link.

Mike Hodson (50:54)
interesting.

There's a blog post there somewhere, there? On this. I reckon there must be a study. Yeah, I think there's probably is something in it. As in, I can see similarities between kind of how I can see how maths and music would go together. ⁓ Yeah, worth a look, think. Worth a research project.

David Morgan-Gumm (50:58)
Yeah, yeah, yeah, how it matches. Yeah.

Mm.

Yeah, definitely.

And last one. ⁓ Is there one piece of tech that you wish existed right now that doesn't?

Mike Hodson (51:30)
I would like to have the the bfg from doom 2 but if you're familiar with that that would be quite a fun piece of tech to have if it was available but assuming not teleporter Yeah

David Morgan-Gumm (51:37)
Okay.

teleporter, you can just

get around anywhere instantly.

Mike Hodson (51:49)
Exactly. like travel. I just don't like the bit in between. If you see what I mean.

David Morgan-Gumm (51:52)
So you

dematerialize on the spot and then rematerialize on the Costa del Sol.

Mike Hodson (51:58)
Yeah, exactly that. That's

exactly that. Yeah, you read my mind. Yeah, no queuing. Yeah. Yeah. I'd actually if you could give me the rest of it, I wouldn't be bothered about which drink it was. Wouldn't have to be a margarita. But yeah, that that kind of thing. Who knows? Who knows? It may be coming at some point.

David Morgan-Gumm (52:02)
Yeah. In a sunbed with a margarita.

I'll say it.

I love my margaritas at the moment. I don't know what it is about them. I've been making them a lot at home too. I've been putting maple syrup in instead of simple syrup, which I find just gives it bit of more of deeper flavor, but that is a bit of a tangent. You've never had a margarita? All right. Well, I'll have to invite you around at some point and make you one of my margaritas. ⁓

Mike Hodson (52:36)
I don't think I've ever had one, ever. Y'all think so? No.

Yeah, I've had a margarita pizza if that count but now

I don't think I have

David Morgan-Gumm (52:51)
Very, very different things.

That's an idea for crazy Pedro's, isn't it? Just try and get, you know, change the Margarita pizza to be act drink style Margarita pizza. There's something there. Yeah. Yeah. All right. ⁓ kind of final topic then, ⁓ we've been over. ⁓

Mike Hodson (53:04)
Yeah, there's definitely, there's at least like a Marguerite night that they can do there. Isn't there, there's some kind of marketing thing. Yeah.

And we say.

David Morgan-Gumm (53:18)
A lot of the AI stuff we've chatted about recruitment and usage of data and insights and analysis. the things I always like to focus on on this podcast is the future. We're very, very future thinking. We're big idea people wanting to get people interested in the world of data and the industry as a whole. Want to speculate about what the future holds for data and the industry.

looking ahead, I suppose, in your research conferences that you've been to, insights that you've got, is there any technologies that you're really excited about and then the use cases for them?

Mike Hodson (54:01)
Good question.

I the one that springs to mind that I'm excited to see how it plays out and progresses over the next few years is vibe coding. So, you know, this thing where you alluded to it earlier, but where in theory you don't need to know how to code to be able to, manipulate data and get insights. You could do it. You could do it yourself. I'll say a CEO you wouldn't need in theory, you wouldn't need a data person in between. So I'm excited.

to see how it plays out and see in five years, what is actually going on? Are the, there a mess that data professionals are being brought in to sort out from CEOs that have been directly using vibe coding to write stuff that then has turned out to be not secure or maintainable, or has it been, has it just gone perfectly? All those little flaws have been ironed out and that is now what's happening.

David Morgan-Gumm (54:37)
Mm-hmm.

Mike Hodson (55:00)
And instead of a team of data professionals, you've an AI expert just kind of sat in the middle, pulling all the levers. Yeah, maybe doing human checks. Don't know. It's, it's going to be interesting and exciting to find out.

David Morgan-Gumm (55:01)
Mm-hmm.

Helping everyone out, yeah.

yeah, that would be my answer as well. There was a meetup that I went to that had a talk and a demonstration a couple of months ago. And it's the first time where I actually became scared for people's jobs. I mentioned AI Foundry earlier on Azure. The talk that happened was the guy basically wanted to

Yeah, develop an application end to end had a backend database, backend API, a middleware service layer, and then a front end application on the back of it for access on the web. He created three AI agents within AI Foundry. One was a product manager. The other was a tech leader.

and the other was a business expert. And so he gave these three AI agents the knowledge that they needed and basically access just links to certain documentation, best practices and processes. had a ⁓ GitHub repository set up, which are basically with a read me file, which was just very

very basic and high level about what he wanted to implement. he on Visual Studio Code using the agent with Visual Studio Code, you know, he had his MCP stuff set up to GitHub to AI Foundry. And he just pressed, you know, run, make this application for me. And it just went away and did it like the the AI agents were talking to each other.

basically sending, when he expanded them, what looks like emails to each other. He even gave them names and personalities, and they were calling each other by their different names, and they were responding in the personality types that they're saying. And

know, by the end of it, he had a fully working application that matched the high level spec and the MVP that he wanted. And how was it done? I was, I was absolutely stunned. I couldn't believe what I was seeing.

Mike Hodson (57:40)
I imagine you needed a

margarita after that. I can see why that would worry someone because the bit I've not heard anyone do before is that thing of creating the personas. I've not heard anyone do that. that's, that's like you say, you know, the concern about fearing for the jobs. That's he's literally done it in that example, hasn't he? The jobs are there.

David Morgan-Gumm (57:42)
I did, yeah.

Yeah, exactly. Yeah. You need to have a look

at AI Foundry. mean, things like Confluence, for example, you can take an agent, a business expert, an SME, you know, and give it access and train it on all of your Confluence documentation for your business and business process and how it all works. And then you can just go to that agent and ask it questions on...

how it needs to work. So while you're developing, while you're vibe coding, you can go to an agent to ask questions about how something needs to be implemented rather than going to a person, having a meeting, having a majorly drawn out process. that's my, that's, I kind of answered it in addition to your stuff, but yeah, it was mad. And that's where we're at now.

Mike Hodson (58:44)
Hmm.

David Morgan-Gumm (58:55)
Currently, you you listen to some podcasts, industry, you know, industry experts, things that Microsoft, open AI and Google are actually creating. And they, they all, they all say that what you've got now nowhere near what we were actually at. ⁓ and we just need to, you know, it needs to be released future in a way that doesn't cause a load of additional harm,

Yeah, I mean, even, you know, companies like Meta have set up super intelligence teams, seeing them paying tens of millions of dollars to open AI employees to poach them to get them to work on this super intelligence team. Yeah, so they, they think that super intelligence is less than five years away, is mental, absolutely mental, just sci-fi, isn't it?

Mike Hodson (59:35)
seen that, yes, sign on bonuses.

Mm.

David Morgan-Gumm (59:50)
science fiction.

Mike Hodson (59:52)
It is, and the bit I'm intrigued in is kind of how long that's going to take. And the example you've just given, I can imagine that looking amazing in a demo and terrifying, but then it's like, say it's well, if that is that, if what, if he then goes and deploys that, is it maintainable? Is it secure? If it's not, is it anyone's ⁓ whose fault is it? Who's going to fix it? So there's.

David Morgan-Gumm (1:00:03)
Yeah.

Mike Hodson (1:00:21)
all those considerations, but it is interesting and it's exciting.

David Morgan-Gumm (1:00:24)
it is. You have to have someone who's, you know, a senior technical person or an advisor to review it all and make sure that it matches the code, you know, coding standards and processes and that it's secure and all that, which is why I mentioned earlier about juniors pros and cons. going to be harder for junior people to get into the industry. ⁓ They're going to have to, they're going to have to make use of AI.

to keep up with the AI. for a message for the entire audience out there is if you're a junior, you're in university doing computer science, looking at a career or a job in data, AI or development, no choice. You have to use it. Otherwise you're going to fall behind.

Mike Hodson (1:01:12)
Yeah, you have to lean into it. It's not like...

David Morgan-Gumm (1:01:14)
Yeah, you just got to go

full force and make the most of it. Because I think that's the only way that, you know, in the modern day or within five years that they end up becoming successful with it all. Because if you don't, you're just going to fall so far behind the efficiency levels of those that use it.

Mike Hodson (1:01:31)
I agree. It's got to be a collaboration mindset rather than a competitor mindset. Because if you're, if you're going for a junior role and you're kind of pitting yourself against AI, the AI is going to win because it's, you know, enthusiastic, confident, keen, probably makes the odd mistake, but probably so do you as a junior. So you really, you want to see it as a collaboration, don't you? Where you can.

David Morgan-Gumm (1:01:35)
Yeah, exactly. Yeah.

Mike Hodson (1:01:59)
Use it to do some of the donkey work for you, but you're still doing the strategic thinking Did you see there was a study recently from the university of amsterdam about the One of the reasons you still need the human check. This is one of the things that reassures me something that ai ai can't do at the moment anyway is navigation as well as humans so the The test they did all the study they did was

David Morgan-Gumm (1:02:09)
Mm-hmm.

Okay.

Mike Hodson (1:02:28)
If you put an environment in front of a human. So if you imagine an actual scene where you've got a river going across, you've got a bridge, you've got a path going, winding up a mountain. When you show that to a human, the human doesn't just see the things, the river, the bridge, the mountain. Without trying, you automatically navigate a path to get. You can, you go, ⁓ I could go over the, over the bridge.

David Morgan-Gumm (1:02:54)
I never thought about that.

Mike Hodson (1:02:56)
I could

walk around that path and I could get up the mountain to the top. Humans do that instinctively. AI struggles with it. So AI sees the things. It doesn't do the mapping out the navigation bit. So.

David Morgan-Gumm (1:02:59)
Hmm.

So that's

a big area of improvement, especially for like humanoid robots and things like that, or the stuff that's being built to replace transportation. you've got to kind of go through obstacles like that and make decisions on the fly.

Mike Hodson (1:03:26)
Exactly. For my teleporter, probably going to need to sort that out, the navigation. So that's something that gives me hope or, you know, makes me think, well, you still need the human chat. There's still things that AI can't do. Now, whether in five years, AI can do that or whether it's just one of them things and it's not going to be able to, we'll see. I'm sure someone's working on it.

David Morgan-Gumm (1:03:28)
Yeah.

Yeah, yeah,

changed.

Yeah, I'm sure they are.

I think, ⁓ I think we're towards the end then. ⁓ I haven't got anything else to, to go through. So I'm not sure if you got any, closing remarks or, ⁓ any additional discussion points that you want to go over.

Mike Hodson (1:04:09)
Only thank you for having me on and it's been an absolute pleasure usually and so people try and stop me talking about this stuff so to be encouraged to do it is a delight so thank you very much for having me on.

David Morgan-Gumm (1:04:10)
You

You're welcome.

Yeah. I encourage people. Yeah. No, you're very welcome.

So yeah. And thank you for joining me on the podcast. I know it's been, a while coming. ⁓ and I, I definitely hope to have you on again sometime in the future in, in some way, shape or form, whether that's another podcast we can even maybe, maybe let's get back together in three years time and talk about, you know, how, how this has all developed since, since the last recording, but, ⁓ you know,

your insights that you've given today on your career journey, agile and scrum, data and recruitment, the discussions on what we think the future of AI would be have been absolutely invaluable to me and I'm sure they'll be invaluable to the people listening as well.

Mike Hodson (1:05:02)
Thank you very much. Put me in for three years time and we'll do it again.

David Morgan-Gumm (1:05:04)
Yeah.

lastly, thank you to all our listeners for tuning in to another thought provoking episode. If you

this ⁓ conversation as engaging as we did, ⁓ please remember to subscribe, leave us a review and share it with your network, wherever your network, ex LinkedIn, wherever. You can find all the relevant links and handles in the show notes below and to connect with us or Mike on social media. That information will always be there. And if you want to visit our website for more resources, go to www.

SQL Squared.co.uk. ⁓ And if any of you have any questions or any discussion points that you'd like to cover in the podcast or you just want to get in contact with me, please send me an email at ⁓ mailbag at SQL Squared.co.uk or fill in the contact form on the website. It's totally up to you. So until next time, keep exploring and keep innovating. Thank you so much, Mike.

you again sometime soon.

Mike Hodson (1:06:14)
A pleasure. Thank you very much, David.

David Morgan-Gumm (1:06:15)
Cheers, bye bye.