Latest / The DEX Show: A Show for IT Change Makers / From Shadow AI to Strategy: The Six-Month AI Imperative (w/ Charlene Li)

Transcript

From Shadow AI to Strategy: The Six-Month AI Imperative (w/ Charlene Li) · The DEX Show: A Show for IT Change Makers

0:01
Speaker 1: Welcome it, change Makers to the Deck Show with Tim
0:05
Speaker 1: Flower and Tom McGraw. Let's get into it.
0:09
Speaker 2: Hello, change Makers, Welcome back to the Deck Show. I'm
0:12
Speaker 2: flying solo today. It's Tom McGraw and I've got I've
0:16
Speaker 2: got an interesting mission to fly everybody, because we've had
0:20
Speaker 2: repeat guests on the Deck Show before and we're delighted
0:25
Speaker 2: today to welcome back for the second time Charlie Lee.
0:28
Speaker 2: She's a business author, speaker, analyst, and thought leader known
0:32
Speaker 2: for her work on digital transformation, the future of work
0:36
Speaker 2: and what we want to talk to her about today
0:38
Speaker 2: AI strategy. Welcome back to the show, Charlie. How are
0:41
Speaker 2: you good.
0:43
Speaker 3: I'm happy to be here.
0:45
Speaker 2: And what I was going to say I was going
0:46
Speaker 2: to qualify it, Charlie, is that while we have had
0:48
Speaker 2: repeat guests, not so many, you know, it always means
0:51
Speaker 2: a great deal when we want somebody back on. We've
0:55
Speaker 2: never done it quite like this. So the first time
0:56
Speaker 2: you're on the show, Tim was flying solo and Tim
0:59
Speaker 2: did they interview? Now it's me. So although you're the
1:02
Speaker 2: second time guest, I've never actually met you before, so
1:05
Speaker 2: it's it's almost like a completely new guest for me.
1:08
Speaker 4: At least well this is this is gonta be interesting then,
1:11
Speaker 4: but you've used you know, the other episodes, so you
1:13
Speaker 4: know what I said, I'm.
1:14
Speaker 2: Not going to ask you exactly the same questions now,
1:16
Speaker 2: I'm not. I'm not. We we actually know you're part
1:18
Speaker 2: of a sort of special series of shows where we're
1:21
Speaker 2: really delving into AI transformation. And particularly because you know,
1:26
Speaker 2: sponsor of the show, Next Think is just released its
1:30
Speaker 2: AI Drive, which gives it teams a kind of unprecedented
1:35
Speaker 2: real visibility into AI usage of their organization. So so
1:39
Speaker 2: many things to talk about in and around that. But
1:42
Speaker 2: first of all, so it's been about a year since
1:44
Speaker 2: you since you came on a show with Tim and
1:46
Speaker 2: the people who were able to listen. How have you been.
1:51
Speaker 2: What have you been up to lately? Has the last year.
1:53
Speaker 4: Been It's been good. We've been my co arth Bearcatcher
1:57
Speaker 4: Whilsh and I have been putting the finishing touches on
2:00
Speaker 4: our book, a long awaited book called Winning with Ai.
2:04
Speaker 4: And I've been working with a lot of people with
2:06
Speaker 4: speaking workshops and advisory to just hear what they're doing
2:11
Speaker 4: and helping them create an AI roadmap that creates value.
2:14
Speaker 4: So it's been an interesting year. It's lots of changes, happy,
2:17
Speaker 4: lots of.
2:18
Speaker 2: Changes, right. I mean, it's fair to say, it's fair
2:22
Speaker 2: to say that AI is changing all of our working lives,
2:25
Speaker 2: in our lives, and I was interested to ask how
2:28
Speaker 2: has your own regular use of AI evolved over the
2:31
Speaker 2: last twelve months.
2:33
Speaker 4: One of the things I've done is created a lot
2:35
Speaker 4: of custom GPTs. They're sometimes in GPT their projects and
2:40
Speaker 4: Claude their gems and Google. They all have the pros
2:43
Speaker 4: and cons and I have everything from obviously, one that's
2:47
Speaker 4: helping me write the book has all my interviews and
2:50
Speaker 4: past writings, so it kind of knows all of my content.
2:53
Speaker 4: It's sort of my second brain. But I also have
2:55
Speaker 4: one for adventure travel, so I've given it my past
2:58
Speaker 4: itinerary sounderstands what I like. I have one that helps
3:02
Speaker 4: me cook, another one that helps me with gardening, nutrition, bookkeeping,
3:07
Speaker 4: because I can't stand bookkeeping, like how do I do
3:11
Speaker 4: this versus that? And I think it's just very helpful
3:15
Speaker 4: to have sort of these customized versions of whatever it
3:19
Speaker 4: is that I do on a routine basis. I have
3:21
Speaker 4: also been doing some bibe coding along the way, just
3:23
Speaker 4: to experiment my fingers in it. I don't know any
3:27
Speaker 4: of the technology. It is fantastic, but it definitely helps
3:32
Speaker 4: if you know some coating and knows like what does
3:34
Speaker 4: it mean to commit to a GitHub? I still don't
3:38
Speaker 4: quite understand what that is, but I know it's something
3:41
Speaker 4: that you have to do, So it's one of those
3:44
Speaker 4: things that I'm experimenting with.
3:46
Speaker 2: Very cool selection of all of those things, what would
3:48
Speaker 2: you find it hardest to live without? Do you think?
3:51
Speaker 4: Oh, definitely all of my customs gputs because I rely
3:56
Speaker 4: on them, my bookkeeping one in particular, because when I
4:00
Speaker 4: have an issue going and using the traditional Google searches,
4:04
Speaker 4: this doesn't work because it doesn't understand me. And sometimes
4:08
Speaker 4: I'm asking the same question again and or a different
4:11
Speaker 4: variation on it, and it has that memory so it
4:13
Speaker 4: understands my situation and my circumstances. So that has been
4:17
Speaker 4: a lifesaver.
4:19
Speaker 2: Last question of this front, Charlie, are there any are
4:23
Speaker 2: there any ways in which you're using these different and
4:25
Speaker 2: these different custom what you call from.
4:29
Speaker 3: Custom one custom GPTs custom GPTs.
4:32
Speaker 2: Are there any ways that you're using these custom GPTs
4:36
Speaker 2: in such a way as to mitigate some kind of
4:39
Speaker 2: side effect they might have on your on your creativity
4:42
Speaker 2: or your imagination, because these are all quite imaginative and
4:44
Speaker 2: creative well many of them, not the bookkeeping, but many
4:47
Speaker 2: of them are quite imaginative and creative areas to use
4:50
Speaker 2: them in. And are are there ways you need to
4:52
Speaker 2: moderate or navigate it which which bring the best out
4:55
Speaker 2: of you and make sure the technology doesn't kind of
4:57
Speaker 2: overwhelm what you do. That's your approach to that.
5:00
Speaker 4: Well again, people who ask me all the times like
5:02
Speaker 4: does it help you or hurt you, especially for things
5:05
Speaker 4: like my creative writing, And one of the things I
5:09
Speaker 4: do is I don't use it to just write. I
5:11
Speaker 4: use it to help me brainstorm for it actually adds
5:15
Speaker 4: to my creativity because it's kind of like having another
5:18
Speaker 4: version of myself to brainstorm with. And there aren't very
5:21
Speaker 4: many people who understand the context of things I have
5:24
Speaker 4: to explain to them. They don't quite underget it. This
5:27
Speaker 4: thing just knows everything about me, and frankly, it remembers
5:31
Speaker 4: more of the stuff that I've written than I have
5:33
Speaker 4: in the past, So it'll bring up things that I've
5:37
Speaker 4: said or done and talked about make connections with things
5:41
Speaker 4: that I may not have remembered. So I do think
5:44
Speaker 4: of it as a second brain that I can careery.
5:48
Speaker 4: And the way I use any of these checkbots is
5:53
Speaker 4: when I put in a query at the end, I
5:55
Speaker 4: ask before you start asking me any clarifying questions, because
6:00
Speaker 4: it turns it from an answer machine into a conversation.
6:03
Speaker 4: And that creativity and that that expansiveness comes from the
6:07
Speaker 4: conversation that I have with it.
6:09
Speaker 2: Oh that's a cool prompt. I've never heard like that
6:12
Speaker 2: a lot. It is amazing, right, like, uh, these these
6:16
Speaker 2: these are applications. They know us as well as anybody
6:20
Speaker 2: could ever know us, right, right, you know, it's astonished
6:23
Speaker 2: thing I mean, And do you have any do you
6:25
Speaker 2: have any misgivings about that level of transparency and intimacy?
6:30
Speaker 3: I do. I'll give you a little exercise.
6:33
Speaker 4: Someone sent me a little thing that said ask especially
6:37
Speaker 4: use the one that you use the most. So I
6:38
Speaker 4: tend to use chatchy Bet because it's sort of all
6:41
Speaker 4: around decent.
6:43
Speaker 3: Not the best at everything, but it's all around decent.
6:46
Speaker 3: So it's my sort of default.
6:48
Speaker 4: And I asked it tell me everything you know about me,
6:51
Speaker 4: which is an interesting exercise anyways. And then the second
6:55
Speaker 4: prompt is go deeper. And the third prompt is do
7:00
Speaker 4: a full cycle analysis of me. And then the fourth
7:04
Speaker 4: prompt is write a letter.
7:06
Speaker 3: To my soul.
7:08
Speaker 2: Oh my yarness.
7:09
Speaker 3: What I was so curious, like, what's it going to say?
7:15
Speaker 4: And it was incredibly moving. I mean, these things are
7:20
Speaker 4: built to do that. I mean, if it's going to
7:22
Speaker 4: write to your soul, it's going to speak. It understands
7:25
Speaker 4: what moves you. And again it helps if you do
7:28
Speaker 4: it across multiple domains of your life. And I'm not
7:32
Speaker 4: to say that it has it knows everything about me,
7:35
Speaker 4: but it's kind of like when you go see a
7:37
Speaker 4: fortune teller or you get your palm red. It's kind
7:40
Speaker 4: of an interesting exercise. You don't take it too seriously.
7:43
Speaker 4: You take it with a grain of salt. But it
7:46
Speaker 4: was incredibly moving. And I sent this to some friends
7:49
Speaker 4: and they go, this, This brought me to tears.
7:53
Speaker 3: So I.
7:55
Speaker 4: Look at it again with a big grain of sult.
7:58
Speaker 4: It doesn't know anything. It doesn't true really know me, right,
8:01
Speaker 4: it has a lot of information for me about me.
8:04
Speaker 4: It could extraperate a lot of understanding and pull together
8:08
Speaker 4: things in a very interesting way. And some parts of
8:13
Speaker 4: it truly resonated with me. I'm like, how did it
8:16
Speaker 4: figure that out? So somehow we did it just again
8:20
Speaker 4: with those prompts. It was set up in the right way.
8:24
Speaker 2: I keep saying, one more question in this direction, you know,
8:26
Speaker 2: or more, just one more something? How how how far
8:31
Speaker 2: are we from the first AI centric religion? I know
8:38
Speaker 2: that that sounds like a left field question, but I
8:40
Speaker 2: mean you you know, people, what did you just cite
8:43
Speaker 2: that you cited? People go into mystics, People go into
8:47
Speaker 2: people with you know, like effectively little religious pilgrimages, right like,
8:52
Speaker 2: you know, and and then you're getting a resonance and
8:54
Speaker 2: a sense of comprehension which is both self comprehension, which
8:58
Speaker 2: is mysterious moving meaningful. You know, like, how are people
9:04
Speaker 2: going to react to that? At scale? Some people are
9:07
Speaker 2: going to go very very far with that connection, right.
9:10
Speaker 4: Well, I want to draw a difference between religion, which
9:12
Speaker 4: is an organized uh or a place where beliefs and
9:17
Speaker 4: rituals and traditions, sacraments, the that's of religion. There's a
9:23
Speaker 4: practice to it, there's a ritual to it, and spirituality,
9:27
Speaker 4: which is diving deeper into this whole mystical world. And
9:32
Speaker 4: I don't know if you want to call it spirituality,
9:35
Speaker 4: but I do believe AS want to help us understand
9:37
Speaker 4: ourselves better.
9:39
Speaker 3: In the world of.
9:41
Speaker 4: Knowledge where AI can know everything, the area where it
9:45
Speaker 4: doesn't really know is us, like what's inside of this head?
9:50
Speaker 3: How do people think?
9:52
Speaker 4: And if it can help us understand each other better
9:55
Speaker 4: and help us smooth the communications that we have. Just
10:00
Speaker 4: look at within organizations it communicating with HR two groups
10:05
Speaker 4: that really need to work together very closely when it
10:08
Speaker 4: comes to transformation. They speak completely different languages, or a
10:13
Speaker 4: startup and a VC speak to completely different languages with
10:17
Speaker 4: common goals, but they have to do some sort of
10:19
Speaker 4: subtle translation.
10:21
Speaker 3: Or let's just take even our relationships.
10:24
Speaker 4: I have said so many times in my relationships, with
10:26
Speaker 4: friendships and with partners, I say something I asked them
10:29
Speaker 4: repeated and it comes back as something completely different. I'm like,
10:32
Speaker 4: what am I doing that they can't understand me? So
10:37
Speaker 4: I think that AI can help us understand ourselves and
10:39
Speaker 4: each other better, and if that is an opportunity, I
10:43
Speaker 4: think to be able to use these tools to help
10:46
Speaker 4: us understand what is it that I don't understand about myself,
10:50
Speaker 4: so that I can be better at whatever it is
10:53
Speaker 4: that I want to accomplish well, I think.
10:55
Speaker 2: This has been an amazing way to label foundations of
10:58
Speaker 2: the rest of the conversation because I've we've got to
11:00
Speaker 2: the heart of some of the more extreme and interesting
11:04
Speaker 2: use cases and potentialities of this technology, and now we
11:07
Speaker 2: have to imagine what that means for our workplaces and
11:10
Speaker 2: for our professional lives, and for our careers. But specifically
11:14
Speaker 2: the first of us, specifically for the workplace, and something
11:16
Speaker 2: you've written about extensively is the concept of AI leadership,
11:21
Speaker 2: and you touched upon it with Tim and your last appearance,
11:23
Speaker 2: But maybe just recap to begin with what the core
11:26
Speaker 2: principles in your view of good AI leadership looks like
11:30
Speaker 2: in this era.
11:31
Speaker 4: Well, I think good AI leadership starts with just good leadership.
11:35
Speaker 4: And one of the things that's interesting about leadership is
11:39
Speaker 4: that it becomes ultimately incredibly important when things are uncertain,
11:45
Speaker 4: when they're unpredictable, because if things are not changing, it's
11:50
Speaker 4: sort of just maintaining their status quo. That's where managers
11:53
Speaker 4: are really good at. But when change is happening and
11:56
Speaker 4: changes required, that's when you need leadership. And leaders what
12:01
Speaker 4: they do really well is they create structure, They create
12:06
Speaker 4: a purpose and a why, an underlying why do we
12:10
Speaker 4: have to go do this change? We don't want to
12:13
Speaker 4: it's hard. And they create the strategy, the roadmap, and
12:18
Speaker 4: they also create accountability, and they're the ones who establish
12:24
Speaker 4: credibility and a relationship so that people will trust them.
12:28
Speaker 4: And I have always believed that with leadership. Classic leadership
12:32
Speaker 4: has always been a relationship between people who aspire to
12:36
Speaker 4: create change and the people who are inspired to follow them,
12:41
Speaker 4: and that has been the truth for a millennia.
12:45
Speaker 3: So nothing changes around that now.
12:49
Speaker 4: Yeah, And the thing when it comes to AI leadership
12:51
Speaker 4: is that that has to come forward even more and
12:53
Speaker 4: people forget that. And this is a moment where I
12:59
Speaker 4: think people are not standing in leadership. They see that,
13:03
Speaker 4: they know that AI is important, they may or may
13:06
Speaker 4: not be doing it. They're not AI fluet for the
13:09
Speaker 4: most part, and without that fluency, it's very difficult to
13:14
Speaker 4: lead a change. They're not showing up as leaders. They're
13:19
Speaker 4: not saying this is the direction we need to go in.
13:23
Speaker 4: That's all go in that direction, and how can I
13:27
Speaker 4: support you through this change? So again I think great
13:33
Speaker 4: AI leadersharps. Leadership starts with a mandate that says this
13:38
Speaker 4: is why we have to change. We see some CEOs
13:41
Speaker 4: now from places like Shopify, from Amazon. It's mostly technology
13:46
Speaker 4: companies because they're kind of more on it. But even
13:49
Speaker 4: like the CEO of Walmart has said, this is the
13:51
Speaker 4: future of what we're going to do, and what you
13:54
Speaker 4: see Walmart doing is rolling it out to all of
13:56
Speaker 4: their associates and stores and giving them access to very
14:00
Speaker 4: powerful AI tools in the existing applications that they use.
14:05
Speaker 4: So if this is something that again there's leadership happening,
14:09
Speaker 4: but also it's not just enough to talk about it,
14:12
Speaker 4: you actually have to follow up with the actions.
14:15
Speaker 2: So interesting and just on that, I just sat an
14:18
Speaker 2: interesting post just yesterday is who should be leading AI
14:22
Speaker 2: transformation in an organization? Is it the CIO, the CEO,
14:25
Speaker 2: even throughout the CMO, which I you know even you
14:28
Speaker 2: are its cause not completely preposterous, right, I don't imagine
14:33
Speaker 2: you're going to name one, But what are your thoughts
14:35
Speaker 2: on that question of who in the existing structures should
14:39
Speaker 2: be could be taking the leader?
14:41
Speaker 4: It's the person who can understand the technology and how
14:46
Speaker 4: it works, and they don't have to be a technologist
14:48
Speaker 4: and the person who understands your business strategy because in
14:52
Speaker 4: the end AI has to be in service of your strategy.
14:56
Speaker 4: It is not about technology, it is about how it
14:59
Speaker 4: can be use as a strategic initiative. So at Moderna,
15:04
Speaker 4: the pharmaceutical company that does mRNA co vaccines, for example,
15:10
Speaker 4: AI lives with the CHRO HR. In fact, they took
15:15
Speaker 4: their technical team and put it under the HR team
15:20
Speaker 4: reporting to the CHRO because they realized that this transformation
15:25
Speaker 4: was going to be human led. It is about transforming
15:29
Speaker 4: that people who are going to use the technology, so
15:32
Speaker 4: that people come first and the technology comes second. The
15:36
Speaker 4: technology is in service of the people who will execute
15:39
Speaker 4: a strategy. So that is that's a big outlier. I
15:43
Speaker 4: would say that is not the norm. A lot of
15:45
Speaker 4: people by default give it to the IT and I
15:48
Speaker 4: would say, really think about this and say, is your
15:52
Speaker 4: IT leader, the person who's going to be driving this,
15:55
Speaker 4: do they understand your business strategy?
15:57
Speaker 3: Are they focused.
15:59
Speaker 4: On ex cuting and supporting business or are they focused
16:02
Speaker 4: on primarily technology and procurement And that's not going to
16:08
Speaker 4: work A I would tie a very slow and painful
16:11
Speaker 4: death if you give it to a person who just
16:14
Speaker 4: doesn't connect the technology to the business strategy.
16:18
Speaker 2: I think that let's say that as an incentive for
16:23
Speaker 2: our listeners who are in it, and I T leadership
16:25
Speaker 2: who are being given these responsibilities to take responsibility for
16:29
Speaker 2: that dimension, right like, because a lot of it, you know,
16:33
Speaker 2: and and and so so you know that becomes that
16:37
Speaker 2: becomes what you just said becomes an incentive for it.
16:41
Speaker 2: If it's going to be their responsibility in an organization, A,
16:44
Speaker 2: it's a huge responsibility, and B it might require a
16:48
Speaker 2: reorientation of their perspective. Would you say, right, how would
16:51
Speaker 2: you how would you turn that advice around to it?
16:54
Speaker 2: Leaders who are being asked to affect this change.
16:58
Speaker 4: They should be constantly asking, there's one question, what is
17:02
Speaker 4: our business strategy and how can AI support it? What
17:07
Speaker 4: are the biggest challenges we have and our business strategy
17:11
Speaker 4: the biggest problems, the biggest opportunities, and how can AI
17:15
Speaker 4: address them. We typically have three, let's say, typical business
17:20
Speaker 4: objectives that we have. It's hard to keep check more
17:23
Speaker 4: than three, and any use case. And I'd like to
17:27
Speaker 4: say use cases are not a strategy. So when you
17:30
Speaker 4: have a whole long list of AI use cases, that's
17:35
Speaker 4: not a strategy. Strategy is this is? These are our
17:39
Speaker 4: business objectives, These are our top three. This is how
17:42
Speaker 4: AI is going to support that, and you can talk
17:44
Speaker 4: about that until the cows come home. So everybody goes, so,
17:49
Speaker 4: what's our AI strategy going to be while we're using
17:52
Speaker 4: AR to support these three business strategies which you all
17:54
Speaker 4: know and are supportive of, we all agree on that is,
17:58
Speaker 4: you don't have to rethink that.
18:00
Speaker 3: And this is why we're doing it now. This is
18:04
Speaker 3: what we're doing.
18:05
Speaker 4: Here's a rogue map of what we're going to do,
18:07
Speaker 4: and this is how we're going to actually implement it.
18:10
Speaker 4: These are the supports, these are the the training we
18:13
Speaker 4: should definitely come and talk about in a second. And
18:19
Speaker 4: this is the change we're going to have. And it's
18:23
Speaker 4: not necessarily going to be impossible, but it is going
18:26
Speaker 4: to be changed, and we need to be prepared for this. Again,
18:31
Speaker 4: that's leadership, it's strategy. It's about having a concerted focus,
18:37
Speaker 4: and focus says you have to choose what you will
18:39
Speaker 4: do and what you won't do. And right now there's
18:44
Speaker 4: not a lot of choices being made. It's just a
18:46
Speaker 4: lot of people trying a lot of different things.
18:50
Speaker 3: I like what your sponsor does.
18:51
Speaker 4: It's looking at how AI is actually being used in
18:54
Speaker 4: the organization because that can give you information about where
18:58
Speaker 4: the value is being found. And if you're pilots, if
19:02
Speaker 4: you're an endless pilot purgatory, that's not a good sign.
19:05
Speaker 4: It's an indicator that you do not have a strategy
19:08
Speaker 4: of how to use AI. And so I say, stop
19:12
Speaker 4: doing the pilots, really focus on doing the top strategic
19:17
Speaker 4: AI initiatives that are going to support your strategy and
19:20
Speaker 4: put all your eggs in those baskets, because you're not
19:24
Speaker 4: going to do visibility studies. You're not going to do pilots.
19:27
Speaker 4: You're going to say this is really important to us,
19:30
Speaker 4: We're going to invest, we're going to figure out the technology.
19:34
Speaker 4: We're going to connect this to outcomes for our customers
19:37
Speaker 4: and impact every single quarter. And that's what a coherent
19:41
Speaker 4: strategy in roadmap looks like when it comes to AI.
19:44
Speaker 2: And I mean, if it leaders can get that kind
19:47
Speaker 2: of level of visibility into how AI and IT are
19:51
Speaker 2: actually being used across for organizations. What do you think
19:55
Speaker 2: would surprise the board or management of the C suite
19:59
Speaker 2: most about what it? What it would tell.
20:00
Speaker 4: Them, Oh, that there's so much happening. We call it
20:05
Speaker 4: shadow AI sort of like shadow it anybody with a
20:09
Speaker 4: browser can use it. They don't need the permission of it.
20:14
Speaker 4: They don't need anyone's permission to go and start using
20:17
Speaker 4: these things. And I think they would be surprised by
20:20
Speaker 4: how much value is actually being created. People send to
20:23
Speaker 4: me like, oh, I see a little of efficiency here
20:25
Speaker 4: and there, but is there true value being created? Go
20:29
Speaker 4: and look at how people actually are using this. They
20:32
Speaker 4: wouldn't be using it, but wasn't creating value. So its
20:38
Speaker 4: value is absolutely being created. You don't know exactly how.
20:41
Speaker 4: How I can tell is when they're given access to
20:45
Speaker 4: these tools. If you try to take it away from them,
20:48
Speaker 4: they go with my cold dead fingers on it when
20:51
Speaker 4: you take it away, So like, no, I need this
20:55
Speaker 4: to do my job. There's value there, but I don't
21:00
Speaker 4: know if you can measure it in specific ROI, which
21:04
Speaker 4: is why I think ROI is the worst measure you
21:06
Speaker 4: could try to be looking at. It's the one question
21:09
Speaker 4: what's the roy of AI? It's kind of like saying,
21:12
Speaker 4: so what can AI do? It could do so many things,
21:16
Speaker 4: but is it creating value that matters to you? So
21:20
Speaker 4: ask instead of what can AI do? Ask what can
21:24
Speaker 4: AI do for us to help us? Achieve our biggest
21:29
Speaker 4: strategic objectives and what.
21:31
Speaker 2: Kind of more meaningful metrics or user patterns should we
21:36
Speaker 2: be looking for in place if something as reductive as ROI.
21:40
Speaker 4: Well, again, you already have business metrics, you have objectives
21:46
Speaker 4: and metrics associated with that that tells you whether you're
21:49
Speaker 4: moving towards your objectives better, faster, cheaper. And I would
21:55
Speaker 4: add another metric, which is safer. And so we talked
22:00
Speaker 4: about safety, security, privacy all the time. When it comes
22:03
Speaker 4: to AI, it could potentially help you do those because
22:07
Speaker 4: it can highlight the things that may be going wrong. Again,
22:11
Speaker 4: it's a security list too, but it can also tell you, hey,
22:14
Speaker 4: this thing may not be aligned because it can give
22:17
Speaker 4: it very clear objectives. This is our objectives, this is
22:20
Speaker 4: our strategic objective. Highlight anything that doesn't help us get
22:23
Speaker 4: to that point. Highlight anything that could be standing in
22:27
Speaker 4: the way or could be jeopardizing it. So again, AI
22:31
Speaker 4: is very objective driven, and if you tell it this
22:35
Speaker 4: is our objective, this is our stage real objective.
22:37
Speaker 3: How do we get there? How do we use AI
22:41
Speaker 3: to help us get there?
22:42
Speaker 4: So I kind of think it's meta that use AI
22:45
Speaker 4: to figure out how to use AI.
22:47
Speaker 3: So may as well use the tools available to like.
22:50
Speaker 2: Macro prompting in a way. Yeah, exactly. Yeah, what about
22:56
Speaker 2: the question of like productivity metric and specifically and maybe
23:02
Speaker 2: a difference between the qualitative and the quantitative, right, which
23:05
Speaker 2: is something that I feel is you know, when you
23:08
Speaker 2: talk about that employee lead AI transformation, you know that
23:13
Speaker 2: for me can be reduced to that contrast between qualitative
23:17
Speaker 2: and quantitative. You know, we all know that AI could
23:21
Speaker 2: reel off a library worth of pretty bad books for
23:25
Speaker 2: us and with you know, a series of prompts. But
23:28
Speaker 2: that's not what we're after, is it. We're after work
23:32
Speaker 2: which cuts through, which retains personality and originality. And that's
23:36
Speaker 2: is that an area where you feel the employee has
23:40
Speaker 2: to be in the driver's seat, most most importantly, most significantly.
23:44
Speaker 4: Well, there's a disconnect because employees know what the problems
23:48
Speaker 4: are with their job, like these are the bottlenecks, these
23:52
Speaker 4: are the things that are pain and to do do
23:55
Speaker 4: and but there's a disconnect between them knowing what the
23:58
Speaker 4: problem is and how use AI to solve it. So again,
24:03
Speaker 4: I think employees are incredibly smart. People generally are smart.
24:07
Speaker 4: They will not do things that are not productive for them. Okay,
24:10
Speaker 4: this is this does not include procrastination, so that's something
24:13
Speaker 4: completely different. But this is again about us understanding that
24:19
Speaker 4: if something is going to help us, if you're given
24:21
Speaker 4: a tool to help us do our jobs better, we're
24:24
Speaker 4: going to use it. And so I can tell you like, oh,
24:29
Speaker 4: I can do this a lot better, the results are better.
24:33
Speaker 4: I don't have to I can be more creative and
24:36
Speaker 4: come up with better solutions. But you give them the tool,
24:40
Speaker 4: and they also have to give them the training, and
24:42
Speaker 4: the training comes in two flavors.
24:45
Speaker 3: They're sort of top.
24:46
Speaker 4: Down, centralized, very organized. It's great for understanding what AI does,
24:53
Speaker 4: and it's great for responsible and ethical use of AI.
24:57
Speaker 4: But when it comes to that additional level of fluid
25:00
Speaker 4: and where you can apply AI to your job until
25:03
Speaker 4: your work, that requires much more of a decentralized approach.
25:08
Speaker 4: It's about pure learning and support. It's about cross fertilization
25:13
Speaker 4: across departments. Like we learned this, this could apply to
25:15
Speaker 4: you too as well. And then the fourth area of
25:18
Speaker 4: fluency is can you use and talk about how to
25:23
Speaker 4: share this? Again, that sharing, the teaching of it, the
25:27
Speaker 4: collaboration that happens is that fluency there to be able
25:32
Speaker 4: to talk about and share and learn collectively too as well.
25:36
Speaker 4: So those are types of individual fluency, but also organizational
25:40
Speaker 4: fluency with AI that's really important to develop. And I'll
25:44
Speaker 4: put a line in the sand here. I believe that
25:47
Speaker 4: organizations should be focused on getting their AI, getting their
25:51
Speaker 4: most important workforce, especially the knowledge based workforce. Everyone fluent
25:57
Speaker 4: with AI in the next six months. There's no excuse.
26:03
Speaker 4: We know this is the future. So invest the money,
26:07
Speaker 4: the time, and it's both to get people to be fluent,
26:13
Speaker 4: get the technology in place where everybody can get access
26:16
Speaker 4: to these powerful models. You can do it at fairly
26:19
Speaker 4: low costs if you go and get an enterprise license
26:23
Speaker 4: use against some of the managed services like Amazon Bedrock
26:28
Speaker 4: to be able to get access to this. It's fairly
26:31
Speaker 4: turnkey and can be a lot less expensive, and frankly,
26:35
Speaker 4: it brings that shadow AI into the light so that
26:40
Speaker 4: you can truly talk about it and advance the use
26:43
Speaker 4: of AI in your organization, because there are some risks.
26:46
Speaker 2: With shadow AI as well. Right, it's not just to this,
26:49
Speaker 2: you know, there is. Maybe you could phrase it frame
26:53
Speaker 2: it as a duty of care on an organizational level,
26:56
Speaker 2: maybe a duty of self protection and security. You know
27:01
Speaker 2: it needs to be about transparency.
27:02
Speaker 4: Right absolutely, I feel like people should declared AI amnesty.
27:07
Speaker 4: It's like, okay, if you're using it, we forgive you,
27:11
Speaker 4: but you have to come forward and here's the timeframe.
27:16
Speaker 2: And with my prompt into to submit, that made me
27:24
Speaker 2: do it.
27:25
Speaker 4: No exactly, it's just like again, that's the only way
27:29
Speaker 4: you can bring all the people out from the edges,
27:31
Speaker 4: and you have to give them amnesty and and so
27:35
Speaker 4: they won't get in trouble because I'm not going to
27:38
Speaker 4: talk about everybody because I've been doing this rogue.
27:40
Speaker 3: Okay, Okay, we know this has been happening.
27:42
Speaker 4: We have the data, we already know that you're using this,
27:45
Speaker 4: so just come forward, do this in the appropriate way,
27:49
Speaker 4: and and say, look, we may have to shut it
27:52
Speaker 4: down because it has these risks, or we can do
27:55
Speaker 4: it in this way. But it's it's about having this
27:59
Speaker 4: really thoughtful conversation.
28:01
Speaker 2: It's about AI. How do I religions earlier, Charlie, maybe
28:04
Speaker 2: we should call it my absolution A right way to
28:09
Speaker 2: frame it. So, where have you seen real world organizations
28:15
Speaker 2: supporting this process of transformation in imagine ty if effective ways?
28:21
Speaker 2: And I guess what advice would you give to leaders
28:23
Speaker 2: looking to support their employees with different temperaments or attitudes
28:28
Speaker 2: towards the extreme novelty and power of this technology.
28:33
Speaker 4: Yeah, we've worked with this and interviewed this company that's
28:37
Speaker 4: a very large call center and they implemented AI and
28:42
Speaker 4: they saw huge productivity gains. But it wasn't about replacing people.
28:46
Speaker 4: It was really about taking the bottom twenty five percent,
28:51
Speaker 4: for example, and bring them up to the level.
28:53
Speaker 3: Of their top performers.
28:55
Speaker 4: And it was really focused on helping people do their
28:58
Speaker 4: best work. And so there are productivity games because of it,
29:03
Speaker 4: and so they can absolutely anticipate that their numbers will
29:06
Speaker 4: go down over time. But they were saying that we're
29:10
Speaker 4: just going to manage that through attrition. And we know
29:12
Speaker 4: that some jobs have changed. We know that jobs are
29:15
Speaker 4: going to change. Some jobs are just going to go away.
29:18
Speaker 4: Somebody gave me the example of reviewing checks, like somebody
29:22
Speaker 4: actually used to look at checks and key them into
29:26
Speaker 4: a system. And now with AI and computer vision, those
29:30
Speaker 4: jobs are just going to go away. There's no way
29:33
Speaker 4: of just it's saying that it's not going to be there.
29:37
Speaker 4: So you need a small number of people to look
29:39
Speaker 4: at the exceptions and this is a reality and we
29:43
Speaker 4: need to prepare ourselves as an organization for that. And
29:48
Speaker 4: again this has been the case for decades now with
29:52
Speaker 4: technology coming in. This is not new and unique to AI,
29:56
Speaker 4: and so having the responsibility to take care of your
30:00
Speaker 4: people I think is a really important leadership trade here
30:03
Speaker 4: to be able to say change is coming.
30:07
Speaker 3: We'll do our.
30:07
Speaker 4: Best to upskill you, to reskal you, but overall, there
30:12
Speaker 4: may be some jobs that just simply aren't going to
30:16
Speaker 4: be here. So we can't guarantee your job, we can't
30:19
Speaker 4: guarantee your role is going to be here, but we
30:21
Speaker 4: can do our best to prepare you for this new
30:23
Speaker 4: future because you understand our organization. You're a valuable worker,
30:28
Speaker 4: because you've been coming here and you know us. We
30:31
Speaker 4: know you to take them possible.
30:33
Speaker 2: We'd love to keep you very cool.
30:36
Speaker 4: And I think it's irresponsible of companies to give two
30:40
Speaker 4: week notices. It's like saying you just don't have your
30:43
Speaker 4: app together first as saying we can see this change happening,
30:47
Speaker 4: we're going to prepare for it.
30:48
Speaker 3: There's a long runway. These are the ways we can
30:51
Speaker 3: do this.
30:55
Speaker 2: I became interested in recent we actually had some of
30:57
Speaker 2: the show was talking about study about how they'd showed
31:02
Speaker 2: for excessive use or use of AI can sometimes sometimes
31:06
Speaker 2: have detrimental effects on creativity. And you know, and I
31:11
Speaker 2: couldn't I could imagine this being true in some circumstances, right.
31:16
Speaker 2: Could you imagine a future where there's mitigation of the
31:22
Speaker 2: use of AI or control the fuse of AI within
31:24
Speaker 2: organizations where they're looking at specific tasks and they're maybe
31:28
Speaker 2: providing prompts above the AI app saying you know, look,
31:31
Speaker 2: maybe step away, maybe carve out some more more autonomy
31:35
Speaker 2: in this process. Similarly, with you see stories charline about
31:41
Speaker 2: AI addiction, AI over use, say, unhealthy attachments to it.
31:45
Speaker 2: These these are all you know, There's going to be
31:49
Speaker 2: so many known unknowns that come about from a new
31:52
Speaker 2: powerful technology such as this that could have implications for
31:57
Speaker 2: HR departments, for leaderships. In my position, what are you
32:00
Speaker 2: forced in this direction?
32:02
Speaker 4: Well, again, this is where I think education helps tremendously,
32:07
Speaker 4: because there was just an article in the New York
32:11
Speaker 4: Times talking about how somebody became very delusional about their
32:15
Speaker 4: ability to do math because AI is kind of like
32:18
Speaker 4: telling you, yeah, you're a math genius. And then when
32:20
Speaker 4: I actually put the formulas and everything into another AI.
32:25
Speaker 4: It's that, oh no, this is bunk, like he goes,
32:27
Speaker 4: why didn't I think to do that?
32:29
Speaker 3: Like?
32:29
Speaker 4: How did I get sucked into this? And again, one
32:33
Speaker 4: of the best practices is critical thinking. We cannot let
32:37
Speaker 4: go of critical thinking because AI is just going to
32:40
Speaker 4: tell us exactly what it is that we want to hear.
32:44
Speaker 4: It's designed to do that for beautiful words. Now we
32:46
Speaker 4: can talk about designing AI so that the objectives are
32:51
Speaker 4: not just to answer the things that you want, but
32:54
Speaker 4: to also look at a body of knowledge that's considered
32:57
Speaker 4: quote right, whatever that means. Again, if we can't even
33:00
Speaker 4: agree on what is right, because there are multiple versions of.
33:03
Speaker 3: That, it is very hard for AI to do that.
33:05
Speaker 4: So I go back to an organization defining what does
33:09
Speaker 4: right look like? And I'll give you an example of fairness.
33:12
Speaker 4: We talk about bias being built into AI. There is
33:16
Speaker 4: bias in everything that we do because we are human
33:19
Speaker 4: and AI was built by humans. We come and show
33:22
Speaker 4: up with our biases based on our experiences. They're not
33:26
Speaker 4: good and bad necessarily, but there are biases, and so
33:31
Speaker 4: understanding what fairness means to an organization because there is
33:35
Speaker 4: no universal definition of fairness, What does fairness mean? What
33:39
Speaker 4: does unbiased mean? And then building that into your AI
33:43
Speaker 4: and to the way that it delivers results is one
33:46
Speaker 4: way to mitigate those inavertent points that you don't want
33:51
Speaker 4: to have happening. So I do believe that teaching people
33:56
Speaker 4: about critical thinking, how to use AI, understand that it
34:01
Speaker 4: can hallucinate, Understand that you can fall into these straps,
34:05
Speaker 4: that it's just going to keep you the answer, so
34:07
Speaker 4: you have to ask it and say, you know, take
34:10
Speaker 4: the other perspective, challenge me on these ideas, use a
34:14
Speaker 4: different engine to check the results of another engine. So
34:17
Speaker 4: these are best practices that you can learn. But right now,
34:22
Speaker 4: you know something like seventy eight percent of employees see
34:27
Speaker 4: the future about being AI and they know it's going
34:29
Speaker 4: to impact their jobs and they want training, and only
34:32
Speaker 4: thirty one percent say they're getting the training that they need.
34:35
Speaker 4: It's a huge gap, and then disconnect. Organizations are just
34:38
Speaker 4: not investing in the training that people need in order
34:42
Speaker 4: to be successful with AI.
34:44
Speaker 2: Fascinating Sholing, I'm soorry. I'm so glad we had to
34:48
Speaker 2: have you back on because I had to have a
34:49
Speaker 2: chance to talk to you. It's not fair. It wasn't
34:51
Speaker 2: fair that any team go to do it. So that
34:54
Speaker 2: was absolutely scintillating conversation. Where can people find you in
34:57
Speaker 2: your work?
35:00
Speaker 4: And find me at my website Charlie Lee dot com.
35:02
Speaker 4: This is my name. I am also very active on LinkedIn.
35:05
Speaker 4: It's a great place to connect with me. So hope
35:07
Speaker 4: to connect with and many of you. Please write me.
35:10
Speaker 4: I tell people connect with me, write me. I read
35:13
Speaker 4: everything and almost nobody does so, but I'm very sucere
35:17
Speaker 4: about business.
35:18
Speaker 2: This is all right to you tomorrow, okay, very correact
35:21
Speaker 2: and I hear you. I hear you're in You're in
35:23
Speaker 2: the UK presently where I am?
35:25
Speaker 3: Yes, I'm right in the middle of London right now.
35:27
Speaker 2: Yeah, I'm in London too. So yeah. Nice to connect
35:31
Speaker 2: on this in the same city, even if virtually, and
35:33
Speaker 2: a pleasure to meet and talk to you.
35:35
Speaker 3: Thank you so much for having me.
35:36
Speaker 2: Thank you, Shenny. Thank you.
35:38
Speaker 1: To make sure that you never miss an episode, subscribe
35:41
Speaker 1: to the show in Apple Podcasts, Spotify for your favorite
35:44
Speaker 1: podcast player, and if you're listening on Apple podcasts, make
35:47
Speaker 1: sure to leave a rating of the show. Just have
35:49
Speaker 1: the number of stars you think the podcast deserves. If
35:52
Speaker 1: you'd like to learn more about how next Thing can
35:54
Speaker 1: help me improve your digital employee experience, head over to
35:58
Speaker 1: next think dot com. Thank you so much for listening.
36:00
Speaker 1: Until next time, M