The 5 Future-Proof Skills Every Tech Pro Needs Now

00:00:01 Mark Smith
Welcome to the ecosystem show. We're thrilled to have you with us here. We challenge traditional mindsets and explore innovative approaches to maximising the value of your software estate. You don't expect you to agree with everything, challenge us, share your thoughts and let's grow together. Now let's dive in. It's Showtime. Welcome back, everybody. For another session of ecosystem show, there's three of us today, Chris, Anna and myself. Welcome everybody. Yay.

00:00:32 Ana Welch
Hello. Hello.

00:00:32 Mark Smith
Thing it's been. A while since we've been done a show, I feel. It's.

00:00:36 Ana Welch
Been a while.

00:00:37 Chris Huntingford
It does feel like it's been a while.

00:00:39 Ana Welch
Yeah, it's been a while. Maybe because.

00:00:40 Mark Smith
And I feel that.

00:00:41 Ana Welch
We're so tired. Like look at my face. I feel so tired.

00:00:45 Mark Smith
I'll be traveling. I just come back in Australia.

00:00:45 Chris Huntingford
You look, you, you, you were vision Adam.

00:00:50 Mark Smith
Well. There's a lot to discuss cause a lot is changing in the in the in the AI world particularly, we're finding across our landscape. I think tonight why don't we have a chat about? The salary report I think, Chris, that you highlighted recently around you know what the AI space is looking at #2 would be good to discuss would be. Red teaming, I feel Red teaming is a is a growing category that we we should be focusing on the the other area that's caught my attention lately. There's there's been a prediction that within two years our LM's large language models as we know today will be.

00:01:19 Chris Huntingford
Yeah.

00:01:34 Mark Smith
History and will have been replaced with what are called world models and Google announced this at the end of their session, moving to a world model, and the world would model the difference. Let me just summarize slightly, is that a language model is all based on language and therefore generally English, right? But that's a language. Based model, a world model is based on all 5 senses that we have plus potential future sensors that we don't have. In other words, spatial awareness is part of the the language models of what we see, what we hear, what we touch, what we taste, what we smell could all be part of the world model. So it's basically how do we, you know, integrate into the world around us is the future modeling coming out? And also much closer aligned to the way our brains work and that our brains don't have infinite memory. It has ability to focus on the things that are top of mind. Right now. We have, of course, our subconscious, which is long term memory, but it's still a lot of the decisions we make is not necessarily. Touchable to something under that, that type of Model 1 of the weird things about LA LM's right now is that because they know all knowledge, no human does that, and it's so therefore you can. I don't know. I I'm just finding lately that I can read a lot of content and can tell pretty quickly that it's been generated by an LLM.

00:03:03 Ana Welch
So.

00:03:03 Mark Smith
And and what concerns me is not like, yes, I expect people to be assisted, you know, with AI nowadays. But it's kind of like. They're giving up on actually proofreading or going actually, does that make sense? Like, would I say that?

00:03:22 Ana Welch
Exactly. I, we, Andrew and I put our house up on the market, the London House because we love it in Spain and when the agent came to the House to like, talk to me about the, you know, their organization, he sold to me all of this technology. Which obviously I immediately bought. I was like, yeah, that's the like. That's the way to go. And he told me about this software that they have bought. To match, you know, prospective customers with the property that we've got and their needs and like statistics and social media, even all of the public stuff and and and so on and so forth. And then they created a brochure and that brochure. Was shocking. Like they left out important bits of information, but then they put up. I know the fact that we have aluminum plugs like who cares about that, you know in a in. A in a brochure or they. Three pictures of our bathrooms but no 0 pictures of our terraces like what is happening. So I think you're absolutely right. Because when you started talking and you were like when I read a large piece of information or a big text, I thought that what you were about to say was. I I find that I immediately can understand it and like going to the essence and stuff like that. I find it. And that and then you went on to say. Something else, right? But. When you started like that, there was like, Oh my God, I feel the exact opposite because so much of the text out there is outputted by MLM's and it just doesn't make sense and like confuses my brain. Just makes me want to take a nap.

00:05:18 Mark Smith
It's it's. I do find it. I feel that what I have learned is is better grammar, better English from a I definitely. But then the way that it's better is also one of the tells. Like, you know, there's kind of three things that are giveaway on AI right now. And I'm not talking about the phrases because there's a heap of phrases. But the three giveaways is the EM Dash now the EM-IN. English. Is actually really important. That's underutilized because it's never been on the keyboard, right? It's not a a, a key that you can tap on the keyboard, but it is great for shortening longer. What would potentially be longer sentences, but still get the concept across. The next one is the Oxford comma, which it's something I've always used because I'm bad at commas, so I overuse them rather than. Underuse them because my natural inclination would be underuse them. And then the third one, of course, is emojis in the in the text and for ages there was a period of time that I thought of. Modis did help with readability. Now I feel it's a blocker to people, people like, that's I. That's like AI generated. What? Why would I read it type thing and I think yeah, it's going to change in time. Of course.

00:06:33 Chris Huntingford
Yeah, I dude. So I struggle with AI generated content because I feel like it's it's very seamy. You know, it's like I'm gonna make a terrible reference. I haven't read the book I'm gonna make. A. Bad reference. So somebody said to me recently that AR was a bit like reading 50 Shades of Grey and I was like, OK, why like, what does? It mean it's like well. The grammar is good, but it's terribly written and I was like actually that that makes a lot of sense, right? Like I understand, I understand why they would say that. Because. The grammar is good like it it is good, but the actual verbiage, the things that it says it's like very semi. I feel like every person that makes an AI driven generated post without reviewing and putting their own personality into. It Yeah, is very sany, so I don't get that like personality from the person and. That sucks, right? Like I I don't think that's why I would want to go and. To a social media platform or read a book that was generated by AI. Because I don't. I don't care. Like it's very Sammy. It's very boring. Like so let's let's take let's let's use Andrew as an example. Actually Andrew's a good example. Yes, Andrew's. A very good. Writer, like he's very articulate and you can, when I read Andrew's work, I can hear Andrew, OK.

00:07:47 Ana Welch
Yes.

00:07:48 Chris Huntingford
Like when I read Anna's word, I can hear Anna when I read social media that is generated by an AI box. I feel very like void of any sort of response and that's why that's why I think it's a.  Problem. It doesn't drive any sort of emotion in a person.

00:08:08 Mark Smith
Yeah. Now one of the other interesting announcements that happened last week was Grok 4 came out.

00:08:15 Chris Huntingford
Mickey Hitler.

00:08:16 Mark Smith
And it. It has mounted all the kind of assessments against everything in in the space about how good it is, but one of the problems with this is that what these LLM companies are doing are kind of priming. The AI for the assessment and not the assessment is right. So they it's it's ranking higher because it's kind of like we've given you. I'm not saying given the answers, but you've been you're leading the witness almost and how this has been done. So it's becoming a bit of a. But what I did notice is that the US government awarded a massive contract. Grok for the defence department.

00:09:00 Chris Huntingford
Wow. Wow, I didn't know. That hmm.

00:09:03 Mark Smith
Yeah.

00:09:05 Chris Huntingford
That is interesting. So.

00:09:07 Mark Smith
Very interesting. And of course, right around the corner, you know, open AI is about to release CHPT 5 O, which will be, you know, they're saying some quantum leap ahead. So it's it's a, it's an interesting part of the year right now. And and what is going on. And and what is changing? Chris, earlier this week you said to us. Salary survey of what's going on in the AI space and it's the first thing I've seen around salaries really in this. Nice. Can you share some stuff there?

00:09:41 Chris Huntingford
This was really interesting, right? So 10th Revolution Group are probably one of the most. Large I I wouldn't call them a recruiting company. The bigger than that. Right. But they are, they're they're. How do I frame this? They're very good at driving relationships with people and customers and things like that. And something I learned very quickly was that. They do these salary surveys and I've been part of the salary survey with them for years, right? Like I've I've done this a lot of times. OK, this is the first time I've been featured in the forward, which I was like, well, OK, that's pretty cool. So I had a call with them and the first thing I did was we had a call with about 400 of the people on the call, and we were talking AI and like what it means for the wider job market and everything. They off the cover they went and did a whole ton of research. OK? Like a whole bunch of research. And they came back with their salary survey, which we can share the link in the chat right now. What's interesting about this crowd is that I think a lot of people think that they are just recruiters. But they really aren't. They do a lot of like matching between partners and customers and it's way bigger than you would think, right. And they are by far they by far have the largest lens into. The corporate market that I've ever. Seen. And when I say that, I mean they have like thousands of people talking to thousands of other people. So it's a bit like a neural network anyway. So they released this and I was like, OK, cool, you know, so I stuck in the cloud lighthouse messaging. And I was like, yeah, you know, embedded AI extensively, I differentially, I, blah blah blah. And then I read the rest of the survey that talks about what people are getting paid to do these jobs. OK, so are you all ready for for this right. I'm going to just screen share if that's cool because the this blew my mind. OK, so this is the celery city.

00:11:27 Mark Smith
Yeah, do a screenshot.

00:11:33 Chris Huntingford
OK, so it's got your usual, you know, like blah blah people talking about stuff, sex, facts, facts, facts, facts. Then you get into the nuts and bolts, man. This is real interesting. So let's take a look at the the UK. So I like how they break it down. So they break it down into medians, right. So you've got a lower quartile, A median and an upper quartile. OK. So here, let's pick the Chief data officer. So Chief data officer in the UK is getting paid at the upper quartile around 200. KA year OK. Chief AI officer is getting paid 235K a year. Let's take a data scientist right. So data scientist is getting paid in the upper quartile OF84K a year.

00:12:15 Mark Smith
Wow.

00:12:16 Chris Huntingford
This is a new job called a prompt engineer and a prompt engineer is getting paid 115K a year. That is 20 grand more than a data scientist is getting paid to speak English to an LM.

00:12:27 Mark Smith
Amazing.

00:12:30 Chris Huntingford
Is there wild? My favorite is the AIF, the Sist, so an AI ethicist is getting paid around 107 K. Yeah, right. That's some. So remember when they started out, it was called ethical AI, not trustworthy computing. So that was around 2018, right? So an AI ethicist is somebody who understands the.

00:12:31 Ana Welch
Yeah, yeah.

00:12:38 Mark Smith
What's that?

00:12:42 Mark Smith
Yeah, yeah.

00:12:51 Chris Huntingford
Core qualities of what an AI model does and does not promote things like anthropomorphization and that type of stuff. So.

00:12:57 Ana Welch
Yeah, yeah.

00:12:59 Chris Huntingford
This is really interesting. So I was like, OK, cool, cool. Now let me tell you why I think this is interesting. When you take a look at the power platform and I'm going to use power platform here on purpose, right? When you take a look at the power platform salary guide, a par platform architect, right, that's an AI architect, a power platform architect gets paid around 110K year on the upper quartile. OK.

00:13:19 Ana Welch
Yes.

00:13:21 Mark Smith
We're talking pounds, right?

00:13:21 Chris Huntingford
What is?

00:13:22 Ana Welch
Yeah. Pounds. Yeah, yeah.

00:13:25 Chris Huntingford
What is the difference between a power platform architect and AI architect? OK so I. Was like what? Like, I'll buy let's let's figure this out. So I went through the process of trying to understand and. It. Turns out that because part platform is the kind of core provider for extensible AI inside of the Microsoft framework, you have a lot of like core qualities that make you an AI architect. So you can use those qualities to learn things like foundry, extend your Python coding skills and your C# coding skills and. Blah blah blah. But that is like a 30 to 40 Grand leap on your salary. Yeah, isn't that wild? Also look at the difference between the UK and the US.

00:14:08 Ana Welch
Ohh yeah there's we cannot look at that difference like that. The I don't know. The salaries are really, really different, different from the US to the UK.

00:14:17 Mark Smith
Are they not lower in the UK, is that?

00:14:18 Ana Welch
What you're saying? Yeah, a lot lower. And in fact, looking at Europe in general, not even comparing to. Not even comparing it to. The US UK salaries are plateaued like, I don't know. They they are definitely comparable to salaries in Romania. Our.

00:14:43 Mark Smith
Wow. Wow. Are you serious? It's incredible.

00:14:47 Chris Huntingford
England pay? Awfully like, awfully badly, yeah.

00:14:51 Ana Welch
I'm serious. Yeah, because and it's. And that's like a ripe market we were. I got back into the Economist this week because I I finished Ezra clients book abundance. Really, really good. I recommend it. And. In the Economist last week, I think they were talking about how the US is moving their offshore. Workforce from India to the UK because the price difference isn't significant enough and you know, obviously you go to the UK and you get people who speak native English and they have like an accent that Americans think you know, makes you a clever person. And all of these other advantages, right? So they go with their work there instead. Like the political landscape and everything else kind of goes into this, but also the way firms in the UK unfortunately are still doing business and the fact that they're still. Like not investing in those prompt engineers. Yeah, you know, to get enough value out of them, unfortunately. And Chris, you say that about an AI engineer and how how a power platform architect. Can easily be an AI architect because they have. All of the tools. But at the same time. It does take away a lot of the work that they were normally doing, like a lot of the power platform architects out there do focus on power platform alone. They know specific things, so moving them away and into AI architecture is not as easy. As. As it seems like I would hate to create a second wave of the secrets of the citizen architect, this time it's just not a.

00:17:02 Chris Huntingford
No, no, no. No, no, no.

00:17:03 Ana Welch
Thing. Yeah, yeah.

00:17:04 Chris Huntingford
No, we don't want that. We need technical architect.

00:17:07 Ana Welch
Exactly.

00:17:08 Chris Huntingford 
We don't need citizen architects.

00:17:11 Mark Smith
I was watching a I find this interesting about the UK. Michael one thing I'll add. When Meg moved to the UK, UK and Google actually and Google Australia moved to Google UK. She took a massive pay cut, going to the UK. Yeah. Yeah. And the reason is there's a glut of people wanting every job over there as opposed to in Australia. And so that may be looking at the lens from the power platform. And what I found the higher salaried people and the power platform, guess what country they're in for, whatever the role. Singapore.

00:17:41 Chris Huntingford
Wow. Wow.

00:17:42 Mark Smith
And one of the reasons is I think they've got a flat tax rate of 8%, so massively low the next best country, Australia.

00:17:52 Chris Huntingford
Wow.

00:17:52 Mark Smith
New Zealand. Very bad. You get paid. Very low, more like the UK, but it's interesting Australia there is. You get good income, good salaries there comparable in the market so. Interesting to note that and, but I was watching something this week where I'm I'm I've got this hyper focus at the moment in the latter half of this year that I'm focusing on, which is the the need for understanding what skills that we should be acquiring. To set us up. For the future that we're all moving into quite rapidly. And you know, one of the skills I'd never thought of is your ability. To have part of your team. BAI.

00:18:44 Chris Huntingford
Yeah. Yeah. So it's being able to observe, observe that layer of like, independent digital minions.

00:18:50 Mark Smith
Yeah. And so your colleagues going forward, there's gonna be a bunch of what we could classify as your colleagues that are going to be dedicated agents, which are gonna be critical in your team. I noticed in build recently one of the things I've. Done around intra ID. Is that they are allocating every agent an intra ID like they're getting same credential type access as though they are people and I think they need to do it Microsoft because that's the. Way they're ultimately going to. It got count for revenue, right they need.

00:19:24 Chris Huntingford
But but but dude, it is it, it makes sense.

00:19:25 Ana Welch
No, I think it's also from a it's I think it's also from a consumer perspective, you need to know you need to have the same monitoring for agents As for human. Colleagues.

00:19:36 Mark Smith
Yeah, agreed, agreed. Security is definitely a. Big part of it, but just back to this video. I watched 5 skills that. You need to develop for the future number one skill. AI augmenter.

00:19:48 Chris Huntingford
Yeah.

00:19:49 Mark Smith
So whatever your job, you've gotta be thinking how can I augment AI into what I am doing?

00:19:55 Chris Huntingford
You should be doing that now.

00:19:57 Mark Smith
Yeah, but I tell you. What I reckon in all the conversations we're having. People are they're not using AI. Maybe every couple of days, not every day. Are they looking? Are they asking that question? How could I make me smarter here? How could I assist me in doing what I'm doing?

00:20:15 Ana Welch
And even and even in initiatives that are happening right now, I am still seeing some where I'm like, OK and what are what's your AI ambition and they're like ohh. It's out of scope for this particular workshop. We're talking about data here. We're trying to like unlike.

00:20:28 Mark Smith
Yeah.

00:20:34 Ana Welch
Exactly. We are talking about data here. 

00:20:38 Mark Smith
I was just talking to my niece by marriage on the weekend in Australia. She works one of the largest pharmaceutical companies in the world and she is about to become a country manager. So she's quite high up right representing this pharmaceutical brand. And I said, hey. And how using AI. Ohh no it's been blocked, no one's allowed to touch it. No one's allowed to touch anything when it's been blocked on the no type of AI is allowed in the organization or why? Because we've got an antiquated IT department. And they've just said no. And I'm like, oh, my gosh, you guys are getting. You're gonna be so far behind by, you know, your competitors will be using it and you guys are just saying, ah, no. I'm just like this crazy this and I I don't. I don't think she's an outlier.

00:21:28 Chris Huntingford
Very the The thing is, I have these conversations with people all the time and like. Sadly. There are people that think that this is a phase and it's just going to pause and I'm like, crack on my friends. Like, I'm past the point of trying to sell this. Yeah, I'm like, this is like trying to explain spell check to some. 1 Yeah, like it's happening. You just just if you don't do it, fine, have terribly written documents. That's cool. Someday I. So if you don't do it, good luck with your business in the next two years. Because when I cancel you, this is. Gonna be like.

00:21:59 Ana Welch
I think that's something you do really well. Chris is actually using it like in the day-to-day work. Every single meeting will have some form of artificial intelligence, and every time you, every time you do it, it looks so seamless that I remember.  Yeah. When Mark first said maybe 2 dynamics months ago. In a panel you said Mark AI will be like electricity. We just turn on our lights and nobody thinks about it. Yeah. And I feel like Chris is really doing that in actual workshops that we do together and meetings and like everywhere it's like. He just used it. He just used Excel or PowerPoint or he turned on the light.

00:22:51 Mark Smith
And and look at the results, right, they're phenomenal results and particularly when there's a good data set behind things, right.

00:22:58 Chris Huntingford
Right. Oh dude, Gimme Gimme, like, right now. I'm busy doing a POC for a utilities org and I'm mad. I'll tell you why I'm mad because the data structure that I have is 17,000,000 records. I can't just fling that into a GBT and say, rationalize this. I'm mad. I'm like this should be so easy. Why can't I do that? So I've found ways to do it right, obviously with Power BI and fabric and everything, but like I'm angry because I know I can do this quicker than I am. Like it's the same with. The Azure DevOps thing like. Getting all of the records out of Azure Dev OPS and rationalizing that without having a million meetings.

00:23:35 Mark Smith
Yeah.

00:23:36 Chris Huntingford
It makes sense like in my brain, it just makes sense.

00:23:39 Mark Smith
Do you know what? For Pete? For tech folks that are listening, I think that you need to become gurus at applying RAG. In any situation. Like how can I get my data into that state that it's fully vectorized tokenised etcetera that I? Can use a. Massive data set at scale and mean I have to know which rag tool am I gonna use to be able to drive the outcomes that I need because as you said, Chris, when you're dealing with those you can't just, you know, upload a file. And say let's go right, because it's it's gonna the context window is gonna lose so much if it isn't vectorized properly and usable in a structure that you know can really drive value on that data. No.

00:24:22 Chris Huntingford
But that makes me mad. Like I get angry because I'm like I should be able to do this way easier. Like, let me give you this. So the use case here was. Here is a **** load of data. Tell us what we don't know. So like, I love that. You know, I love that type of stuff. I'm like, OK, I will demystify the hell out of this and find the things that you don't understand. It's a treasure trove. To me. Date is a treasure trove. I love it. So, like, cool and looking at this, I'm like. I should just be able to upload. Why? Like why can't I do that? Why can I not just upload this? Because you can't take 1000 records from 17,000,000 yeah, cause that's not a good enough subset. That's not how.

00:24:59 Mark Smith
No, no, no.

00:24:59 Chris Huntingford
It works.

00:25:00 Ana Welch
Yeah, to me. Simply the thing that you are that you are thinking I should be able to do this faster.

00:25:08 Chris Huntingford
Yeah.

00:25:08 Ana Welch
Even that is a mindset change because before, like, think about the simplest form of. I don't know uploading data in a power platform world. Yeah, we all come from that. Like, whatever, that's an Excel spreadsheet that you upload into your model driven. Let's just say if that doesn't work, you're like, OK, so this just doesn't work. I'm going to have to work harder to split into smaller files to get a better query. That. To clean it up, but I never ever thought. This should work better, but this is not a me problem, I there is a tool out there that's going to help me fix this.

00:25:44 Chris Huntingford
Yeah. But but you know.

00:25:48 Mark Smith
Do you know what's not a me problem as? Well, that we're a. We're so deep in the marketing as agents, I don't think agents is anywhere near easy enough that the marketing sells it. It is or anything close to a. No. No.Dude. Intake.

00:26:02 Chris Huntingford
Dude, it's not remotely agented like This is why Donna and I, like get sparkling automation, because at at at the very best and we got the German people to do this as well. By the way, Spock, I wanna mention so.

00:26:08 Mark Smith
 Yeah.

00:26:14 Chris Huntingford
At very best, at the very best. If you breakdown what an agent is right like there's action, there's orchestration, there's memory, there's all sorts of things, of the things that make an agent. We don't have half of them. We don't have orchestration. Like there's no planning like in in. I've said this in this podcast before. There's no planning. Yeah, in. Memory. There is no like actual memory in an ad like that. This doesn't exist, so why are we calling it agents anyway? This is gonna put me in. A soapbox, so. I'm gonna try and like not do that but like.

00:26:45 Mark Smith
Yeah.

00:26:46 Ana Welch
Yeah, try not to like we have enough people renting.

00:26:50 Chris Huntingford
No, I'm not gonna rant. I'm just gonna say what I'm gonna do is I'm gonna put this out there for. The 5th time. Show. And agents live. I put this out there to the world, to Microsoft, to partners like you. All are saying that you are making agents especially masses. Multi agent systems. Yeah. Show me like yeah. Show me how to do it because I I've tried. Yeah. And I've succeeded to an extent, but I want to see it working.

00:27:15 Mark Smith
Yeah. But but. There's so much work. To make it a reality, and I'm just like, hey, it's me to make my life. Simpler not make. Me have to learn 50 new things. To make it work so. I'm with you, but hey. We're we're time just about. That's what a rap. With. I feel that retaining yes is not getting the I. I don't know whether I'm ahead of the curve thinking companies need to be paying much more attention to this. I feel like it's considered an edge thing and a lot of companies that are doing stuff in AI and not even considering the role of a red team in the organization. Let alone having methodologies software tools. Repeatability train stuff to do this. UM.

00:28:02 Ana Welch
Yeah.

00:28:03 Mark Smith
What do you think as and what are you guys thoughts? I know you got opinions.

00:28:07 Ana Welch
From my perspective, after this over the past. Two weeks I've been doing some research, field research, meeting people, meeting people who have recently had their changed their jobs or are about to change their jobs or they're unhappy about the jobs. And as you. So I've been working with Transylvania University as well because we're working with them for virtual forum in the winter. So I went into quite a lot of details with regards to what disciplines are kids being taught in university, what sort of jobs people do. And when you want to go and do a new job. What sort of? What sort of things do you need? And this is all granted based in my hometown bash of Romania over there, people are what we call pro code. All of them are writing stuff from scratch, and I know we've been talking about platform a lot and. How you like can upskill towards AI from power platform, but we also know the sort of stuff that AI takes away from you as a power platform person because. It. Just. Builds it on its own.

00:29:29 Mark Smith
Yeah, yeah.

00:29:31 Ana Welch
OK, so going back to programming and how these people are taught, build careers and deliver stuff. They are finding right now that when it comes to progressing in your career or wanting to get a new job, you no longer have to be an engineer back end engineer or front end or full stack. Now right now your full stack. You're also develops, you know, cyber security, you know cloud. You know everything. Because these organizations have found ways to do tasks that otherwise you would have done right with artificial intelligence. So. Bringing it back to what you were saying earlier, Mark and you were like ohh, these are the skills that people should learn. That's how you know, we calibrate for a world of AI. I think and I would love it not right now, but maybe in a future episode if we could go deeper into, OK, fine. You have these skills, but how do you actually start? Because you come from a different discipline. You may come from pro code or you may be a business analyst or you may be a PM. Or like whatever you do, where do you actually start? And you know, for writing in particular, because that was my interest, you know, when talking to these people, like, how will? Professional developers be used in retiming and I think they're going to be used plenty. I think that there there's like a whole new job market for them, especially with vibe coding. All of these people who know, you know how code is written and how it should behave. But also. There is more generally I guess there is a course on red teaming that our friend Yuana compiled. I'm happy to kind of share my screen for just one second.

00:31:49 Mark Smith
Yeah.

00:31:49 Ana Welch
To kind of show you it, this is it. It's got like 10 episodes and it's free, right? So it's just on YouTube and go ahead and check it out if you want. I know like. Head start like where to start type of situation.

00:32:08 Mark Smith
I like it.

00:32:09 Chris Huntingford
I am, I feel very passionately about this. And I'll tell you why I'm I'm I'm in an interesting debate with some fool called Paul Swider on Twitter. Yeah, he's moron. Anyway, it's getting interesting because.

00:32:20 Mark Smith
So that.

00:32:25 Chris Huntingford
It's it's heated up a little bit where I think that reteaming is not necessarily what we think it is. I think that is very much more psychological. Like way more psychological I did, I'd like. I think this is going to create a whole void of of things that need to be done in the AI verse and I feel like doing that one time in the very beginning I was like, OK, I reckon we're gonna need lawyers, right? I think we're gonna need lawyers because of this. Now I'm saying I think we're gonna need therapists. And I'll tell you why. I think that the people doing the right teaming in certain aspect. It's. Are going to be so brutally challenged at the end of it because of all the **** that people put into other limbs and all the terrible prompts people put in. And I can tell you, I've seen some of them. It's like, it's awful. Like, it's honestly awful.

00:33:11 Mark Smith
And you're not talking. Just let's clarify here. You're not talking about. Ohh. That's a weirdly formed prompt you're talking about. Sorry. People doing gross things.

00:33:20 Chris Huntingford
Dude, it's more than gross. It's actually appalling. Yeah. Like what? People are typing in and other people are having to review is utterly appalling. And it's like, I'm not gonna. I'm not even gonna mention it. It's just gross. Like, it's disgusting and people are having to read that on the other end, right. And they're going to have. They're having to read that and apply responsible AI filters. So. Yeah. What's happening is that. Imagine you spend your entire day reading these awful things that people are. Into windows dude. Like I couldn't do that. I couldn't do that for a living. I could not physically because I'm very sensitive to bullying and that type of. So when people are putting that type of stuff in, I I I hate it like I think it's gross. I get angry. Yeah. And I think that rape teaming is going to be one of the most challenging areas in AI. To come, and I think that we're gonna need help. We're gonna need psychologists and therapists and things to brain shower our all the crap that comes into our. Heads.

00:34:15 Mark Smith
100% and it's different. There's different layers to this. You know, I've had some very interesting conversations with people on the red team inside Microsoft and they have had to stipulate their no go areas. I don't want to be involved in any racism based.

00:34:26 Chris Huntingford
Yes.

00:34:31 Mark Smith
Rig teaming. I don't wanna be involved in child or Kitty or anything like that's not me. Like they're very clear. What their. Proud of and and homophobia that the list goes on right. Every this whole you know categories.

00:34:39 Chris Huntingford
God.

00:34:46 Mark Smith
And they have.

00:34:47 Chris Huntingford
Also.

00:34:47 Mark Smith
To opt out like what they're not comfortable with doing because they know psychologically it will be detrimental. It reminds me a bit of the police force. I had a friend of mine become a cop. And I notice he's a good friend. His personality changed the longer he was in the police force. He liked people less and less because he saw the scum of society and he just became a. A very cold person, you know. And then I I also remember having a friend who was a police photographer, so she went in and photographed crime scenes and she had to quit because it was just.

00:35:22 Chris Huntingford
Going through that.

00:35:26 Mark Smith
She couldn't. She was getting to the point that she couldn't sleep anymore at night and stuff because of the gruesomeness that she had seen humans do to humans. And unfortunately, you know, that's the world we.

00:35:35 Chris Huntingford
Dude, people are awful. Live in people are awful.

00:35:38 Ana Welch
I have a friend who worked in an ICU department at a very prestigious Hospital in London, and she also had to quit for the exact same reason because when you're there, you're saving people who have been. Who have had stuff done to them by other humans.

00:35:56 Mark Smith
Yeah, just terrible crazy.

00:35:58 Chris Huntingford
People, people are awful. Like, can I tell you something? Like to wrap this up. OK. So do you know like? I can't even watch horror movies or like brutal I can't watch Game of Thrones and crap like that. I can't watch it like I can't dude. It makes me so physically angry. Yeah.

00:36:13 Ana Welch
Yeah, I'll. I'll watch it with you.

00:36:14 Chris Huntingford
No, I won't. Dude. I won't do it. I I like I I had this conversation with some the other day and I'm like.

00:36:18 Mark Smith
And it's like there's people at different, like different things trigger them different things.

00:36:20 Chris Huntingford
I just can't do it.

00:36:21 Mark Smith
Don't. Right? Correct.

00:36:22 Chris Huntingford
Yeah, I I can't. I can't watch it, dude. Like, I get so physical. Mad and I'm like, there's enough blood and guts and gross stuff in this world already. Like, I don't need more of it in my life. So for me, just for me personally. And This is why I get so sensitive to when people talk about rape teaming, I'm like.

00:36:33 Mark Smith
I get that.

00:36:37 Chris Huntingford
I I've seen it, dude. I've seen the things that people type in. Yeah. And let me tell you, it's ******* awful. Like it's. Yeah. People are gross. Like I I genuinely think that people are revolting in that sense where where a lot of people where they think it's OK to, like, research that type of stuff will do that type of thing. So that's why I get sensitive to it. And that's why I tell.

00:36:46 Mark Smith
Yeah, yeah, yeah.

00:36:58 Chris Huntingford
This I'm like, don't think this is a joke like this is really, really gonna affect us.

00:37:04 Mark Smith
Yeah, OK. I wanna wrap on a higher note than than that serious topic. And so I want to share this with you. And this is a book I read some time ago. So not recently there it is on my Kindle which is called range and it says why generalists triumph in a specialized world I think.

00:37:08 Chris Huntingford
Yes.

00:37:24 Mark Smith
That because AI is gonna give us such depth of specialty. Just like Anna, you said before, the developer needs to know Dev OPS. They need to know architecture. They need to know. Like. I think that our skill set. And for years I talked about, you know, I've, I've come from a consulting background of where we were taught the T consultant you should be knowledgeable, you know, an inch deep and a mile wide. But you need to be strong and really. Depth, you know, technical, technical depth. In one area, I think we actually the, the, the, the next World that we're moving into is going to the generalists that you have a wide range of knowledge around a a whole. Range of things. Things and know how to bring specialist speciality into specific areas. And so yeah, I think that the the generalists might be the that where things are at in the future and so something to mix up.

00:38:17 Ana Welch
I love it and I just I. Just bought it. Because I think that it's.

00:38:22 Mark Smith
Yeah.

00:38:22 Ana Welch
It's an audible because I have a three-year old and no time, yes. Yeah. I guess that my recommendation is that whenever Mark recommends a book and you feel like oh. My God, it's like. 1000 pages. I'm never going to go through it just by inaudible. It's just like it's still very good.

00:38:39 Mark Smith
Yes. So this is what Bill Gates is about. This book. He's actually quoted it fascinating. If you're a journalist who have felt overshadowed by a specialist colleagues, this book is for you, Bill Gates. So it's got some. It was actually a Microsoft Stew person. Somebody in the Stew that I was on a call with and he brought it up and it's caught my ear about a year and a half ago and a Microsoft until call. And I was like, I got it and I read it and I'm just like, oh, like, I sent it to my son, who just started university.

00:38:55 Chris Huntingford
Good.

00:39:12 Mark Smith
Yeah, because that he doesn't have to think that he has to just become a super knowledgeable lawyer, only that he can go and explore it because it really encourages to explore other parts of your life, not just go. Deep on one thing.

00:39:27 Chris Huntingford
Yes, I like that. I like that Look. It's a good.

00:39:31 Mark Smith
It's a good listen, good. Listen. Anyhow, guys, I think we're at well over time. Thank you. It's been fun. And we'll see you on the next one.

00:39:36 Chris Huntingford
Thank you. Always rock'n'roll hey.

00:39:40 Mark Smith
Thanks for tuning into the ecosystem show. We hope you found today's discussion insightful and thought provoking and maybe you had a laugh or two. Remember your feedback and challenges help us all grow, so don't hesitate to share your perspective. Stay connected with us for more innovative ideas. And strategies to enhance your sufferer state and until next time, keep pushing the boundaries and creating value. See you on the next episode.