Latest / How I Work / Why most AI rollouts fail (and how to make sure yours doesn't)
Transcript
- 0:00Speaker 1: Most AI rollouts looks something like this, a big announcement,
- 0:06Speaker 1: lots of enthusiasm, and a few months later things have
- 0:09Speaker 1: gone one of two ways, either barely anyone's using the tools,
- 0:14Speaker 1: or everyone's using them to churn out AI slop. Sometimes
- 0:18Speaker 1: both and leadless to say, the huge investment in licenses
- 0:23Speaker 1: isn't paying off. Now, for today's episode, we are releasing
- 0:28Speaker 1: a recording from a live webinar Neo and I ran
- 0:32Speaker 1: recently on why AI rollouts fail and how to set
- 0:36Speaker 1: yours up for success. We get into why these problems
- 0:40Speaker 1: are happening, including the surprising research on how little time
- 0:45Speaker 1: AI is actually saving most people, the training goes that
- 0:48Speaker 1: does move the dial, and the question every leadership team
- 0:52Speaker 1: needs to answer before spending another dollar on licenses. Well,
- 0:57Speaker 1: let's jump in by starting with the biggest problem we
- 1:00Speaker 1: are seeing organizations making Welcome to how IAI with me
- 1:11Speaker 1: Doctor Amantha Imba and Neo Applin, head of Inventium AI.
- 1:15Speaker 1: Each episode we share one practical way to use AI
- 1:19Speaker 1: better at.
- 1:19Speaker 2: Work and in life.
- 1:21Speaker 1: No fluff, no tech jargon, just things you can use
- 1:24Speaker 1: straight away. We are of course talking to you today
- 1:28Speaker 1: about AI rollouts and why we've seen a lot of
- 1:33Speaker 1: them fail and how we can actually make them work.
- 1:37Speaker 1: So to start with, you might be like, oh my god,
- 1:41Speaker 1: who are these randoms? Or maybe like you followed us
- 1:45Speaker 1: for years and you're like you need no introduction. But anyway,
- 1:48Speaker 1: I'm going to assume that, like we're two random people,
- 1:50Speaker 1: So I'll introduce myself first. So my name is Amantha
- 1:54Speaker 1: Imber and with my initials, well, it just qualifies me
- 1:57Speaker 1: to be.
- 1:57Speaker 2: Talking about AI, right, So what more do I need? Anyway?
- 2:01Speaker 1: Aside from that, I'm the founder of Inventium and Inventium
- 2:04Speaker 1: dot AI, which is our sister company where all we
- 2:07Speaker 1: do is help people with AI skills, building and rollouts
- 2:10Speaker 1: and all that sort of great stuff. I was awarded
- 2:12Speaker 1: AI Consultant of the Year last year, so there we go.
- 2:15Speaker 1: I'm award winning, so hey, you should probably trust me.
- 2:18Speaker 2: Neo, who are you?
- 2:19Speaker 3: I'm neo and matrix jokes going in here. So I've
- 2:23Speaker 3: been in it my whole career. I've been building stuff
- 2:25Speaker 3: for big companies, small companies, startup, scale ups and that
- 2:28Speaker 3: kind of stuff. But AI has been the biggest change
- 2:31Speaker 3: I've seen over the time. So AI is all I
- 2:33Speaker 3: do these days. I'm the head of AI and invent him.
- 2:36Speaker 3: AI and I trained companies on how to get the
- 2:39Speaker 3: best out of this, and I usually start at leadership
- 2:41Speaker 3: teams and boards and things like that and then go
- 2:43Speaker 3: throughout the rest of the company to be able to
- 2:45Speaker 3: implement AI awesome.
- 2:47Speaker 2: So together Neo and I and the team at.
- 2:51Speaker 1: Inventium, we have worked with a lot of different organizations.
- 2:54Speaker 2: These are just a smattering.
- 2:57Speaker 1: Who we have helped build AI capability, help them with
- 3:00Speaker 1: roll outs, all that sort of great stuff. So today,
- 3:03Speaker 1: like we're talking from firsthand on the ground experience plus
- 3:07Speaker 1: all the research that AI has of course helped us
- 3:10Speaker 1: get across.
- 3:11Speaker 2: So let's start. Let's start with a really big problem.
- 3:13Speaker 3: Great, So first problem is actually top down. A lot
- 3:17Speaker 3: of leadership teams are saying this AI thing, Yeah, we'll
- 3:20Speaker 3: just get someone else to take care of that. So
- 3:23Speaker 3: it is a problem that needs to be solved at
- 3:25Speaker 3: the leadership level, and it needs to be a problem
- 3:28Speaker 3: that needs to be solved as far as here's what
- 3:30Speaker 3: we need to have for our company, but also needs
- 3:32Speaker 3: to be embodied by the leads as well. So one
- 3:36Speaker 3: of the challenges we've seen is the company might have
- 3:39Speaker 3: an AI policy, say, but everyone meets the policy and goes, oh, no,
- 3:43Speaker 3: I think you can use it. No, I can't use
- 3:45Speaker 3: it for that, And people get confused on what you
- 3:47Speaker 3: can actually use it for because the policy is not
- 3:49Speaker 3: specific about the types of data or the types of
- 3:52Speaker 3: use cases you can use it for. So one department
- 3:54Speaker 3: feels great to be using it, another apartment doesn't use
- 3:57Speaker 3: it much at all. That's one of the challenges there.
- 4:00Speaker 3: So the AI strategy is the first challenge, but within
- 4:03Speaker 3: that strategy challenge, the CEO needs to know what these
- 4:08Speaker 3: AI tools are, where the warts are, where they wanted
- 4:11Speaker 3: to go, where they think it's going to actually change
- 4:13Speaker 3: your company. And we found a lot of leadership teams
- 4:17Speaker 3: will say, oh, that AI thing, will get that done
- 4:19Speaker 3: by it department, but the CEO themselves don't use it.
- 4:22Speaker 3: They don't get their leadership team to use it, so
- 4:25Speaker 3: everyone else in the company is a bit scared to
- 4:27Speaker 3: use it as well. I even had one lead who
- 4:30Speaker 3: said every time anyone walked behind them, if they had
- 4:33Speaker 3: AI opened on their laptop, they'd all tab out of it,
- 4:36Speaker 3: which is an absolute wrong way to do it. You
- 4:37Speaker 3: should be saying this is how I've used AI, and
- 4:40Speaker 3: you should be able to use it in these kind
- 4:42Speaker 3: of ways too, So we're going to walk the talk
- 4:44Speaker 3: here with AI. Otherwise a whole bunch of your staff
- 4:47Speaker 3: are going to be too scared to use the damn thing.
- 4:51Speaker 1: Now, let's move on to another problem near I know
- 4:54Speaker 1: you are seeing this a lot as am I.
- 4:57Speaker 3: Yep, Absolutely, buy the licenses, go right ahead. But you've
- 5:02Speaker 3: got to be able to train your people on how
- 5:04Speaker 3: to use these things. Where is it twenty thirty bucks
- 5:07Speaker 3: a month per user? A few times that buy your workforce.
- 5:11Speaker 3: This is not a cheap investment, but you can get
- 5:14Speaker 3: a lot of this back. We think we get a
- 5:16Speaker 3: hell of a lot more than that back return on
- 5:18Speaker 3: investment for us and the people we train. So thirty
- 5:21Speaker 3: bucks is actually pretty cheap for what you can get
- 5:23Speaker 3: out of AI and the people's time savings. But just
- 5:26Speaker 3: giving them AI and saying now go is actually going
- 5:30Speaker 3: to give you less productivity. What I mean by that
- 5:33Speaker 3: is you're going to get more people getting confused with it,
- 5:35Speaker 3: just writing big emails, creating AI slop. So you need
- 5:39Speaker 3: to be able to train people on how to use AI.
- 5:42Speaker 3: Licenses alone is not going to solve the problem.
- 5:47Speaker 2: Now.
- 5:48Speaker 1: Of course, the big promise with AI is that it's
- 5:51Speaker 1: meant to save us all time.
- 5:54Speaker 2: But when we look at the latest research.
- 5:56Speaker 1: Around is coming out of HBr, they found that the
- 5:59Speaker 1: average time saving is just like guess in your head
- 6:03Speaker 1: before I reveal this number, two point five percent, Like
- 6:09Speaker 1: that is nothing, And there are many reasons why this
- 6:14Speaker 1: time saving is not happening. But I mean, what we
- 6:18Speaker 1: are saying, particularly when you give people the licenses without training,
- 6:22Speaker 1: is that you get this huge proliferation of volume of work,
- 6:27Speaker 1: and generally that work is slop. Like you will have people,
- 6:32Speaker 1: you know, typing a prompt into AI, going write me
- 6:35Speaker 1: a ten page report and here are my three bullet points,
- 6:39Speaker 1: and then that report then gets distributed to people people
- 6:42Speaker 1: presumably have to read that report, or even worse, they're
- 6:46Speaker 1: putting that report into AI and going summarize this report
- 6:49Speaker 1: and they're getting back three completely different bullet points. And
- 6:52Speaker 1: then companies are like, ohy am, I not saving time.
- 6:55Speaker 1: Well yeah, next on our problem list. What we see
- 7:01Speaker 1: is a big problem in how companies are measuring the
- 7:06Speaker 1: ROI of their investment, which is obviously like you know,
- 7:09Speaker 1: really it's a decent sized investment. When you think about
- 7:12Speaker 1: licenses and training and doing this stuff properly.
- 7:14Speaker 2: And what we see a lot of companies doing, and.
- 7:17Speaker 1: Particularly like I know, like on sales calls that I'll
- 7:20Speaker 1: be on, I'll be like, well, how's you know, how's
- 7:22Speaker 1: AIU is going?
- 7:23Speaker 3: Now?
- 7:23Speaker 2: Like where are people at with adoption? And they'll go, oh, well,
- 7:27Speaker 2: you know, we can see the x percentage of stuff
- 7:29Speaker 2: for logging in daily to their AI tool, and some
- 7:33Speaker 2: companies is a little bit more advanced, like they're measuring
- 7:35Speaker 2: token usage.
- 7:36Speaker 1: But the thing with this, this tells us nothing whether
- 7:39Speaker 1: people are logging on or not, Like they could be
- 7:41Speaker 1: logging on and creating slop and then logging off, Like
- 7:43Speaker 1: that's not helping your company. So these vanity metrics are
- 7:48Speaker 1: not helping you. We need to think differently about the
- 7:51Speaker 1: metrics that you're using to go how much value is
- 7:55Speaker 1: AI actually giving us?
- 7:57Speaker 2: Okay, next to agents, Yeah.
- 8:01Speaker 3: That's another vantage metric we've seen as well. Some companies
- 8:04Speaker 3: are going great, we're getting more agents created, and we
- 8:07Speaker 3: want more agents created. So, just for those who don't know,
- 8:10Speaker 3: agents are effectively AI experts that do things for you
- 8:14Speaker 3: the same way every single time. So you might have
- 8:16Speaker 3: an agent that might create for you like a little
- 8:20Speaker 3: the way we create our briefing here, or it might
- 8:22Speaker 3: be an agent that does like a style guide checker,
- 8:24Speaker 3: or it might be an agent that helps you with
- 8:25Speaker 3: your email writing. Those kind of things. Everyone's building agents,
- 8:29Speaker 3: and that's great, really good to see. In fact, we
- 8:31Speaker 3: train people on how to build agents. But here's the thing.
- 8:34Speaker 3: If everyone in your company is building agents, you're going
- 8:37Speaker 3: to get one thing, which is a proliferation of agents.
- 8:40Speaker 3: And the other thing is my agent will work differently
- 8:43Speaker 3: to your agent. And so what you're going to find
- 8:45Speaker 3: is that there'll be too many agents out there. They'll
- 8:48Speaker 3: all work a little bit different to each other. And
- 8:50Speaker 3: a lot of people are building agents. Don't know how
- 8:51Speaker 3: to build these things, so they work the great way
- 8:54Speaker 3: every single time. And so it might work great for
- 8:57Speaker 3: me when I use it, when I give it to you,
- 8:59Speaker 3: you might use work different because you use different words
- 9:01Speaker 3: when you created your agent. So those kind of problems.
- 9:04Speaker 3: The other problem is, because we've got so many agents
- 9:07Speaker 3: being created, no one's actually looking after the agent. Who
- 9:10Speaker 3: owns that agent? If it needs to be updated or changed?
- 9:13Speaker 3: Who owns that? How do I find that within my organization?
- 9:16Speaker 3: And so what we've actually got is a whole bunch
- 9:19Speaker 3: of agents out there floating around that may or may
- 9:22Speaker 3: not be giving value to the company. The other thing
- 9:24Speaker 3: is we're going to have a whole bunch of legacy problems.
- 9:27Speaker 3: Like if I'm got this old agent, if anyone hasn't
- 9:29Speaker 3: used it in two months, can I delete it? If
- 9:32Speaker 3: I do delete it, does that break someone's workflow? I
- 9:34Speaker 3: don't know, because all these agents are just been thrown around.
- 9:37Speaker 3: It's one of the big problems we've got. So agents
- 9:40Speaker 3: are great, but too many agents not done particularly well.
- 9:44Speaker 3: Isn't innovation. It actually is more confusing than anything else.
- 9:48Speaker 1: And interestingly, like on the topic of vanity metrics, I
- 9:52Speaker 1: certainly hear this a lot, particularly when I'm speaking with
- 9:55Speaker 1: leaders who are looking at AI rollouts, and they'd be like, yeah, yeah,
- 9:58Speaker 1: I've got like heaps of people they're bill agents. And
- 10:01Speaker 1: I always wonder, hmm, what do you mean by that,
- 10:04Speaker 1: because like, there's good agent building and there's crap agent building.
- 10:09Speaker 1: If you've spent maybe like five or ten minutes like
- 10:12Speaker 1: putting a system prompt into like agents on Copilot or
- 10:15Speaker 1: a GPT on GPT, and then you're like, Bam, agent created,
- 10:21Speaker 1: it's probably not the best agent because like Neo.
- 10:24Speaker 2: How long do you take to build an agent?
- 10:26Speaker 3: Typically half a day a day. Yeah, and that's because
- 10:30Speaker 3: I'm building it, I'm refining it, I'm testing it, I'm
- 10:33Speaker 3: trying to break it. I'm giving it different inputs, different
- 10:35Speaker 3: outputs need to be put in. I need to make
- 10:37Speaker 3: sure that this works reliably for me and everyone else
- 10:40Speaker 3: I give it to every single time. And the other
- 10:43Speaker 3: thing is I then need to manage and version control
- 10:45Speaker 3: my agents because I'm going to come back to these
- 10:48Speaker 3: ones when the engine changes, when it goes from chetpt
- 10:51Speaker 3: five point three to five point four sometimes day break.
- 10:54Speaker 3: Same thing with Copilot and all the others. So yeah,
- 10:57Speaker 3: agents do take time to build correctly.
- 11:00Speaker 1: I'm curious, folks at your workplace, who's working in an
- 11:03Speaker 1: organization where like there's a bit of a vibe that like, oh,
- 11:07Speaker 1: is AI going.
- 11:07Speaker 2: To replace us?
- 11:08Speaker 1: Like just give us like a thumbs up if that
- 11:11Speaker 1: is your organization.
- 11:14Speaker 2: Yeah, I can see LA lots of them's going up now.
- 11:16Speaker 1: Okay, So I question, like, why would anyone adopt a
- 11:19Speaker 1: tool that they think is there to replace them? This
- 11:21Speaker 1: is a culture problem that very much needs to be
- 11:25Speaker 1: managed from the top, because you know, if I'm working
- 11:28Speaker 1: in an organization, I'm like, Oh, I think AI is
- 11:31Speaker 1: going to replace me? How excited am I going to
- 11:33Speaker 1: be to use the technology and really embrace it?
- 11:36Speaker 2: Probably not much at all. Now. The final big.
- 11:40Speaker 1: Problem that we're going to talk about before we get
- 11:42Speaker 1: into solutions and then we'll get into your questions is
- 11:45Speaker 1: brain fry. So I know that this has kind of
- 11:48Speaker 1: been headlines in AI news.
- 11:51Speaker 2: If you're like us.
- 11:53Speaker 1: Inclined and like just you know, follow this stuff every
- 11:56Speaker 1: single day. Brain fry was something that I was really
- 11:59Speaker 1: awesome pair of research and published in HBr, led by
- 12:03Speaker 1: this amazing researcher Gabrielle Obrezen Kellerman, who I've actually had
- 12:08Speaker 1: on how I.
- 12:08Speaker 2: Work a couple of times. But brain fry is what happens.
- 12:11Speaker 1: And I'm curious as to anyone that has experience to this.
- 12:16Speaker 1: When you have got multiple let's call them, like, you know,
- 12:20Speaker 1: multiple AI tabs tasks going on at the same time.
- 12:24Speaker 1: Maybe you know, you asked chat GPT to do one
- 12:27Speaker 1: thing and then you've opened up another tab and you've
- 12:30Speaker 1: got it working on another task and then another. And
- 12:33Speaker 1: this is how a lot of people are working with AI,
- 12:35Speaker 1: and people are getting to the end of the day,
- 12:37Speaker 1: and it's different from burnout. They are feeling like their
- 12:40Speaker 1: brain is fried, like they have intense cognitive fatigue and
- 12:46Speaker 1: not surprisingly, like when we are fatigued like that, it
- 12:49Speaker 1: causes us to make more errors. It causes us to
- 12:52Speaker 1: obviously produce like lower quality work. It is not good
- 12:56Speaker 1: for the output we are delivering in our job. So
- 13:00Speaker 1: different to burn out, which is a little bit more
- 13:01Speaker 1: emotional and some other aspects. But AI brain fright is
- 13:05Speaker 1: a really big problem. I don't know, you know, if
- 13:07Speaker 1: people have started to experience that. I know, I when
- 13:11Speaker 1: I read that research, I'm like, oh my gosh, I
- 13:14Speaker 1: am feeling that.
- 13:15Speaker 2: I feel like the way I work with AI.
- 13:17Speaker 1: Sometimes I am context switching because, like god forbid, I
- 13:21Speaker 1: have to wait, you know, thirty or sixty seconds for
- 13:24Speaker 1: the AI to you know, be in like proper thinking
- 13:27Speaker 1: votes and vasually takes some time to think about what
- 13:29Speaker 1: I'm asking it to do.
- 13:30Speaker 2: But yeah, you've probably experienced brain for as well.
- 13:33Speaker 1: Yeah.
- 13:33Speaker 3: The other part of it also is that that constant checking,
- 13:36Speaker 3: like is that thing that it's given me good? Is
- 13:38Speaker 3: it actually correct? And you're checking your changing your job
- 13:43Speaker 3: from being a creator of email or document whatever to
- 13:46Speaker 3: being effectively a boss and a checker and a policy
- 13:49Speaker 3: and a reviewer and all those kind of things. And
- 13:52Speaker 3: you're constantly vigilant as well as constantly between tabs and
- 13:55Speaker 3: tasks and things you got to do and things you
- 13:57Speaker 3: got to check. So it's a real thing.
- 14:00Speaker 2: So's that's a few problems.
- 14:02Speaker 1: When Nao and I were preparing for this, were like,
- 14:04Speaker 1: we listed about twenty problems that we're seeing all the time,
- 14:06Speaker 1: and we just like got the top ones that you know,
- 14:10Speaker 1: I'm assuming that you can probably relate to at least
- 14:13Speaker 1: one of those.
- 14:13Speaker 2: So let's get into solution mode. And again, there's so
- 14:18Speaker 2: much that.
- 14:18Speaker 1: We could talk about here, but we've just you know,
- 14:21Speaker 1: picked out a few.
- 14:22Speaker 2: Now I am curious, particularly.
- 14:24Speaker 1: For people that are sitting on a senior leadership team,
- 14:27Speaker 1: whether you can answer this question. So before you spend
- 14:30Speaker 1: a dollar on AI. Finished this sentence, we are using AI.
- 14:33Speaker 3: Two.
- 14:33Speaker 2: Okay, this is this is your why why you're doing AI.
- 14:37Speaker 1: And a reason is not because well everyone's doing AI,
- 14:40Speaker 1: like that's the thing that you do. It's not a reason.
- 14:42Speaker 1: It's so it's a reason. I guess it's not a
- 14:44Speaker 1: compelling one, no great one. So you need to be
- 14:47Speaker 1: clear on your why, because everything else when it comes
- 14:50Speaker 1: to your AI strategy will follow.
- 14:53Speaker 2: Right doing AI to you free up time.
- 14:57Speaker 1: To really innovate, to service your customers better to maybe
- 15:01Speaker 1: let your staff work god for be like a thirty
- 15:04Speaker 1: eight hour week.
- 15:06Speaker 2: So be really clear on your why. Now another solution near.
- 15:11Speaker 3: We'll have to mention this one earlier, which is we
- 15:13Speaker 3: often start with and I encourage you start with the
- 15:16Speaker 3: senior leadership team because that dovetails into the why, but
- 15:19Speaker 3: also dovetails into the leadership and what we can do
- 15:22Speaker 3: as an organization with AI, because the finance lead will
- 15:26Speaker 3: know how they could implement it in finance, the operations likewise,
- 15:30Speaker 3: et cetera, et cetera, et cetera. So if those leads
- 15:32Speaker 3: don't understand AI, then there's pretty much no chance that
- 15:37Speaker 3: they'll be able to lead their teams to be able
- 15:40Speaker 3: to use this effectively. So we often start there at
- 15:44Speaker 3: de Lasia teams. Sometimes we start a board, sometimes board
- 15:46Speaker 3: and leadership team, those kind of things, because what we're
- 15:49Speaker 3: finding is that the trailblazing leaders are spending up to
- 15:53Speaker 3: eight hours a week just continually working on AI and
- 15:56Speaker 3: working with their teams on AI to make sure that
- 15:58Speaker 3: it's going to be implemented, so it's upskilling themselves, but
- 16:01Speaker 3: also working with their teams to actually see how it's
- 16:03Speaker 3: being used at the coal face as well. So the
- 16:07Speaker 3: leaders using it in their own routines does a couple
- 16:10Speaker 3: of things. One is keeps them up to date, rather
- 16:13Speaker 3: than being the Ivory Tower leader who doesn't really know
- 16:15Speaker 3: about the AI but thinks it's a great idea. Instead,
- 16:18Speaker 3: they're actually embodying that and they're showing their teams and
- 16:22Speaker 3: everyone below them. Here's how I'm using it. You should
- 16:26Speaker 3: be using it too.
- 16:28Speaker 1: And I think neo because you're like you spent your
- 16:30Speaker 1: entire career in it and tech like you have seen
- 16:35Speaker 1: many organizations go through digital transformations, like ALI is really
- 16:40Speaker 1: different here, and I'm keen to know because like previously,
- 16:42Speaker 1: a digital transformation, it's like you just delegate that and.
- 16:46Speaker 3: CEO of a medium sized company if you've got a
- 16:48Speaker 3: new CRM, like a customer relationship management software, so like
- 16:52Speaker 3: Salesforce or something like that, where we've got our sales
- 16:54Speaker 3: and our deals and our products and all those kind
- 16:56Speaker 3: of things in there. Traditionally, it was get the it
- 16:59Speaker 3: to department to go and buy it the IT department
- 17:03Speaker 3: would then buy it, they'd implement it. They do a
- 17:05Speaker 3: little bit of training for everyone, and then the training
- 17:07Speaker 3: is done a couple of handholds with a couple of people,
- 17:10Speaker 3: a couple of manuals there, and then it's sorted. So
- 17:13Speaker 3: eighty percent of the work is about just putting the
- 17:15Speaker 3: software in. II is really different because most of the
- 17:20Speaker 3: work is actually with the training and embodying that within
- 17:23Speaker 3: your workday. But also this thing changes. This AI thing
- 17:28Speaker 3: is continually evolving, so it's not a one and done
- 17:30Speaker 3: training problem. It's actually an ongoing training thing. And so
- 17:35Speaker 3: if the leads aren't keeping that ongoing themselves, they've got
- 17:38Speaker 3: no hope of being able to lead that for the
- 17:39Speaker 3: rest of the organization.
- 17:41Speaker 1: Now on the topic of training, neo, what should we do?
- 17:44Speaker 1: And we're a comfort like you know, because we see
- 17:46Speaker 1: a lot of bad training. We're often bought into fixed
- 17:49Speaker 1: training that was not particularly effective.
- 17:52Speaker 2: But yeah, what tell it? Tell us about this?
- 17:53Speaker 3: A lot of the training, this part of that delegation thing.
- 17:56Speaker 3: A lot of the leaders are delegating to one department.
- 17:59Speaker 3: They're saying it this is yours to manage and roll out,
- 18:02Speaker 3: or they're saying HR it's a learning and development thing,
- 18:04Speaker 3: so Hi, you're going to roll this thing out. It's
- 18:08Speaker 3: not just one department to do. And what we find
- 18:10Speaker 3: is a lot of the training that happens and sometimes
- 18:12Speaker 3: is with the bigger providers as well. They're very much
- 18:15Speaker 3: based on features. So here, click this button, you can
- 18:18Speaker 3: do this, you can do it. Here's deep research, you
- 18:20Speaker 3: can click that and do some things. But they're not
- 18:22Speaker 3: showing people why they use it to how they can
- 18:25Speaker 3: put it in their day job to be actually able
- 18:27Speaker 3: to change the way they work. So being taught a
- 18:30Speaker 3: whole bunch of features is great, but not knowing how
- 18:33Speaker 3: I can improve my day job, improve my outputs, my clients,
- 18:36Speaker 3: my customers, my internal stakeholders. That's where people get stuck.
- 18:40Speaker 3: And we find a lot of teams get already trained
- 18:42Speaker 3: on the features, and then there's not the kind of
- 18:45Speaker 3: traction that people are seeing who are wanting to see.
- 18:47Speaker 3: They'll see people are using emails. That's really common. You
- 18:49Speaker 3: don't need to be trained to use emails, and so
- 18:52Speaker 3: these get lots of emails and lots of big emails,
- 18:54Speaker 3: but they're not actually getting better work done. And so
- 18:57Speaker 3: what we recommend is people train to tasks, not to features.
- 19:01Speaker 3: So initially there's going to be some kind of acclimation
- 19:04Speaker 3: here's what the buttons are and whatnot, But then quickly
- 19:06Speaker 3: it really shu you go into familiarity of the tool
- 19:09Speaker 3: and how that's going to help me a as a worker,
- 19:12Speaker 3: and so I really go to tasks that people do
- 19:14Speaker 3: and particularly if you can then get it to be
- 19:17Speaker 3: embedded with your workflows as well. So what we found
- 19:21Speaker 3: is there's actually a bit of a threshold with training
- 19:23Speaker 3: with different organizations, the researchers out there and saying that
- 19:28Speaker 3: there's a minimum viable training amounts. They call it the
- 19:31Speaker 3: minimum dose if you'd like, and it's actually five hours
- 19:34Speaker 3: of hands on training. So this is not read a manual.
- 19:37Speaker 3: It's not come to a lunch and learn and watch
- 19:39Speaker 3: someone do it. It's training where people are being taught
- 19:42Speaker 3: their features and how to use that, but also taught
- 19:44Speaker 3: how to put that within their day job and practicing
- 19:46Speaker 3: that minimum dose five hours. But those companies that are
- 19:50Speaker 3: actually getting benefits out of it, they're training up to
- 19:53Speaker 3: eighty hours a year. So that's not just here's what
- 19:57Speaker 3: the feature is. It's like, let's talk about your workflow.
- 20:00Speaker 3: Let's see how we can build AI into those workflows.
- 20:02Speaker 3: Let's practice these things and building an agent and then
- 20:05Speaker 3: sharing it and then tweaking it and changing it, and
- 20:07Speaker 3: let's use different scenarios here, those kind of things. Yeah,
- 20:11Speaker 3: they take some time to sit down with those those people.
- 20:14Speaker 3: You've got change champions, you might have external providers like us,
- 20:17Speaker 3: but that's where the benefit really comes in.
- 20:20Speaker 1: And if we look at that research that Nio just
- 20:22Speaker 1: mentioned that eighty hours a week training is like a
- 20:26Speaker 1: great dose to aim for in the research that showed
- 20:29Speaker 1: people got back fourteen hours per week terms every week
- 20:33Speaker 1: in terms of time saving.
- 20:35Speaker 2: So if you're like going, I am.
- 20:37Speaker 1: I'm going to like do some a train and do
- 20:39Speaker 1: like a one hour lunch and learn and like.
- 20:42Speaker 2: Hope for the best.
- 20:44Speaker 1: Just like think about that research and any research that
- 20:46Speaker 1: we mentioned, just like hit us up. We'll share links
- 20:49Speaker 1: with you. Okay, dooksys. So quite often.
- 20:54Speaker 2: Again, when I'm in meetings.
- 20:56Speaker 1: With various leaders is they'll be like, oh, yeah, you know,
- 20:58Speaker 1: we're just we're just you know, rolling out the licenses
- 21:01Speaker 1: and yeah, you know, give us a few months to
- 21:03Speaker 1: do that and then you know, we'll we'll get you
- 21:06Speaker 1: guys into do training and that that scares me. I mean,
- 21:11Speaker 1: we've got we've got plenty of works, so that's totally fine.
- 21:14Speaker 1: But what scares me is if you've got people with
- 21:17Speaker 1: licenses like free or paid. Obviously paid is better ACTU
- 21:20Speaker 1: you can do more stuff, and then there's a real
- 21:23Speaker 1: lag between when people have access to the tool and
- 21:26Speaker 1: when they're getting trained in the tool. To me, like
- 21:30Speaker 1: my organizational psychologist, habit change brain goes, oh man, that's
- 21:34Speaker 1: a lot of time to set some really bad habits
- 21:37Speaker 1: and also to produce a lot of AI slop that
- 21:41Speaker 1: is going to reduce productivity, not increase productivity. And interestingly,
- 21:46Speaker 1: some other research that Neil and I have come across
- 21:48Speaker 1: has found that people with AI I think there's something
- 21:51Speaker 1: like over one hundred percent increase in the amount of
- 21:54Speaker 1: emails that they're sending. And that is because it's now
- 21:58Speaker 1: super easy to write an e like there's going to
- 22:01Speaker 1: be a crap email if there's been no like human
- 22:04Speaker 1: judgment or involvement in that email other than write this
- 22:08Speaker 1: email with these three messages like that is slowing people down.
- 22:13Speaker 1: So make sure that do try to time the training
- 22:18Speaker 1: very close to when people have the licenses to reduce
- 22:22Speaker 1: bad habits forming and a proliferation proliferation. I think I
- 22:27Speaker 1: said that right of aislot. Okay, neo, what's this one about?
- 22:32Speaker 3: Yeah, it's on literacy first, so maybe there's a step
- 22:36Speaker 3: prior to that, but really it's about first focus on
- 22:39Speaker 3: here's what the talking do and he hey, can actually
- 22:41Speaker 3: get it to help you to challenge your thinking and
- 22:44Speaker 3: use it as a buddy or an expert rather than
- 22:46Speaker 3: just use your slave, which is a lot of people
- 22:48Speaker 3: like write me the email that kind of stuff. But yeah,
- 22:50Speaker 3: absolutely focus on that literacy first and get that good
- 22:54Speaker 3: baseline knowledge of how it can fit into mind day
- 22:56Speaker 3: and help me. Then the benefits really come into leverage,
- 23:00Speaker 3: which is one of my workflows. How can it help
- 23:03Speaker 3: fit into my workflow? How can it speed up that
- 23:06Speaker 3: thing that I do that annoys and the crap out
- 23:08Speaker 3: of me? How do I get AI to help me
- 23:09Speaker 3: with that or that big report that I do. Rather
- 23:13Speaker 3: than doing one a month because it's so big, maybe
- 23:16Speaker 3: I can do five a month because we're using AI
- 23:18Speaker 3: so we can get better insights on our customers or
- 23:20Speaker 3: the ones that I love is how do we customize
- 23:23Speaker 3: some of these things so rather than every customer, every client,
- 23:26Speaker 3: every department getting the same response. How do we get
- 23:29Speaker 3: AI to customize these things for people? So, but first
- 23:32Speaker 3: it really depends on that literacy part. So I'll put
- 23:35Speaker 3: it a little diagram here. So here's where a lot
- 23:38Speaker 3: of people start, just at access. So that's where here
- 23:42Speaker 3: take co pilot. There's a common one of companies get
- 23:45Speaker 3: and go for it and people have got it and
- 23:47Speaker 3: they think, yep, AI is now done. We've got our
- 23:51Speaker 3: it guy. He's pretty good at AI, I understand, and
- 23:55Speaker 3: our it guy is going to show people how to
- 23:57Speaker 3: do it while I have a luncheon len. That's great. Done,
- 24:00Speaker 3: that's not bad. It's not a bad start. But really
- 24:03Speaker 3: it's about then giving people the tools and showing them
- 24:06Speaker 3: with buttons to bress and things like that. You're not
- 24:08Speaker 3: going to get any business benefit there, but it's required, right.
- 24:11Speaker 3: You need to give people licenses, you need to give
- 24:13Speaker 3: them their people access, and you need to tell them
- 24:15Speaker 3: what these things are. We can certainly help with that,
- 24:17Speaker 3: but you can do that as well. But the benefits
- 24:19Speaker 3: really start coming in at this next stage, and that's
- 24:21Speaker 3: the literacy thing. Now you'll notice here this isn't a box.
- 24:26Speaker 3: This is a box with an arrow. The reason for
- 24:27Speaker 3: that is this literacy thing is now an ongoing challenge.
- 24:31Speaker 3: Like the AI you're using now is the worst AI
- 24:35Speaker 3: you're going to use for the rest of your life.
- 24:37Speaker 3: It is going to continue to change, not just with feature.
- 24:40Speaker 3: It's going to get smarter, it's going to be able
- 24:41Speaker 3: to be plugged into more systems and all those kind
- 24:43Speaker 3: of things. So the literacy part will have to change
- 24:47Speaker 3: over time. This is an ongoing thing that people need
- 24:50Speaker 3: to be involved in. But certainly get people up to
- 24:54Speaker 3: speed with what the tools are, how they work, how
- 24:56Speaker 3: they think, how they hallucinate from time to time, how
- 24:59Speaker 3: to best work with them so you get fewer hallucinations,
- 25:02Speaker 3: how you can fit it anyr work day, all of
- 25:04Speaker 3: those kind of things, and so that literacy is there
- 25:06Speaker 3: as a baseline for everything that comes next. From there,
- 25:10Speaker 3: we really see those kind of two things, and the
- 25:11Speaker 3: first is I'm calling an individual leverage, which is how
- 25:15Speaker 3: do I get it to help me in my day job?
- 25:19Speaker 3: Knowing what I do, and I've then got some great
- 25:21Speaker 3: literacy on how I can use AI to do that.
- 25:23Speaker 3: Then I can start building things like agents. I can
- 25:25Speaker 3: start building workflows and fitting how putting it into my
- 25:29Speaker 3: workday and getting some benefit out of that. And the
- 25:32Speaker 3: other key on this one is actually sharing it with
- 25:33Speaker 3: your team as well, so making sure that individuals know
- 25:37Speaker 3: how to get the best out of this tool and
- 25:39Speaker 3: know how to look after these things that they're built
- 25:41Speaker 3: as well, so we don't just get old agents that
- 25:43Speaker 3: kind of work, that kind of don't work, and processes
- 25:46Speaker 3: and things like that half broken. So that individual leverage
- 25:49Speaker 3: is the first next step. And then after that, what
- 25:51Speaker 3: we generally find and what we recommend and do help
- 25:55Speaker 3: companies with is that organizational leverage, which is we have
- 25:58Speaker 3: this workflow, how do we get AI to help that?
- 26:02Speaker 3: And the organization and individual leverage often starts with I'm
- 26:06Speaker 3: doing this thing, how do I get AI to help
- 26:08Speaker 3: me do this thing? In the same way just a
- 26:11Speaker 3: bit different. The companies that are really going ahead, they're
- 26:15Speaker 3: the ones who are saying, now I've got AI, how
- 26:17Speaker 3: do I re envisage this workflow? How do I really
- 26:19Speaker 3: re envisage the inputs the outputs that we're getting from
- 26:22Speaker 3: these How do I actually move the company's dial rather
- 26:25Speaker 3: than just shove AI into a step? How do I say,
- 26:28Speaker 3: now I've got this an amazing tool. How do I change
- 26:32Speaker 3: the way we're delivering so it's more effective, more productive,
- 26:34Speaker 3: those kind of things, and that's where the big benefits
- 26:37Speaker 3: come from.
- 26:39Speaker 2: And just on that. It's not just about.
- 26:41Speaker 1: Finding people who are really really good at AI or
- 26:44Speaker 1: really heavy users. It's about training people or finding people
- 26:49Speaker 1: that think in terms of workflow, like like what is
- 26:54Speaker 1: the process? How do I unpack a process step by step?
- 26:57Speaker 2: How do I know what AI can do and what
- 27:00Speaker 2: it can't do?
- 27:01Speaker 1: Because what Neo and I find and the rest of
- 27:03Speaker 1: the team when we're working with organizations on this, there's
- 27:06Speaker 1: a lot of people that think AI is magic, like
- 27:10Speaker 1: it can create high quality data where no data exists,
- 27:14Speaker 1: it's not that good. So it's really important that like
- 27:17Speaker 1: when we're looking at individual leverage for individual workflows and
- 27:20Speaker 1: organization or team or function workflows, is that these people
- 27:24Speaker 1: not only no AI, but they also think like a
- 27:27Speaker 1: workflow designer to get the best stuff. And like, obviously
- 27:31Speaker 1: you know whereby us we're do great training on this,
- 27:33Speaker 1: but you just need to make sure you've got that
- 27:35Speaker 1: skill set in mind.
- 27:35Speaker 2: It's not just about knowing AI, and.
- 27:37Speaker 3: Often we find that it's in different people. So you
- 27:39Speaker 3: might have the IT person of helping out, and you
- 27:42Speaker 3: might have someone who's like a business analyst kind of
- 27:44Speaker 3: processing kind of person helping out. And you might also
- 27:46Speaker 3: have someone who's like that into the SME expert who
- 27:50Speaker 3: knows their organizational lens and the three of those people
- 27:54Speaker 3: are often needed to pull in the needs and the
- 27:57Speaker 3: goods all those kind of things. But yeah, it's not
- 28:00Speaker 3: they get one person to do it or a self serve,
- 28:03Speaker 3: particularly without training, because then you're going to get a
- 28:05Speaker 3: lot of broken processes and upset customers and things like that.
- 28:08Speaker 3: So yeah, absolutely you need to be a whole organization
- 28:12Speaker 3: behind this. And you'll also see this diagram and really
- 28:15Speaker 3: provi should go on forever because this is now an
- 28:17Speaker 3: ongoing challenge, so it's not once again. It's not a
- 28:20Speaker 3: one and done. You might have improved this process, but
- 28:23Speaker 3: the tools will change over time and so we need
- 28:25Speaker 3: to continue revise and look at these to see if
- 28:28Speaker 3: we need to optimize them again.
- 28:29Speaker 1: Okay, one more solution before we go into questions.
- 28:33Speaker 2: So AI will give people their time.
- 28:36Speaker 1: Back, but guess what happens if you don't have a
- 28:40Speaker 1: plan for where that time goes.
- 28:42Speaker 2: They're going to fill it with more work.
- 28:45Speaker 1: And that's probably going to lead to brain fry because
- 28:47Speaker 1: generally we're filling it with like deeper, more cognitively intense work,
- 28:52Speaker 1: particularly when we're reviewing AI's output, which is now a
- 28:55Speaker 1: large part of our day that requires vigilance and critical
- 28:59Speaker 1: thinking and stuff that requires more brain power than the
- 29:03Speaker 1: task that we're outsourcing to AI, which typically require a
- 29:05Speaker 1: little bit less brain power. So this is really important.
- 29:09Speaker 1: So leaders need to really give some thought to this
- 29:12Speaker 1: and communicate it, like, do you want people with all
- 29:16Speaker 1: these time savings that you're getting or are going to
- 29:19Speaker 1: get or have been promised? Like, do you want them
- 29:22Speaker 1: putting that towards innovation where previously they had no time?
- 29:25Speaker 1: Do you want them putting it towards the customer and
- 29:28Speaker 1: learning more about the customer, speaking more with customers, trainer
- 29:31Speaker 1: up your customer satisfaction scores and net promotor scores? Do
- 29:34Speaker 1: you want them putting it back into learning and development?
- 29:37Speaker 1: Because I'm yet to meet an L and D professional
- 29:39Speaker 1: who's like, oh, people have so much time for learning
- 29:41Speaker 1: at our organization. Or do you want them putting it
- 29:44Speaker 1: into god forbid, work life balance? So crazy idea, So
- 29:47Speaker 1: be really thoughtful around that, because if you are not
- 29:51Speaker 1: people will just cram it with more work. Thanks so
- 29:55Speaker 1: much for listening. If you are in the middle of
- 29:56Speaker 1: an AI rollout and want some help getting it right,
- 30:00Speaker 1: you can find us at Inventium dot ai.
- 30:03Speaker 2: We would love to chat. See you next time.
- 30:06Speaker 1: How i AI was hosted by me, Amantha Imber and
- 30:09Speaker 1: Neo Applan. A big thank you to Martin Imber who
- 30:12Speaker 1: does our sound editing, and Jem Rubio for production support,
- 30:16Speaker 1: and thank you to John Kilby who composed the theme music.