WEBVTT - What AI capability really looks like in high-performing teams

0:00:00.640 --> 0:00:05.760
<v Speaker 1>Most organizations measuring their AI progress and measuring the wrong theme.

0:00:06.519 --> 0:00:09.280
<v Speaker 1>They're tracking how many people have a license, how many

0:00:09.320 --> 0:00:13.680
<v Speaker 1>you have done the onboarding training, what percentage are using AI?

0:00:14.000 --> 0:00:18.320
<v Speaker 1>In some form? And on paper it looks like progress,

0:00:18.680 --> 0:00:23.320
<v Speaker 1>but in practice it often isn't because using AI and

0:00:23.440 --> 0:00:28.760
<v Speaker 1>using AI well are two completely different things. One can

0:00:28.760 --> 0:00:32.440
<v Speaker 1>actually make your team less productive because you get buried

0:00:32.440 --> 0:00:35.720
<v Speaker 1>in mediocre output, drowning in AI slop, and none of

0:00:35.760 --> 0:00:39.120
<v Speaker 1>it's moving the needle. So in this episode, Neo and

0:00:39.159 --> 0:00:44.240
<v Speaker 1>I get into what actually separates high performing AI augmented

0:00:44.280 --> 0:00:48.839
<v Speaker 1>teams from the average ones. We cover why adoption metrics

0:00:48.880 --> 0:00:53.240
<v Speaker 1>are almost always vanity metrics, the difference between AI literacy

0:00:53.280 --> 0:00:57.720
<v Speaker 1>and AI leverage, and why most teams nail the first

0:00:58.040 --> 0:01:01.560
<v Speaker 1>and never get to the second. We also talk through

0:01:01.600 --> 0:01:05.039
<v Speaker 1>what it looks like to truly re engineer your workflows,

0:01:05.400 --> 0:01:08.200
<v Speaker 1>not just slot AI into the ones that you already have,

0:01:09.000 --> 0:01:11.320
<v Speaker 1>and by the end you will have a much clearer

0:01:11.440 --> 0:01:14.520
<v Speaker 1>picture of where your team sits and what it would

0:01:14.560 --> 0:01:23.720
<v Speaker 1>take to close the gap. Welcome to how IAI with

0:01:23.880 --> 0:01:28.360
<v Speaker 1>me Doctor Amantha Imber and Neo Applin, head of Inventium AI.

0:01:29.000 --> 0:01:32.119
<v Speaker 1>Each episode we share one practical way to use AI

0:01:32.240 --> 0:01:36.400
<v Speaker 1>better at work and in life. No fluff, no deech jargon,

0:01:36.760 --> 0:01:41.279
<v Speaker 1>just things you can use straight away. So at inventium AI,

0:01:41.480 --> 0:01:44.840
<v Speaker 1>we have worked with one hundreds and hundreds of different

0:01:44.920 --> 0:01:51.840
<v Speaker 1>teams and many organizations around building AI capability, and I

0:01:51.840 --> 0:01:55.240
<v Speaker 1>think there's still a really big difference between what a

0:01:55.360 --> 0:02:00.640
<v Speaker 1>high performing AI augmented team looks like like and what

0:02:00.680 --> 0:02:04.800
<v Speaker 1>an average team looks like. Neo, what are you noticing

0:02:05.000 --> 0:02:07.600
<v Speaker 1>are the biggest difference is because most teams are using

0:02:07.640 --> 0:02:09.960
<v Speaker 1>AI in some shape or form, but how are they

0:02:10.040 --> 0:02:11.840
<v Speaker 1>really high performing ones different.

0:02:13.080 --> 0:02:17.079
<v Speaker 2>They've embedded it within the workday core, so other people

0:02:17.160 --> 0:02:19.760
<v Speaker 2>are using it a little bit, and the poor performing

0:02:19.800 --> 0:02:25.320
<v Speaker 2>teams are actually over using it poorly. So maybe we'll

0:02:25.320 --> 0:02:27.760
<v Speaker 2>start there. So the poor performing teams are the ones

0:02:27.760 --> 0:02:29.960
<v Speaker 2>who are doing AI slop, So they're the ones who

0:02:30.000 --> 0:02:32.200
<v Speaker 2>are looking really efficient. They're creating lots of documents and

0:02:32.200 --> 0:02:33.959
<v Speaker 2>we're creating lots of packs of blah blah blah. They're

0:02:33.960 --> 0:02:36.200
<v Speaker 2>just getting AI to do the lot, and unfortunately they're

0:02:36.240 --> 0:02:39.880
<v Speaker 2>just getting buried in information. So a little bit of training,

0:02:40.760 --> 0:02:43.480
<v Speaker 2>which is basically here's AI, you've got it now, go,

0:02:44.440 --> 0:02:46.320
<v Speaker 2>is actually the worst part of training you can do

0:02:47.200 --> 0:02:50.160
<v Speaker 2>showing people how to correctly use it. Then you're actually

0:02:50.160 --> 0:02:54.200
<v Speaker 2>getting some benefits, but then generally it becomes I make

0:02:54.240 --> 0:02:57.320
<v Speaker 2>a joke where it's my dad brought my grandma when

0:02:57.360 --> 0:03:01.080
<v Speaker 2>you're still alive, a microwave and she for one thing only,

0:03:01.200 --> 0:03:04.880
<v Speaker 2>which was defrosting soup. AI is being used the same

0:03:04.919 --> 0:03:07.040
<v Speaker 2>people know how to use it that one way, and

0:03:07.120 --> 0:03:10.440
<v Speaker 2>so therefore it's being used that one way every single time,

0:03:10.440 --> 0:03:12.560
<v Speaker 2>every single week, so they're not opening up the breadth.

0:03:12.840 --> 0:03:14.840
<v Speaker 2>The big changes that happen when you actually embed it

0:03:14.840 --> 0:03:17.000
<v Speaker 2>in your work date where you're using it in different

0:03:17.040 --> 0:03:20.400
<v Speaker 2>ways to be able to solve problems, to do things better,

0:03:20.639 --> 0:03:24.320
<v Speaker 2>to do things faster, and that's where you get the benefits.

0:03:25.040 --> 0:03:27.640
<v Speaker 2>But that's where everyone needs to be using it for

0:03:27.800 --> 0:03:30.359
<v Speaker 2>similar kind of things and so that everyone is using

0:03:30.440 --> 0:03:32.880
<v Speaker 2>AI to all get better, to get stronger, to get

0:03:32.919 --> 0:03:35.520
<v Speaker 2>a better result. And that's where the higher performing teams work.

0:03:35.760 --> 0:03:38.560
<v Speaker 2>It's not just about using the tool, it's using it

0:03:38.600 --> 0:03:41.480
<v Speaker 2>well so that you're delivering for your team and for

0:03:41.520 --> 0:03:42.680
<v Speaker 2>your customers and clients.

0:03:43.400 --> 0:03:46.680
<v Speaker 1>And this is where I see leaders getting their metrics

0:03:46.720 --> 0:03:50.800
<v Speaker 1>around AAI a little bit wrong. A little bit kind

0:03:50.840 --> 0:03:54.040
<v Speaker 1>of vanity metrics as I would think of them, where

0:03:54.920 --> 0:03:57.680
<v Speaker 1>I would say a lot of organizations that we encounter

0:03:58.040 --> 0:04:01.680
<v Speaker 1>will say we've got an adoption target. So for example,

0:04:01.760 --> 0:04:05.440
<v Speaker 1>we want you know, eighty percent of our workforce to

0:04:05.640 --> 0:04:10.440
<v Speaker 1>have adopted AI, which basically means using AI or maybe

0:04:10.840 --> 0:04:16.440
<v Speaker 1>you know, put through some internally created perhaps not that

0:04:16.720 --> 0:04:21.400
<v Speaker 1>useful training in how to use AI, certainly using it responsibly.

0:04:21.800 --> 0:04:25.240
<v Speaker 1>And if that is your target around how well you

0:04:25.279 --> 0:04:28.680
<v Speaker 1>are going with AI that people are using it, that

0:04:29.320 --> 0:04:33.000
<v Speaker 1>target is just so way off because you can easily

0:04:34.720 --> 0:04:37.600
<v Speaker 1>actually get a whole lot worse when it comes to productivity.

0:04:38.480 --> 0:04:41.080
<v Speaker 1>Let alone, just keep your productivity the same as Neo

0:04:41.240 --> 0:04:46.159
<v Speaker 1>was saying, So Neo, can you share how with inventim

0:04:46.200 --> 0:04:50.200
<v Speaker 1>AI's clients where we take quite a different approach because

0:04:50.240 --> 0:04:53.120
<v Speaker 1>we see this in two parts in terms of first, yes,

0:04:53.160 --> 0:04:56.680
<v Speaker 1>there absolutely is literacy, but then there's leverage. Can you

0:04:57.040 --> 0:04:57.880
<v Speaker 1>talk a little bit about that.

0:04:58.640 --> 0:05:01.440
<v Speaker 2>Yeah, literacy first off is what are these tools, how

0:05:01.520 --> 0:05:03.159
<v Speaker 2>they are different, how do we get the best out

0:05:03.200 --> 0:05:05.880
<v Speaker 2>of those things, and how do we avoid things like

0:05:05.920 --> 0:05:09.320
<v Speaker 2>aislop And that is really important. That is your foundation.

0:05:09.760 --> 0:05:11.680
<v Speaker 2>You need to have a solid foundation on what is

0:05:11.760 --> 0:05:13.760
<v Speaker 2>this thing? And then how's it going to work? And

0:05:13.800 --> 0:05:15.719
<v Speaker 2>some of these things are different for leadership teams, and

0:05:15.720 --> 0:05:18.880
<v Speaker 2>we run leadership teams and boards through this, So like

0:05:18.880 --> 0:05:21.280
<v Speaker 2>if you're setting policies, then you need to know what

0:05:21.320 --> 0:05:23.440
<v Speaker 2>these tools are, how they work, so you can then

0:05:23.520 --> 0:05:26.240
<v Speaker 2>create the policy. But also for knowledge workers, which is

0:05:26.240 --> 0:05:29.039
<v Speaker 2>the bulk of our work, to know how and when

0:05:29.120 --> 0:05:32.040
<v Speaker 2>to use it is really caught. The leverage comes in

0:05:32.120 --> 0:05:34.640
<v Speaker 2>where it's how do I embed this in my day job?

0:05:34.920 --> 0:05:36.799
<v Speaker 2>How do I speed up the things that I'm doing?

0:05:37.320 --> 0:05:40.120
<v Speaker 2>And how do I get a better result? Because sometimes

0:05:40.200 --> 0:05:44.040
<v Speaker 2>speeding up isn't the most important thing. Sometimes better result,

0:05:44.320 --> 0:05:48.000
<v Speaker 2>but slower is better, or maybe a more customer centric

0:05:48.040 --> 0:05:51.160
<v Speaker 2>approach to things is better. So how do I use

0:05:51.360 --> 0:05:53.800
<v Speaker 2>this brand new tool. We've got to be able to

0:05:53.839 --> 0:05:56.240
<v Speaker 2>get the benefit. And it's not about let's get the

0:05:56.240 --> 0:05:58.480
<v Speaker 2>same process we've always had and then get AI to

0:05:58.640 --> 0:06:02.760
<v Speaker 2>step into one of those steps. Yep, that's an interim step.

0:06:02.839 --> 0:06:06.200
<v Speaker 2>That's where a lot of organizations go. It's more about

0:06:06.480 --> 0:06:09.200
<v Speaker 2>now we've got this tool, now we understand what it is,

0:06:09.560 --> 0:06:12.160
<v Speaker 2>how do we reinvent how we do our work, how

0:06:12.200 --> 0:06:15.200
<v Speaker 2>we deliver our services, now we've got this amazing tool,

0:06:15.279 --> 0:06:18.880
<v Speaker 2>and that's where the organizations really go better, and that's

0:06:19.000 --> 0:06:21.760
<v Speaker 2>using the leverage of AI in a new way, because

0:06:21.760 --> 0:06:23.480
<v Speaker 2>this is a new tool we've never had before.

0:06:24.440 --> 0:06:27.960
<v Speaker 1>Can you give an example maybe of a team that

0:06:28.000 --> 0:06:33.359
<v Speaker 1>you've worked with where you've really couldn't double down on going, okay,

0:06:33.360 --> 0:06:37.400
<v Speaker 1>this is how you leave reach AI in things like

0:06:37.440 --> 0:06:38.479
<v Speaker 1>you're doing every day.

0:06:39.160 --> 0:06:41.560
<v Speaker 2>Here's a really simple and we see this a lot

0:06:41.560 --> 0:06:47.039
<v Speaker 2>of places. Back in the day, managers look at a

0:06:47.040 --> 0:06:49.040
<v Speaker 2>spreadsheet and they'd go, I need to get some insights

0:06:49.040 --> 0:06:52.799
<v Speaker 2>out of this. But maybe managers weren't the pivot table experts.

0:06:52.839 --> 0:06:55.320
<v Speaker 2>Maybe they weren't great with Excel formulas because you didn't

0:06:55.360 --> 0:06:57.880
<v Speaker 2>need to be because you'd have that awesome person in

0:06:57.920 --> 0:07:02.400
<v Speaker 2>your department, let's call it. You go to Susie and say, hey, look,

0:07:02.400 --> 0:07:04.760
<v Speaker 2>can you please do some analysis on this spreadsheet for me?

0:07:05.520 --> 0:07:09.000
<v Speaker 2>Now you could say, why don't we get Susie with

0:07:09.080 --> 0:07:12.040
<v Speaker 2>AI so Susie can do it faster. That's one way

0:07:12.080 --> 0:07:13.520
<v Speaker 2>to do it, and certainly a lot of people doing that.

0:07:13.760 --> 0:07:16.520
<v Speaker 2>We're teaching a lot of Susie's and David's and Justin's

0:07:16.800 --> 0:07:19.200
<v Speaker 2>to be able to do that as well, But wouldn't

0:07:19.200 --> 0:07:22.560
<v Speaker 2>it be better if we could get AI to be

0:07:22.640 --> 0:07:25.440
<v Speaker 2>there to be able to service the manager, so everyone

0:07:25.480 --> 0:07:28.040
<v Speaker 2>then can self service rather than going to Susie. It's

0:07:28.080 --> 0:07:31.280
<v Speaker 2>then I can ask questions of the AI, so either

0:07:31.400 --> 0:07:33.720
<v Speaker 2>I can get that information out or even better, I

0:07:33.720 --> 0:07:35.960
<v Speaker 2>can get the report that I need in the way

0:07:35.960 --> 0:07:38.720
<v Speaker 2>that I need every single time. So it's not so

0:07:38.800 --> 0:07:41.240
<v Speaker 2>much me just ask questions of it. It's I can

0:07:41.320 --> 0:07:44.000
<v Speaker 2>just feed it a spreadsheet. It knows exactly what I'm

0:07:44.040 --> 0:07:46.200
<v Speaker 2>after in a particular way I'm after because I've either

0:07:46.200 --> 0:07:48.680
<v Speaker 2>built a workflow or built an agent or a GPT

0:07:48.880 --> 0:07:52.120
<v Speaker 2>or a GEM or a project that then deliver me

0:07:52.480 --> 0:07:55.560
<v Speaker 2>exactly what I need every single time. So it's an

0:07:55.560 --> 0:07:58.880
<v Speaker 2>easy way to re envisage the way that we're doing things.

0:07:59.200 --> 0:08:01.720
<v Speaker 2>But these are all embedded with so many different teams

0:08:01.840 --> 0:08:05.520
<v Speaker 2>we've got, Like here's an example of like a hotel,

0:08:07.120 --> 0:08:09.080
<v Speaker 2>you get lots of Q and A coming through to

0:08:09.480 --> 0:08:11.720
<v Speaker 2>your concierge desk. You know what times a pool open

0:08:11.760 --> 0:08:13.760
<v Speaker 2>and if I made a reservation, can I cancel it?

0:08:13.800 --> 0:08:15.240
<v Speaker 2>And how of far and all those kind of things,

0:08:15.280 --> 0:08:17.320
<v Speaker 2>So you could have to have that information in your

0:08:17.320 --> 0:08:19.800
<v Speaker 2>brain or you can get an agent to be able

0:08:19.840 --> 0:08:23.200
<v Speaker 2>to craft those emails for you for your customer, or

0:08:23.240 --> 0:08:25.960
<v Speaker 2>you can even build a workflow where AI can help

0:08:26.000 --> 0:08:28.160
<v Speaker 2>you to do that kind of thing automatically and it

0:08:28.240 --> 0:08:32.880
<v Speaker 2>drops out the hard ones that you need human to process.

0:08:32.880 --> 0:08:34.960
<v Speaker 2>So maybe it's like eighty twenty eighty percent has been

0:08:35.000 --> 0:08:37.880
<v Speaker 2>automated and twenty percent has been taken care of by people.

0:08:38.160 --> 0:08:40.520
<v Speaker 2>It's now what we've got with this tool, how do

0:08:40.559 --> 0:08:43.480
<v Speaker 2>we solve that problem better and perhaps go closer to

0:08:43.520 --> 0:08:46.000
<v Speaker 2>the source or closer to the person who needs to

0:08:46.120 --> 0:08:48.559
<v Speaker 2>use it. And that's the real leverage.

0:08:49.160 --> 0:08:52.760
<v Speaker 1>And what we're certainly finding and recommending to our clients

0:08:53.160 --> 0:08:57.439
<v Speaker 1>is that, yes, everyone needs a certain amount of literacy

0:08:57.559 --> 0:09:01.680
<v Speaker 1>about AI so that they're getting good output rather than

0:09:01.840 --> 0:09:06.120
<v Speaker 1>just proliferating the organization with aislop, which is going to

0:09:06.160 --> 0:09:10.839
<v Speaker 1>have really detrimental impacts on productivity, reputation, all sorts of things.

0:09:11.200 --> 0:09:14.000
<v Speaker 1>But then what we're typically doing is we're working with

0:09:14.400 --> 0:09:18.480
<v Speaker 1>AI champions and these are spread across different functional groups

0:09:18.520 --> 0:09:21.920
<v Speaker 1>because you need to have that more advanced understanding of

0:09:21.920 --> 0:09:25.440
<v Speaker 1>what AI is capable of in order to get those benefits,

0:09:25.520 --> 0:09:31.320
<v Speaker 1>because most people are not like workflow architects where they

0:09:31.360 --> 0:09:34.559
<v Speaker 1>know how to identify what are the different workflows that

0:09:34.720 --> 0:09:38.600
<v Speaker 1>this individual or this group goes through daily or weekly,

0:09:38.760 --> 0:09:41.520
<v Speaker 1>And then how do you even map out a workflow?

0:09:41.720 --> 0:09:44.679
<v Speaker 1>Like most people don't have those skills, and that is

0:09:44.720 --> 0:09:47.960
<v Speaker 1>a really critical skill set to have in order to

0:09:48.160 --> 0:09:51.040
<v Speaker 1>really leave reach AI. If you can't do that, it's

0:09:51.120 --> 0:09:53.480
<v Speaker 1>really hard to access the gains.

0:09:53.720 --> 0:09:56.959
<v Speaker 2>And it's even harder to go mapping. It is one thing,

0:09:57.040 --> 0:09:59.880
<v Speaker 2>one skill to then go how do I re envisage?

0:10:00.280 --> 0:10:03.640
<v Speaker 2>That's a completely different skill as well, And that kind

0:10:03.640 --> 0:10:05.200
<v Speaker 2>of takes a bit of a bit of a blue

0:10:05.200 --> 0:10:08.280
<v Speaker 2>sky thinker, someone who understands the business and also can

0:10:08.360 --> 0:10:11.360
<v Speaker 2>understand what the technology can do. And so yeah, we

0:10:11.440 --> 0:10:13.840
<v Speaker 2>do definitely help these kind of people to build up

0:10:13.880 --> 0:10:16.160
<v Speaker 2>those skills. We also build a bunch of tools to

0:10:16.200 --> 0:10:18.520
<v Speaker 2>help them to help themselves with AI as well.

0:10:19.880 --> 0:10:23.800
<v Speaker 1>So I would say, if you are listening and relating

0:10:24.040 --> 0:10:27.560
<v Speaker 1>to having some sort of a metric in your organization

0:10:28.559 --> 0:10:30.960
<v Speaker 1>or maybe even just for yourself, to go, yes, have

0:10:31.000 --> 0:10:33.800
<v Speaker 1>I got people using AI? Have I given them licenses?

0:10:33.800 --> 0:10:36.760
<v Speaker 1>So have I put them through some basic training? I

0:10:36.800 --> 0:10:39.680
<v Speaker 1>would say it's really problematic if you are stopping there

0:10:39.800 --> 0:10:43.559
<v Speaker 1>so I would challenge any listener to think about how

0:10:43.559 --> 0:10:48.200
<v Speaker 1>can you really leaverage AI? How can you unpack your workflows?

0:10:48.280 --> 0:10:52.120
<v Speaker 1>How can you really think deeply around where AI is

0:10:52.200 --> 0:10:55.200
<v Speaker 1>best place to help you. Obviously, Neo and I are

0:10:55.280 --> 0:10:57.199
<v Speaker 1>very biased because we do a lot of this training

0:10:57.240 --> 0:11:01.480
<v Speaker 1>at Inventium AI, but by all means go anywhere, but

0:11:01.640 --> 0:11:04.560
<v Speaker 1>make sure you are doing it to unlock the full

0:11:04.840 --> 0:11:08.319
<v Speaker 1>benefits of AI whatever your job looks like.

0:11:09.360 --> 0:11:13.520
<v Speaker 2>And just kind of leveraging that. I've read some studies

0:11:13.640 --> 0:11:17.040
<v Speaker 2>a couple of days ago. The minimum viable dose for

0:11:17.200 --> 0:11:21.560
<v Speaker 2>AI training is five hours minimum viable. Anything less than

0:11:21.600 --> 0:11:24.000
<v Speaker 2>that you're not going to get any benefit in the workplace.

0:11:24.400 --> 0:11:27.880
<v Speaker 2>But that is not about leverage. That is just about baseline.

0:11:28.040 --> 0:11:30.880
<v Speaker 2>Can I use this tool? There are companies that spend

0:11:30.920 --> 0:11:33.520
<v Speaker 2>up to believe it or not, eighty one hours on

0:11:33.559 --> 0:11:36.320
<v Speaker 2>their teams, and what that's about is is what Amantha

0:11:36.360 --> 0:11:38.400
<v Speaker 2>and I are talking about, which is using the skills

0:11:38.520 --> 0:11:42.040
<v Speaker 2>in your workplace to re envisage your workflows and help

0:11:42.080 --> 0:11:44.720
<v Speaker 2>your clients even better than you did before. That's where

0:11:44.720 --> 0:11:46.240
<v Speaker 2>the benefit comes from.

0:11:46.800 --> 0:11:49.200
<v Speaker 1>And Neo, I'm going to do a shameless plug because

0:11:49.240 --> 0:11:51.480
<v Speaker 1>we do have a new program that We've just launched

0:11:51.520 --> 0:11:54.800
<v Speaker 1>an inventium called the AI Agent boot Camp, which is

0:11:54.840 --> 0:11:57.880
<v Speaker 1>designed to solve this very problem where we will be

0:11:58.360 --> 0:12:02.640
<v Speaker 1>taking people through how do you become a workflow architect

0:12:02.760 --> 0:12:06.760
<v Speaker 1>and how do you really double down on understanding where

0:12:06.760 --> 0:12:11.680
<v Speaker 1>can AI help you augment or streamline those workflows get

0:12:11.679 --> 0:12:17.160
<v Speaker 1>massive productivity savings but also get far better output at

0:12:17.200 --> 0:12:18.040
<v Speaker 1>the end of the day.

0:12:18.440 --> 0:12:20.640
<v Speaker 2>Yeah, it's a fun day. And also you bring one

0:12:20.679 --> 0:12:22.760
<v Speaker 2>of your problems and we will help you solve those

0:12:22.760 --> 0:12:25.400
<v Speaker 2>problems with agents during the day, so you've actually got

0:12:25.440 --> 0:12:26.760
<v Speaker 2>us and a whole bunch of AI tools to be

0:12:26.800 --> 0:12:27.960
<v Speaker 2>able to help you with that. So yeah, it's good

0:12:27.960 --> 0:12:28.400
<v Speaker 2>fun day.

0:12:28.600 --> 0:12:30.920
<v Speaker 1>And there is a link to that in the show notes.

0:12:31.840 --> 0:12:35.760
<v Speaker 1>How IAI was hosted by me, Amantha Imber and Neo Applan.

0:12:36.040 --> 0:12:37.960
<v Speaker 1>A big thank you to Martin Imba who does our

0:12:38.080 --> 0:12:42.040
<v Speaker 1>sound editing, and Jim Rubio for production support, and thank

0:12:42.080 --> 0:12:44.680
<v Speaker 1>you to John Kilby who composed the theme music.