WEBVTT - Why Most AI Rollouts Fail (And How to Make Sure Yours Doesn't) 0:00:00.640 --> 0:00:05.680 Most AI rollouts looks something like this, a big announcement, 0:00:06.200 --> 0:00:09.560 lots of enthusiasm, and a few months later things have 0:00:09.640 --> 0:00:13.960 gone one of two ways, either barely anyone's using the tools, 0:00:14.040 --> 0:00:18.840 or everyone's using them to churn out AI slop. Sometimes 0:00:18.840 --> 0:00:22.840 both and leadless to say, the huge investment in licenses 0:00:23.600 --> 0:00:28.160 isn't paying off. Now, for today's episode, we are releasing 0:00:28.200 --> 0:00:31.680 a recording from a live webinar Neo and I ran 0:00:32.120 --> 0:00:36.720 recently on why AI rollouts fail and how to set 0:00:36.840 --> 0:00:40.879 yours up for success. We get into why these problems 0:00:40.880 --> 0:00:44.920 are happening, including the surprising research on how little time 0:00:45.080 --> 0:00:48.720 AI is actually saving most people, the training goes that 0:00:48.840 --> 0:00:52.000 does move the dial, and the question every leadership team 0:00:52.159 --> 0:00:56.920 needs to answer before spending another dollar on licenses. Well, 0:00:57.000 --> 0:01:00.320 let's jump in by starting with the biggest problem we 0:01:00.560 --> 0:01:10.800 are seeing organizations making Welcome to how IAI with me 0:01:11.000 --> 0:01:15.280 Doctor Amantha Imba and Neo Applin, head of Inventium AI. 0:01:15.880 --> 0:01:18.959 Each episode we share one practical way to use AI 0:01:19.120 --> 0:01:19.880 better at. 0:01:19.720 --> 0:01:20.920 Work and in life. 0:01:21.480 --> 0:01:24.560 No fluff, no tech jargon, just things you can use 0:01:24.560 --> 0:01:28.440 straight away. We are of course talking to you today 0:01:28.800 --> 0:01:33.120 about AI rollouts and why we've seen a lot of 0:01:33.160 --> 0:01:36.880 them fail and how we can actually make them work. 0:01:37.240 --> 0:01:41.760 So to start with, you might be like, oh my god, 0:01:41.959 --> 0:01:45.560 who are these randoms? Or maybe like you followed us 0:01:45.600 --> 0:01:48.120 for years and you're like you need no introduction. But anyway, 0:01:48.120 --> 0:01:50.520 I'm going to assume that, like we're two random people, 0:01:50.960 --> 0:01:54.320 So I'll introduce myself first. So my name is Amantha 0:01:54.360 --> 0:01:57.640 Imber and with my initials, well, it just qualifies me 0:01:57.680 --> 0:01:57.840 to be. 0:01:57.880 --> 0:02:01.280 Talking about AI, right, So what more do I need? Anyway? 0:02:01.320 --> 0:02:04.360 Aside from that, I'm the founder of Inventium and Inventium 0:02:04.400 --> 0:02:07.200 dot AI, which is our sister company where all we 0:02:07.240 --> 0:02:10.400 do is help people with AI skills, building and rollouts 0:02:10.400 --> 0:02:12.200 and all that sort of great stuff. I was awarded 0:02:12.240 --> 0:02:14.799 AI Consultant of the Year last year, so there we go. 0:02:15.120 --> 0:02:18.120 I'm award winning, so hey, you should probably trust me. 0:02:18.320 --> 0:02:19.280 Neo, who are you? 0:02:19.760 --> 0:02:23.560 I'm neo and matrix jokes going in here. So I've 0:02:23.600 --> 0:02:25.840 been in it my whole career. I've been building stuff 0:02:25.840 --> 0:02:28.040 for big companies, small companies, startup, scale ups and that 0:02:28.120 --> 0:02:31.320 kind of stuff. But AI has been the biggest change 0:02:31.360 --> 0:02:33.600 I've seen over the time. So AI is all I 0:02:33.680 --> 0:02:36.680 do these days. I'm the head of AI and invent him. 0:02:36.720 --> 0:02:39.560 AI and I trained companies on how to get the 0:02:39.560 --> 0:02:41.480 best out of this, and I usually start at leadership 0:02:41.520 --> 0:02:43.520 teams and boards and things like that and then go 0:02:43.560 --> 0:02:45.120 throughout the rest of the company to be able to 0:02:45.160 --> 0:02:47.320 implement AI awesome. 0:02:47.520 --> 0:02:51.320 So together Neo and I and the team at. 0:02:51.120 --> 0:02:54.720 Inventium, we have worked with a lot of different organizations. 0:02:54.760 --> 0:02:57.000 These are just a smattering. 0:02:57.160 --> 0:03:00.560 Who we have helped build AI capability, help them with 0:03:00.680 --> 0:03:03.120 roll outs, all that sort of great stuff. So today, 0:03:03.280 --> 0:03:07.440 like we're talking from firsthand on the ground experience plus 0:03:07.520 --> 0:03:10.400 all the research that AI has of course helped us 0:03:10.440 --> 0:03:11.000 get across. 0:03:11.240 --> 0:03:13.520 So let's start. Let's start with a really big problem. 0:03:13.639 --> 0:03:17.360 Great, So first problem is actually top down. A lot 0:03:17.440 --> 0:03:20.600 of leadership teams are saying this AI thing, Yeah, we'll 0:03:20.600 --> 0:03:23.480 just get someone else to take care of that. So 0:03:23.800 --> 0:03:25.640 it is a problem that needs to be solved at 0:03:25.680 --> 0:03:28.079 the leadership level, and it needs to be a problem 0:03:28.160 --> 0:03:30.000 that needs to be solved as far as here's what 0:03:30.040 --> 0:03:32.120 we need to have for our company, but also needs 0:03:32.120 --> 0:03:36.040 to be embodied by the leads as well. So one 0:03:36.040 --> 0:03:38.960 of the challenges we've seen is the company might have 0:03:39.160 --> 0:03:43.320 an AI policy, say, but everyone meets the policy and goes, oh, no, 0:03:43.880 --> 0:03:45.440 I think you can use it. No, I can't use 0:03:45.480 --> 0:03:47.560 it for that, And people get confused on what you 0:03:47.600 --> 0:03:49.720 can actually use it for because the policy is not 0:03:49.760 --> 0:03:52.120 specific about the types of data or the types of 0:03:52.200 --> 0:03:54.600 use cases you can use it for. So one department 0:03:54.760 --> 0:03:57.600 feels great to be using it, another apartment doesn't use 0:03:57.640 --> 0:03:59.960 it much at all. That's one of the challenges there. 0:04:00.360 --> 0:04:03.600 So the AI strategy is the first challenge, but within 0:04:03.720 --> 0:04:07.960 that strategy challenge, the CEO needs to know what these 0:04:08.120 --> 0:04:11.120 AI tools are, where the warts are, where they wanted 0:04:11.160 --> 0:04:13.680 to go, where they think it's going to actually change 0:04:13.760 --> 0:04:17.279 your company. And we found a lot of leadership teams 0:04:17.320 --> 0:04:19.480 will say, oh, that AI thing, will get that done 0:04:19.480 --> 0:04:22.640 by it department, but the CEO themselves don't use it. 0:04:22.960 --> 0:04:25.040 They don't get their leadership team to use it, so 0:04:25.120 --> 0:04:27.359 everyone else in the company is a bit scared to 0:04:27.480 --> 0:04:30.400 use it as well. I even had one lead who 0:04:30.440 --> 0:04:33.120 said every time anyone walked behind them, if they had 0:04:33.240 --> 0:04:35.880 AI opened on their laptop, they'd all tab out of it, 0:04:36.120 --> 0:04:37.720 which is an absolute wrong way to do it. You 0:04:37.720 --> 0:04:40.440 should be saying this is how I've used AI, and 0:04:40.520 --> 0:04:42.080 you should be able to use it in these kind 0:04:42.120 --> 0:04:44.120 of ways too, So we're going to walk the talk 0:04:44.240 --> 0:04:47.320 here with AI. Otherwise a whole bunch of your staff 0:04:47.400 --> 0:04:49.600 are going to be too scared to use the damn thing. 0:04:51.040 --> 0:04:54.360 Now, let's move on to another problem near I know 0:04:54.400 --> 0:04:56.960 you are seeing this a lot as am I. 0:04:57.480 --> 0:05:02.479 Yep, Absolutely, buy the licenses, go right ahead. But you've 0:05:02.480 --> 0:05:04.080 got to be able to train your people on how 0:05:04.120 --> 0:05:07.440 to use these things. Where is it twenty thirty bucks 0:05:07.440 --> 0:05:11.000 a month per user? A few times that buy your workforce. 0:05:11.080 --> 0:05:14.480 This is not a cheap investment, but you can get 0:05:14.520 --> 0:05:16.200 a lot of this back. We think we get a 0:05:16.240 --> 0:05:18.560 hell of a lot more than that back return on 0:05:18.640 --> 0:05:21.320 investment for us and the people we train. So thirty 0:05:21.320 --> 0:05:23.760 bucks is actually pretty cheap for what you can get 0:05:23.839 --> 0:05:26.880 out of AI and the people's time savings. But just 0:05:26.960 --> 0:05:30.800 giving them AI and saying now go is actually going 0:05:30.800 --> 0:05:33.280 to give you less productivity. What I mean by that 0:05:33.440 --> 0:05:35.760 is you're going to get more people getting confused with it, 0:05:35.960 --> 0:05:39.840 just writing big emails, creating AI slop. So you need 0:05:39.880 --> 0:05:42.400 to be able to train people on how to use AI. 0:05:42.600 --> 0:05:46.680 Licenses alone is not going to solve the problem. 0:05:47.200 --> 0:05:47.599 Now. 0:05:48.000 --> 0:05:51.039 Of course, the big promise with AI is that it's 0:05:51.120 --> 0:05:53.440 meant to save us all time. 0:05:54.000 --> 0:05:56.440 But when we look at the latest research. 0:05:56.040 --> 0:05:58.919 Around is coming out of HBr, they found that the 0:05:59.160 --> 0:06:03.839 average time saving is just like guess in your head 0:06:03.880 --> 0:06:09.000 before I reveal this number, two point five percent, Like 0:06:09.080 --> 0:06:13.839 that is nothing, And there are many reasons why this 0:06:14.040 --> 0:06:18.240 time saving is not happening. But I mean, what we 0:06:18.279 --> 0:06:21.960 are saying, particularly when you give people the licenses without training, 0:06:22.400 --> 0:06:27.880 is that you get this huge proliferation of volume of work, 0:06:27.960 --> 0:06:31.719 and generally that work is slop. Like you will have people, 0:06:32.080 --> 0:06:35.520 you know, typing a prompt into AI, going write me 0:06:35.560 --> 0:06:39.000 a ten page report and here are my three bullet points, 0:06:39.400 --> 0:06:42.200 and then that report then gets distributed to people people 0:06:42.440 --> 0:06:46.080 presumably have to read that report, or even worse, they're 0:06:46.080 --> 0:06:49.159 putting that report into AI and going summarize this report 0:06:49.160 --> 0:06:52.719 and they're getting back three completely different bullet points. And 0:06:52.800 --> 0:06:55.159 then companies are like, ohy am, I not saving time. 0:06:55.279 --> 0:07:01.440 Well yeah, next on our problem list. What we see 0:07:01.720 --> 0:07:05.880 is a big problem in how companies are measuring the 0:07:06.040 --> 0:07:08.840 ROI of their investment, which is obviously like you know, 0:07:09.160 --> 0:07:12.040 really it's a decent sized investment. When you think about 0:07:12.080 --> 0:07:14.840 licenses and training and doing this stuff properly. 0:07:14.840 --> 0:07:17.280 And what we see a lot of companies doing, and. 0:07:17.480 --> 0:07:20.240 Particularly like I know, like on sales calls that I'll 0:07:20.280 --> 0:07:22.440 be on, I'll be like, well, how's you know, how's 0:07:22.480 --> 0:07:23.280 AIU is going? 0:07:23.360 --> 0:07:23.520 Now? 0:07:23.640 --> 0:07:27.520 Like where are people at with adoption? And they'll go, oh, well, 0:07:27.680 --> 0:07:29.960 you know, we can see the x percentage of stuff 0:07:29.960 --> 0:07:33.240 for logging in daily to their AI tool, and some 0:07:33.280 --> 0:07:35.480 companies is a little bit more advanced, like they're measuring 0:07:35.480 --> 0:07:36.280 token usage. 0:07:36.320 --> 0:07:39.200 But the thing with this, this tells us nothing whether 0:07:39.200 --> 0:07:41.200 people are logging on or not, Like they could be 0:07:41.200 --> 0:07:43.360 logging on and creating slop and then logging off, Like 0:07:43.960 --> 0:07:48.240 that's not helping your company. So these vanity metrics are 0:07:48.280 --> 0:07:51.720 not helping you. We need to think differently about the 0:07:51.760 --> 0:07:55.280 metrics that you're using to go how much value is 0:07:55.360 --> 0:07:57.240 AI actually giving us? 0:07:57.960 --> 0:08:02.320 Okay, next to agents, Yeah. 0:08:01.960 --> 0:08:04.840 That's another vantage metric we've seen as well. Some companies 0:08:04.840 --> 0:08:07.000 are going great, we're getting more agents created, and we 0:08:07.040 --> 0:08:10.440 want more agents created. So, just for those who don't know, 0:08:10.600 --> 0:08:14.640 agents are effectively AI experts that do things for you 0:08:14.760 --> 0:08:16.640 the same way every single time. So you might have 0:08:16.680 --> 0:08:19.680 an agent that might create for you like a little 0:08:20.280 --> 0:08:22.480 the way we create our briefing here, or it might 0:08:22.520 --> 0:08:24.400 be an agent that does like a style guide checker, 0:08:24.440 --> 0:08:25.800 or it might be an agent that helps you with 0:08:25.800 --> 0:08:28.920 your email writing. Those kind of things. Everyone's building agents, 0:08:29.080 --> 0:08:31.040 and that's great, really good to see. In fact, we 0:08:31.320 --> 0:08:34.160 train people on how to build agents. But here's the thing. 0:08:34.640 --> 0:08:37.640 If everyone in your company is building agents, you're going 0:08:37.640 --> 0:08:40.200 to get one thing, which is a proliferation of agents. 0:08:40.520 --> 0:08:43.160 And the other thing is my agent will work differently 0:08:43.280 --> 0:08:45.319 to your agent. And so what you're going to find 0:08:45.480 --> 0:08:48.000 is that there'll be too many agents out there. They'll 0:08:48.000 --> 0:08:50.040 all work a little bit different to each other. And 0:08:50.080 --> 0:08:51.800 a lot of people are building agents. Don't know how 0:08:51.840 --> 0:08:54.240 to build these things, so they work the great way 0:08:54.400 --> 0:08:56.920 every single time. And so it might work great for 0:08:57.040 --> 0:08:59.160 me when I use it, when I give it to you, 0:08:59.160 --> 0:09:01.840 you might use work different because you use different words 0:09:01.880 --> 0:09:04.839 when you created your agent. So those kind of problems. 0:09:04.840 --> 0:09:06.960 The other problem is, because we've got so many agents 0:09:07.040 --> 0:09:10.600 being created, no one's actually looking after the agent. Who 0:09:10.640 --> 0:09:13.760 owns that agent? If it needs to be updated or changed? 0:09:13.920 --> 0:09:16.880 Who owns that? How do I find that within my organization? 0:09:16.920 --> 0:09:18.920 And so what we've actually got is a whole bunch 0:09:19.000 --> 0:09:22.200 of agents out there floating around that may or may 0:09:22.240 --> 0:09:24.480 not be giving value to the company. The other thing 0:09:24.520 --> 0:09:26.960 is we're going to have a whole bunch of legacy problems. 0:09:27.000 --> 0:09:29.880 Like if I'm got this old agent, if anyone hasn't 0:09:29.920 --> 0:09:32.079 used it in two months, can I delete it? If 0:09:32.120 --> 0:09:34.640 I do delete it, does that break someone's workflow? I 0:09:34.679 --> 0:09:37.560 don't know, because all these agents are just been thrown around. 0:09:37.800 --> 0:09:40.560 It's one of the big problems we've got. So agents 0:09:40.640 --> 0:09:43.760 are great, but too many agents not done particularly well. 0:09:44.000 --> 0:09:47.559 Isn't innovation. It actually is more confusing than anything else. 0:09:48.040 --> 0:09:51.920 And interestingly, like on the topic of vanity metrics, I 0:09:52.000 --> 0:09:54.920 certainly hear this a lot, particularly when I'm speaking with 0:09:55.040 --> 0:09:58.400 leaders who are looking at AI rollouts, and they'd be like, yeah, yeah, 0:09:58.400 --> 0:10:01.280 I've got like heaps of people they're bill agents. And 0:10:01.320 --> 0:10:04.160 I always wonder, hmm, what do you mean by that, 0:10:04.440 --> 0:10:09.120 because like, there's good agent building and there's crap agent building. 0:10:09.200 --> 0:10:12.400 If you've spent maybe like five or ten minutes like 0:10:12.440 --> 0:10:15.839 putting a system prompt into like agents on Copilot or 0:10:15.880 --> 0:10:20.599 a GPT on GPT, and then you're like, Bam, agent created, 0:10:21.520 --> 0:10:24.160 it's probably not the best agent because like Neo. 0:10:24.280 --> 0:10:26.200 How long do you take to build an agent? 0:10:26.280 --> 0:10:30.160 Typically half a day a day. Yeah, and that's because 0:10:30.360 --> 0:10:33.040 I'm building it, I'm refining it, I'm testing it, I'm 0:10:33.040 --> 0:10:35.600 trying to break it. I'm giving it different inputs, different 0:10:35.640 --> 0:10:37.600 outputs need to be put in. I need to make 0:10:37.640 --> 0:10:40.160 sure that this works reliably for me and everyone else 0:10:40.200 --> 0:10:43.600 I give it to every single time. And the other 0:10:43.640 --> 0:10:45.800 thing is I then need to manage and version control 0:10:45.880 --> 0:10:48.040 my agents because I'm going to come back to these 0:10:48.040 --> 0:10:51.360 ones when the engine changes, when it goes from chetpt 0:10:51.520 --> 0:10:54.240 five point three to five point four sometimes day break. 0:10:54.480 --> 0:10:56.960 Same thing with Copilot and all the others. So yeah, 0:10:57.000 --> 0:10:59.280 agents do take time to build correctly. 0:11:00.400 --> 0:11:03.840 I'm curious, folks at your workplace, who's working in an 0:11:03.880 --> 0:11:06.840 organization where like there's a bit of a vibe that like, oh, 0:11:07.080 --> 0:11:07.760 is AI going. 0:11:07.760 --> 0:11:08.440 To replace us? 0:11:08.480 --> 0:11:11.600 Like just give us like a thumbs up if that 0:11:11.840 --> 0:11:13.280 is your organization. 0:11:14.160 --> 0:11:15.880 Yeah, I can see LA lots of them's going up now. 0:11:16.000 --> 0:11:19.000 Okay, So I question, like, why would anyone adopt a 0:11:19.080 --> 0:11:21.760 tool that they think is there to replace them? This 0:11:21.920 --> 0:11:25.320 is a culture problem that very much needs to be 0:11:25.400 --> 0:11:28.720 managed from the top, because you know, if I'm working 0:11:28.760 --> 0:11:31.160 in an organization, I'm like, Oh, I think AI is 0:11:31.200 --> 0:11:33.800 going to replace me? How excited am I going to 0:11:33.840 --> 0:11:36.679 be to use the technology and really embrace it? 0:11:36.880 --> 0:11:40.880 Probably not much at all. Now. The final big. 0:11:40.720 --> 0:11:42.520 Problem that we're going to talk about before we get 0:11:42.520 --> 0:11:45.400 into solutions and then we'll get into your questions is 0:11:45.480 --> 0:11:48.600 brain fry. So I know that this has kind of 0:11:48.640 --> 0:11:51.840 been headlines in AI news. 0:11:51.920 --> 0:11:53.920 If you're like us. 0:11:53.800 --> 0:11:56.240 Inclined and like just you know, follow this stuff every 0:11:56.240 --> 0:11:59.480 single day. Brain fry was something that I was really 0:11:59.520 --> 0:12:03.679 awesome pair of research and published in HBr, led by 0:12:03.760 --> 0:12:08.280 this amazing researcher Gabrielle Obrezen Kellerman, who I've actually had 0:12:08.320 --> 0:12:08.720 on how I. 0:12:08.720 --> 0:12:11.600 Work a couple of times. But brain fry is what happens. 0:12:11.800 --> 0:12:15.640 And I'm curious as to anyone that has experience to this. 0:12:16.160 --> 0:12:20.240 When you have got multiple let's call them, like, you know, 0:12:20.360 --> 0:12:24.520 multiple AI tabs tasks going on at the same time. 0:12:24.640 --> 0:12:27.920 Maybe you know, you asked chat GPT to do one 0:12:27.960 --> 0:12:30.120 thing and then you've opened up another tab and you've 0:12:30.160 --> 0:12:32.800 got it working on another task and then another. And 0:12:33.200 --> 0:12:35.200 this is how a lot of people are working with AI, 0:12:35.800 --> 0:12:37.760 and people are getting to the end of the day, 0:12:37.960 --> 0:12:40.559 and it's different from burnout. They are feeling like their 0:12:40.559 --> 0:12:46.520 brain is fried, like they have intense cognitive fatigue and 0:12:46.920 --> 0:12:49.760 not surprisingly, like when we are fatigued like that, it 0:12:49.840 --> 0:12:52.560 causes us to make more errors. It causes us to 0:12:52.559 --> 0:12:56.120 obviously produce like lower quality work. It is not good 0:12:56.240 --> 0:12:59.800 for the output we are delivering in our job. So 0:13:00.040 --> 0:13:01.880 different to burn out, which is a little bit more 0:13:01.920 --> 0:13:05.079 emotional and some other aspects. But AI brain fright is 0:13:05.080 --> 0:13:07.920 a really big problem. I don't know, you know, if 0:13:07.920 --> 0:13:11.360 people have started to experience that. I know, I when 0:13:11.360 --> 0:13:14.080 I read that research, I'm like, oh my gosh, I 0:13:14.160 --> 0:13:15.000 am feeling that. 0:13:15.160 --> 0:13:17.719 I feel like the way I work with AI. 0:13:17.679 --> 0:13:21.640 Sometimes I am context switching because, like god forbid, I 0:13:21.679 --> 0:13:24.360 have to wait, you know, thirty or sixty seconds for 0:13:24.640 --> 0:13:27.360 the AI to you know, be in like proper thinking 0:13:27.440 --> 0:13:29.320 votes and vasually takes some time to think about what 0:13:29.320 --> 0:13:30.080 I'm asking it to do. 0:13:30.240 --> 0:13:33.280 But yeah, you've probably experienced brain for as well. 0:13:33.360 --> 0:13:33.560 Yeah. 0:13:33.559 --> 0:13:36.040 The other part of it also is that that constant checking, 0:13:36.120 --> 0:13:38.439 like is that thing that it's given me good? Is 0:13:38.480 --> 0:13:42.960 it actually correct? And you're checking your changing your job 0:13:43.000 --> 0:13:46.600 from being a creator of email or document whatever to 0:13:46.679 --> 0:13:49.600 being effectively a boss and a checker and a policy 0:13:49.720 --> 0:13:51.800 and a reviewer and all those kind of things. And 0:13:52.320 --> 0:13:55.760 you're constantly vigilant as well as constantly between tabs and 0:13:55.800 --> 0:13:57.240 tasks and things you got to do and things you 0:13:57.320 --> 0:13:58.760 got to check. So it's a real thing. 0:14:00.360 --> 0:14:02.319 So's that's a few problems. 0:14:02.360 --> 0:14:04.240 When Nao and I were preparing for this, were like, 0:14:04.320 --> 0:14:06.800 we listed about twenty problems that we're seeing all the time, 0:14:06.960 --> 0:14:10.120 and we just like got the top ones that you know, 0:14:10.720 --> 0:14:13.080 I'm assuming that you can probably relate to at least 0:14:13.120 --> 0:14:13.720 one of those. 0:14:13.840 --> 0:14:18.360 So let's get into solution mode. And again, there's so 0:14:18.480 --> 0:14:18.920 much that. 0:14:18.880 --> 0:14:21.040 We could talk about here, but we've just you know, 0:14:21.240 --> 0:14:22.080 picked out a few. 0:14:22.360 --> 0:14:25.000 Now I am curious, particularly. 0:14:24.520 --> 0:14:27.160 For people that are sitting on a senior leadership team, 0:14:27.320 --> 0:14:29.920 whether you can answer this question. So before you spend 0:14:30.040 --> 0:14:32.520 a dollar on AI. Finished this sentence, we are using AI. 0:14:33.040 --> 0:14:33.280 Two. 0:14:33.600 --> 0:14:37.280 Okay, this is this is your why why you're doing AI. 0:14:37.560 --> 0:14:40.640 And a reason is not because well everyone's doing AI, 0:14:40.920 --> 0:14:42.640 like that's the thing that you do. It's not a reason. 0:14:42.640 --> 0:14:44.560 It's so it's a reason. I guess it's not a 0:14:44.600 --> 0:14:47.720 compelling one, no great one. So you need to be 0:14:47.760 --> 0:14:50.480 clear on your why, because everything else when it comes 0:14:50.640 --> 0:14:53.560 to your AI strategy will follow. 0:14:53.840 --> 0:14:57.280 Right doing AI to you free up time. 0:14:57.920 --> 0:15:01.440 To really innovate, to service your customers better to maybe 0:15:01.600 --> 0:15:04.840 let your staff work god for be like a thirty 0:15:04.880 --> 0:15:05.800 eight hour week. 0:15:06.200 --> 0:15:10.840 So be really clear on your why. Now another solution near. 0:15:11.160 --> 0:15:13.720 We'll have to mention this one earlier, which is we 0:15:13.760 --> 0:15:16.200 often start with and I encourage you start with the 0:15:16.240 --> 0:15:19.800 senior leadership team because that dovetails into the why, but 0:15:19.880 --> 0:15:22.520 also dovetails into the leadership and what we can do 0:15:22.600 --> 0:15:26.800 as an organization with AI, because the finance lead will 0:15:26.840 --> 0:15:30.000 know how they could implement it in finance, the operations likewise, 0:15:30.080 --> 0:15:32.920 et cetera, et cetera, et cetera. So if those leads 0:15:32.960 --> 0:15:37.000 don't understand AI, then there's pretty much no chance that 0:15:37.000 --> 0:15:39.960 they'll be able to lead their teams to be able 0:15:40.000 --> 0:15:44.000 to use this effectively. So we often start there at 0:15:44.040 --> 0:15:46.520 de Lasia teams. Sometimes we start a board, sometimes board 0:15:46.560 --> 0:15:49.840 and leadership team, those kind of things, because what we're 0:15:49.840 --> 0:15:53.440 finding is that the trailblazing leaders are spending up to 0:15:53.480 --> 0:15:56.160 eight hours a week just continually working on AI and 0:15:56.200 --> 0:15:58.680 working with their teams on AI to make sure that 0:15:58.720 --> 0:16:01.880 it's going to be implemented, so it's upskilling themselves, but 0:16:01.960 --> 0:16:03.800 also working with their teams to actually see how it's 0:16:03.840 --> 0:16:06.960 being used at the coal face as well. So the 0:16:07.120 --> 0:16:10.600 leaders using it in their own routines does a couple 0:16:10.640 --> 0:16:13.200 of things. One is keeps them up to date, rather 0:16:13.240 --> 0:16:15.880 than being the Ivory Tower leader who doesn't really know 0:16:15.880 --> 0:16:18.720 about the AI but thinks it's a great idea. Instead, 0:16:18.720 --> 0:16:22.560 they're actually embodying that and they're showing their teams and 0:16:22.600 --> 0:16:26.200 everyone below them. Here's how I'm using it. You should 0:16:26.240 --> 0:16:27.480 be using it too. 0:16:28.080 --> 0:16:30.280 And I think neo because you're like you spent your 0:16:30.440 --> 0:16:34.760 entire career in it and tech like you have seen 0:16:35.160 --> 0:16:40.200 many organizations go through digital transformations, like ALI is really 0:16:40.240 --> 0:16:42.720 different here, and I'm keen to know because like previously, 0:16:42.960 --> 0:16:46.240 a digital transformation, it's like you just delegate that and. 0:16:46.520 --> 0:16:48.920 CEO of a medium sized company if you've got a 0:16:48.960 --> 0:16:52.360 new CRM, like a customer relationship management software, so like 0:16:52.440 --> 0:16:54.800 Salesforce or something like that, where we've got our sales 0:16:54.800 --> 0:16:56.600 and our deals and our products and all those kind 0:16:56.600 --> 0:16:59.560 of things in there. Traditionally, it was get the it 0:16:59.840 --> 0:17:03.360 to department to go and buy it the IT department 0:17:03.400 --> 0:17:05.680 would then buy it, they'd implement it. They do a 0:17:05.720 --> 0:17:07.600 little bit of training for everyone, and then the training 0:17:07.640 --> 0:17:10.679 is done a couple of handholds with a couple of people, 0:17:10.840 --> 0:17:12.960 a couple of manuals there, and then it's sorted. So 0:17:13.240 --> 0:17:15.760 eighty percent of the work is about just putting the 0:17:15.800 --> 0:17:20.080 software in. II is really different because most of the 0:17:20.080 --> 0:17:22.960 work is actually with the training and embodying that within 0:17:23.040 --> 0:17:28.080 your workday. But also this thing changes. This AI thing 0:17:28.200 --> 0:17:30.840 is continually evolving, so it's not a one and done 0:17:30.960 --> 0:17:35.199 training problem. It's actually an ongoing training thing. And so 0:17:35.240 --> 0:17:37.960 if the leads aren't keeping that ongoing themselves, they've got 0:17:38.000 --> 0:17:39.679 no hope of being able to lead that for the 0:17:39.680 --> 0:17:40.679 rest of the organization. 0:17:41.000 --> 0:17:44.399 Now on the topic of training, neo, what should we do? 0:17:44.520 --> 0:17:46.800 And we're a comfort like you know, because we see 0:17:46.840 --> 0:17:49.480 a lot of bad training. We're often bought into fixed 0:17:49.480 --> 0:17:51.760 training that was not particularly effective. 0:17:52.000 --> 0:17:53.439 But yeah, what tell it? Tell us about this? 0:17:53.760 --> 0:17:55.959 A lot of the training, this part of that delegation thing. 0:17:56.200 --> 0:17:59.119 A lot of the leaders are delegating to one department. 0:17:59.160 --> 0:18:02.000 They're saying it this is yours to manage and roll out, 0:18:02.200 --> 0:18:04.719 or they're saying HR it's a learning and development thing, 0:18:04.760 --> 0:18:08.160 so Hi, you're going to roll this thing out. It's 0:18:08.240 --> 0:18:10.680 not just one department to do. And what we find 0:18:10.760 --> 0:18:12.840 is a lot of the training that happens and sometimes 0:18:12.960 --> 0:18:15.120 is with the bigger providers as well. They're very much 0:18:15.119 --> 0:18:18.040 based on features. So here, click this button, you can 0:18:18.359 --> 0:18:20.280 do this, you can do it. Here's deep research, you 0:18:20.280 --> 0:18:22.399 can click that and do some things. But they're not 0:18:22.440 --> 0:18:25.600 showing people why they use it to how they can 0:18:25.760 --> 0:18:27.720 put it in their day job to be actually able 0:18:27.720 --> 0:18:30.879 to change the way they work. So being taught a 0:18:30.880 --> 0:18:33.320 whole bunch of features is great, but not knowing how 0:18:33.320 --> 0:18:36.320 I can improve my day job, improve my outputs, my clients, 0:18:36.359 --> 0:18:40.000 my customers, my internal stakeholders. That's where people get stuck. 0:18:40.240 --> 0:18:42.399 And we find a lot of teams get already trained 0:18:42.680 --> 0:18:45.159 on the features, and then there's not the kind of 0:18:45.200 --> 0:18:47.040 traction that people are seeing who are wanting to see. 0:18:47.080 --> 0:18:49.600 They'll see people are using emails. That's really common. You 0:18:49.600 --> 0:18:51.919 don't need to be trained to use emails, and so 0:18:52.160 --> 0:18:54.240 these get lots of emails and lots of big emails, 0:18:54.320 --> 0:18:57.000 but they're not actually getting better work done. And so 0:18:57.160 --> 0:19:01.119 what we recommend is people train to tasks, not to features. 0:19:01.600 --> 0:19:04.240 So initially there's going to be some kind of acclimation 0:19:04.320 --> 0:19:06.679 here's what the buttons are and whatnot, But then quickly 0:19:06.720 --> 0:19:09.040 it really shu you go into familiarity of the tool 0:19:09.240 --> 0:19:12.040 and how that's going to help me a as a worker, 0:19:12.320 --> 0:19:14.480 and so I really go to tasks that people do 0:19:14.920 --> 0:19:17.359 and particularly if you can then get it to be 0:19:17.440 --> 0:19:21.560 embedded with your workflows as well. So what we found 0:19:21.680 --> 0:19:23.840 is there's actually a bit of a threshold with training 0:19:23.880 --> 0:19:28.120 with different organizations, the researchers out there and saying that 0:19:28.160 --> 0:19:31.639 there's a minimum viable training amounts. They call it the 0:19:31.680 --> 0:19:34.400 minimum dose if you'd like, and it's actually five hours 0:19:34.400 --> 0:19:37.080 of hands on training. So this is not read a manual. 0:19:37.119 --> 0:19:38.960 It's not come to a lunch and learn and watch 0:19:39.000 --> 0:19:42.080 someone do it. It's training where people are being taught 0:19:42.119 --> 0:19:44.480 their features and how to use that, but also taught 0:19:44.520 --> 0:19:46.840 how to put that within their day job and practicing 0:19:46.960 --> 0:19:50.600