How AI Is Quietly Rewriting Every Job
What if AI isn't replacing jobs, but quietly redesigning the tasks inside them?
In the most recent HR Leaders Podcast episode, I had an important conversation with Sultan Murad Saidov, Former CEO / Founder and Investor at Beamery. He explains why task intelligence could become critical to workforce planning, how AI is changing the work inside existing roles, and why deciding what gets automated should become an HR question, not just a technology question.
Drawing on more than a decade of building Beamery and working at the intersection of AI, skills and workforce data, Sultan explains why skills alone are no longer enough to understand where work is heading.
5 things you’ll learn from this episode:
Why looking at tasks instead of job titles changes how leaders understand AI's impact on work
The research Sultan references suggesting 75% of jobs may need redesign because of AI augmentation
How digital twin teams could help leaders model workforce decisions before making them
Why reducing junior roles may create a long-term talent and career pipeline problem
Why Sultan believes HR and technology teams should jointly lead the redesign of work
What if AI agents were not just another way to create more learning content?
What if AI agents could help talent teams find real skills gaps, build the right intervention, and deliver it at the moment employees need it most?
That is what Arist is building with the first end-to-end AI agent for enablement.
Instead of relying on slow needs analysis, outdated content libraries, and one-size-fits-all training, Arist helps organisations move from guessing what people need to identifying real gaps in real time.
Its AI agents can interview employees, uncover talent gaps, create personalised learning, and deliver it directly through tools like SMS and MS Teams.
Because the future of enablement will not be built by more content or more manual workflows. It will be built by teams that can spot business needs faster, deliver learning that fits the moment, and prove impact.
With Arist, AI agents become a way to close skills gaps faster, improve performance, and help employees get the right learning before the business has already moved on.
WEBVTT
1
00:00:00.151 --> 00:00:05.760
We need to treat this as a central decision-making process for how are we
2
00:00:05.790 --> 00:00:09.390
redesigning work, because that's what agents are doing, and that should be an HR
3
00:00:09.480 --> 00:00:13.395
question, not a tool question or just a single line of business question,
4
00:00:13.831 --> 00:00:17.581
particularly because these things interact across the company, not just within the
5
00:00:17.610 --> 00:00:21.780
function. Now, a lot of companies are starting to create these sort of hybrid HRIT,
6
00:00:21.960 --> 00:00:26.701
HRAI pockets. That should be more the norm in every company.
7
00:00:26.790 --> 00:00:31.590
So I think that sort of hybrid of tech and HR teams tackling this problem is
8
00:00:31.830 --> 00:00:32.580
probably the best way.
9
00:00:48.630 --> 00:00:50.970
Hey, Salta, welcome to the show, my friend. How are you?
10
00:00:51.090 --> 00:00:52.830
I am fantastic. Good to see you, Chris.
11
00:00:52.981 --> 00:00:56.759
I can't believe that you've been in London this entire time, and we haven't met.
12
00:00:56.821 --> 00:00:58.080
Yeah, around the corner too.
13
00:00:58.290 --> 00:00:58.651
I know.
14
00:00:59.160 --> 00:00:59.580
Less than a week or so.
15
00:00:59.640 --> 00:01:01.410
Yeah. Us antisocial Londoners.
16
00:01:03.630 --> 00:01:05.040
Anyway, how you been, first and foremost?
17
00:01:05.640 --> 00:01:11.610
I've been well. It's a good time to be in tech, AI, HR, all the things that
18
00:01:11.640 --> 00:01:14.040
I do, and definitely a good time to be living in London.
19
00:01:14.100 --> 00:01:18.180
Yeah. I love the way you're the opposite of all the CTOs I interview every day.
20
00:01:18.810 --> 00:01:23.760
They're like: "It's a crazy time to be in AI and I'm drowning.
21
00:01:23.821 --> 00:01:27.660
I don't know what I'm doing, and I'm overwhelmed." And we're like: "It's an amazing
22
00:01:27.720 --> 00:01:29.940
time to be..."
23
00:01:29.970 --> 00:01:32.700
Well, I think it goes just in a different order.
24
00:01:32.836 --> 00:01:38.310
I probably felt the it's overwhelming in early '23 when we were
25
00:01:39.000 --> 00:01:42.150
grappling with how fast AI suddenly started working.
26
00:01:42.180 --> 00:01:44.940
And I think now we've had enough time to sort of settle into it and make sure we're
27
00:01:45.120 --> 00:01:51.090
doing the right things with it. But I think there's a joy to be
28
00:01:51.150 --> 00:01:53.730
found when problems are hard, but important.
29
00:01:54.510 --> 00:01:59.295
Yeah. I know, I love it. For me, it's difficult because one of the things we talk
30
00:01:59.310 --> 00:02:01.230
about internally here is seek discomfort.
31
00:02:03.211 --> 00:02:07.215
Shane and I've always, even as when we were athletes, we always seek discomfort,
32
00:02:07.260 --> 00:02:08.866
because that's where the growth and the magic happens.
33
00:02:09.211 --> 00:02:12.660
It doesn't feel good when you're in it, but when you come out the other side,
34
00:02:13.290 --> 00:02:15.000
that's when the magic really happens, right?
35
00:02:15.330 --> 00:02:15.480
Sure.
36
00:02:15.660 --> 00:02:20.145
Yeah, as well. So it's just kind of building up that, putting the reps in and
37
00:02:20.190 --> 00:02:24.060
building up that resilience, so when you are in it.
38
00:02:24.360 --> 00:02:27.300
The key thing is to know when you are in a sprint.
39
00:02:28.650 --> 00:02:32.610
Because if you don't, then you can burn out along the way.
40
00:02:32.700 --> 00:02:35.760
But anyways, before we get into the details, tell everyone a bit about your
41
00:02:35.820 --> 00:02:39.480
background personally and your journey to where we are now.
42
00:02:40.171 --> 00:02:46.005
So I started Beamerie about 12 years ago, and it
43
00:02:46.050 --> 00:02:51.615
was while I was working in finance at Goldman Sachs during the tail end of the last
44
00:02:51.690 --> 00:02:57.375
recession. And the genesis of the company was actually not with the intention of
45
00:02:57.630 --> 00:02:59.970
starting a software company or AI company.
46
00:03:00.000 --> 00:03:05.476
It was initially an experiment, because I was doing some coaching for kids from
47
00:03:05.520 --> 00:03:10.305
underprivileged backgrounds to help get them jobs in the city, and suddenly lots of
48
00:03:10.500 --> 00:03:12.360
people are unemployed, lots of careers are cut.
49
00:03:12.840 --> 00:03:18.390
And the experiment was, can you look at labor market data to figure out what
50
00:03:18.451 --> 00:03:20.610
actually makes one job similar to another?
51
00:03:21.150 --> 00:03:24.510
Particularly living in London and working in finance at the time, everyone I knew
52
00:03:24.541 --> 00:03:29.101
was called an analyst or an associate, and these sort of amorphous titles and jobs.
53
00:03:29.220 --> 00:03:34.710
And obviously, you had this sort of growing Cambrian explosion of startups and tech
54
00:03:34.740 --> 00:03:39.345
and different types of jobs. And I'd started a company before, which
55
00:03:40.770 --> 00:03:43.845
meant that I did a bit of coding, a bit of data analysis, and essentially this
56
00:03:43.860 --> 00:03:48.360
experiment showed that you could look at which jobs are adjacent to other jobs and
57
00:03:48.390 --> 00:03:50.744
look at sort of underlying things like skills within jobs.
58
00:03:51.451 --> 00:03:56.835
And I then just started trying to speak to any HR leaders that I could meet or that
59
00:03:56.835 --> 00:04:02.490
would listen and try to understand what is it that makes us not really look at data
60
00:04:02.520 --> 00:04:06.525
when it comes to people decisions, why things sort of are so based on precedent and
61
00:04:06.541 --> 00:04:07.860
where people went to school and so forth.
62
00:04:08.430 --> 00:04:12.675
And one thing led to another, and it led to us essentially trying to figure out how
63
00:04:12.690 --> 00:04:17.970
do you try and take the principles of what make companies smart and data-driven in
64
00:04:18.000 --> 00:04:20.955
how they approach their customers, and then start treating their employees and
65
00:04:20.970 --> 00:04:23.940
their candidates like customers. And that was the genesis of Beamerie.
66
00:04:24.600 --> 00:04:29.701
Love that. And it's so interesting how you never know where you are now, right?
67
00:04:30.060 --> 00:04:35.265
Back then, looking at where you started, was it more of a challenge back then?
68
00:04:35.310 --> 00:04:39.105
Because the type of conversation you just described is very top of mind right now,
69
00:04:39.180 --> 00:04:41.550
and everyone's kind of talking about this.
70
00:04:41.610 --> 00:04:45.945
But 10 years ago, you're probably entering rooms with HR leaders and others
71
00:04:45.960 --> 00:04:48.570
thinking, "What is this guy talking about?"
72
00:04:50.190 --> 00:04:50.369
Yeah.
73
00:04:50.640 --> 00:04:52.244
It's so far away from-
74
00:04:52.320 --> 00:04:57.750
It's interesting to look back. Definitely, what people generally wanted is their
75
00:04:57.781 --> 00:04:58.875
job boards to work better and-
76
00:04:58.951 --> 00:04:59.701
Yeah, exactly.
77
00:05:00.270 --> 00:05:01.560
Yeah.
78
00:05:01.574 --> 00:05:01.574
Yeah.
79
00:05:01.980 --> 00:05:06.781
But it happened quite quickly. I think the interesting thing 12 years ago was also
80
00:05:06.900 --> 00:05:12.540
being a company that wasn't based in San Francisco was odd if you're trying to be a
81
00:05:12.600 --> 00:05:13.140
tech company.
82
00:05:13.531 --> 00:05:14.100
Yeah.
83
00:05:14.341 --> 00:05:19.725
Even that was odd. My co-founder is my brother, and the two of us moved out and
84
00:05:19.770 --> 00:05:22.470
lived in New York and San Francisco for about six months.
85
00:05:23.190 --> 00:05:28.500
And one of the things that actually made us build our team, our core technology
86
00:05:28.560 --> 00:05:31.350
team in the UK, is that it was a talent advantage.
87
00:05:31.410 --> 00:05:37.350
And even more so, over time, as London became a center for AI thanks to
88
00:05:37.380 --> 00:05:43.230
companies like Google. But at the time, we were essentially a non-California
89
00:05:43.290 --> 00:05:47.115
AI company that was trying to come up with a new product in HR.
90
00:05:47.400 --> 00:05:48.600
The first couple of years were super hard.
91
00:05:49.455 --> 00:05:49.455
Yeah.
92
00:05:49.470 --> 00:05:54.360
And actually, what ended up helping us get through that was partnering.
93
00:05:54.570 --> 00:05:57.720
We ended up taking the approach of we don't want to be all things to all people.
94
00:05:57.750 --> 00:06:01.800
We want to be the best at AI and data and the people side of what we look at.
95
00:06:01.860 --> 00:06:05.940
And we ended up partnering with companies like Microsoft and Workday, who both
96
00:06:05.970 --> 00:06:09.570
invested in us, and I think that approach helped us to stay focused on the problem
97
00:06:09.600 --> 00:06:12.180
and not try to convince everybody about everything.
98
00:06:12.240 --> 00:06:14.400
Yeah. I'm going through a similar process
99
00:06:16.230 --> 00:06:19.230
right now, so that makes sense. One of the things I'm really excited to talk to you
100
00:06:19.290 --> 00:06:25.081
about is the rise of task intelligence, which you guys are at the forefront.
101
00:06:25.095 --> 00:06:29.910
Probably for our audience, probably good to describe what is that first, and then
102
00:06:29.939 --> 00:06:32.341
we can jump into it, because maybe it'll be the first time some of them have come
103
00:06:32.369 --> 00:06:33.091
across that phrase.
104
00:06:33.690 --> 00:06:37.320
I'll start by saying that the thing I just mentioned of how we started out by
105
00:06:37.350 --> 00:06:43.320
looking at how jobs evolved back 12 years ago to figure out how can you
106
00:06:43.350 --> 00:06:47.100
help people get discovered for different types of careers.
107
00:06:47.130 --> 00:06:51.526
At the time, what we were doing is analyzing things like job descriptions and
108
00:06:51.570 --> 00:06:54.315
career transitions. And the thing about jobs is
109
00:06:56.100 --> 00:06:59.430
they don't mean much as titles. And for the last 10 years, there's also been a lot
110
00:06:59.460 --> 00:07:03.510
of talk about skills, but people have skills, but jobs aren't really described with
111
00:07:03.540 --> 00:07:04.951
skills in any way that makes sense.
112
00:07:05.716 --> 00:07:05.716
Mm-hmm.
113
00:07:05.790 --> 00:07:10.951
And tasks was actually something that consultants have been using for decades.
114
00:07:10.980 --> 00:07:16.860
There's books like "Work Without Jobs" by Ravin from Mercer, because in order to
115
00:07:16.890 --> 00:07:21.901
understand how to break down work and redesign it, we have to look at the tasks
116
00:07:21.930 --> 00:07:25.890
that make up the work. But technology providers didn't really go down that route
117
00:07:25.920 --> 00:07:29.490
because it was very hard to know what to do with that.
118
00:07:29.910 --> 00:07:35.505
You couldn't really extract tasks from jobs until recently, and it was hard to sort
119
00:07:35.520 --> 00:07:37.890
of figure out, well, why would we even bother extracting it?
120
00:07:38.490 --> 00:07:43.980
But suddenly, with the advent of AI and robotics and the speed of technology, tasks
121
00:07:44.010 --> 00:07:49.290
have become one of the most important things to understand, because AI augments
122
00:07:49.305 --> 00:07:52.470
tasks or automate tasks. It doesn't replace jobs.
123
00:07:52.591 --> 00:07:56.460
There's been studies, there's a recent one from McKinsey which highlighted that
124
00:07:56.520 --> 00:08:02.341
while only 10% of jobs are being fully automated today, 75%
125
00:08:02.460 --> 00:08:06.270
are already in need of full redesign because of the way that they're being
126
00:08:06.300 --> 00:08:11.175
augmented. And that requires us to understand how tasks don't just look today, but
127
00:08:11.190 --> 00:08:15.390
how tasks are changing. And the task intelligence refers to the principle of how do
128
00:08:15.420 --> 00:08:21.195
we understand the tasks that we need to do in our jobs, in our company, and how
129
00:08:21.240 --> 00:08:26.880
those tasks are evolving, either through technology or through the way that the
130
00:08:26.970 --> 00:08:32.160
needs of how people work alongside copilots and agents are changing.
131
00:08:32.280 --> 00:08:36.795
And it's also a way of looking at things like labor market data, not just as how
132
00:08:36.841 --> 00:08:41.160
many people are hiring for these jobs, but who are we truly competing with in terms
133
00:08:41.190 --> 00:08:45.135
of what the jobs look like? If the titles all look the same, but the work is
134
00:08:45.181 --> 00:08:49.515
changing, understanding the task intelligence helps you be more thoughtful around
135
00:08:49.590 --> 00:08:53.850
how you hire, how you train people, how you upskill people to actually where work
136
00:08:53.940 --> 00:08:57.719
is heading, rather than just where work is today, and so on.
137
00:08:57.840 --> 00:08:59.940
Yeah. We've gone a bit on this journey, right?
138
00:09:00.000 --> 00:09:03.630
So we've gone role-based, skill-based, now we're at task.
139
00:09:05.250 --> 00:09:08.880
What's the consequences for companies that don't make this transition?
140
00:09:11.172 --> 00:09:16.796
The skill-based piece only really worked in narrow contexts.
141
00:09:16.962 --> 00:09:22.842
I think the skill journey was very important, particularly in the post-2020
142
00:09:22.872 --> 00:09:27.537
era, when companies needed to figure out how do we start using some of these
143
00:09:27.537 --> 00:09:31.212
evolving AI models to match people to learning and to careers.
144
00:09:31.242 --> 00:09:35.202
And so we haven't moved beyond skill-based in the sense that skills are still the
145
00:09:35.262 --> 00:09:36.312
currency-
146
00:09:36.327 --> 00:09:36.327
Sure
147
00:09:36.327 --> 00:09:37.258
... that helps us understand
148
00:09:38.802 --> 00:09:42.387
what employees and candidates may be good at, and how systems speak to each other.
149
00:09:43.092 --> 00:09:46.842
I think the task piece is what actually allows us to scale to all decisions and
150
00:09:46.872 --> 00:09:52.812
jobs, and actually connect the way that HR teams support their businesses to
151
00:09:52.872 --> 00:09:56.922
business decisions. Because the language of the business is work, not skills, the
152
00:09:56.952 --> 00:09:59.952
language of whether you're bidding on a project or trying to decide
153
00:10:01.212 --> 00:10:05.113
what is our goal in this department or for the company.
154
00:10:05.127 --> 00:10:10.633
And so I think the opportunity that is already happening with the
155
00:10:10.782 --> 00:10:16.542
view of tasks is it's actually putting AI and HR
156
00:10:16.633 --> 00:10:22.287
using AI at the center of business decisions rather than being something that is
157
00:10:22.332 --> 00:10:26.846
being driven just by the technology teams or the CIO or others who are kind of
158
00:10:26.863 --> 00:10:29.562
deploying tools without understanding how they impact people.
159
00:10:29.577 --> 00:10:29.636
Mm-hmm.
160
00:10:29.772 --> 00:10:34.407
And so I think for the current chapter that HR is in, task intelligence is probably
161
00:10:34.452 --> 00:10:37.182
one of the biggest opportunities to elevate in business impact.
162
00:10:37.602 --> 00:10:42.072
Yeah. With us being kind of hybrid remote, does that also make a difference now,
163
00:10:42.102 --> 00:10:47.607
the ability to be able to really truly leverage our workforce globally, to be able
164
00:10:47.652 --> 00:10:53.532
to match tasks and skills to work? Is that also something you see as a big
165
00:10:53.592 --> 00:10:55.002
opportunity as well?
166
00:10:55.062 --> 00:11:00.972
I think the opportunity for having the best people, no matter where
167
00:11:01.032 --> 00:11:01.422
they are-
168
00:11:01.633 --> 00:11:01.647
Yeah
169
00:11:01.647 --> 00:11:07.302
... to do work is definitely something that is being enabled by
170
00:11:07.782 --> 00:11:11.532
being able to work remotely and hire, and train talent remotely.
171
00:11:12.192 --> 00:11:18.102
I do think that there's more nuance to consider for how do teams actually work
172
00:11:18.162 --> 00:11:22.512
best together on a problem, and how do teams work together alongside technologies
173
00:11:22.542 --> 00:11:28.242
and tools. The way that a company that is building
174
00:11:28.272 --> 00:11:32.772
physical objects will think about how do you optimize the components you have, you
175
00:11:32.802 --> 00:11:35.682
can suddenly get the best component from a different country, just like you could
176
00:11:35.712 --> 00:11:40.812
hire a remote person. But the way you think about it today would be by digitally
177
00:11:40.842 --> 00:11:45.282
modeling it. There are companies like Fincantieri that build digital twin ships.
178
00:11:45.642 --> 00:11:48.042
Rather than saying, "We're going to order all these components and it's going to be
179
00:11:48.133 --> 00:11:52.827
better," they simulate it. And they then can say, "Well, before building the ship,
180
00:11:52.902 --> 00:11:56.532
we're going to simulate the weather, the water," and it's something that's in
181
00:11:56.592 --> 00:11:58.242
manufacturing been known as digital twins-
182
00:11:58.752 --> 00:11:58.962
Yeah
183
00:11:58.977 --> 00:12:03.597
... for decades. I think the way I would look at your question of how do you think
184
00:12:03.672 --> 00:12:07.633
about whether remote is best, instinctively it is, but you can look at it in a more
185
00:12:07.662 --> 00:12:12.613
data-driven way by actually modeling digital twin teams and digital twin
186
00:12:12.672 --> 00:12:13.363
organizations.
187
00:12:13.932 --> 00:12:13.947
Mm-hmm.
188
00:12:14.202 --> 00:12:17.442
And tasks is actually one of the things that enables that, because you can model
189
00:12:17.532 --> 00:12:21.162
digitally the tasks that are being done and need to be done and model out the
190
00:12:21.192 --> 00:12:23.802
people we have, the skills they have, whether they're remote, and how they work.
191
00:12:24.732 --> 00:12:28.557
And then you can answer those types of questions in a much more precise and
192
00:12:28.872 --> 00:12:33.402
forward-looking way, and go beyond kind of the instinct of remote teams that should
193
00:12:33.432 --> 00:12:36.372
be effective to where would that be most effective and where would that be a
194
00:12:36.402 --> 00:12:36.702
problem.
195
00:12:37.152 --> 00:12:40.467
Wow. You just reminded me of a conversation I had a while back with Boeing, and
196
00:12:40.512 --> 00:12:44.052
they showed me the digital twin of their airplane.
197
00:12:44.113 --> 00:12:44.442
Yeah.
198
00:12:44.502 --> 00:12:47.786
And the fact that they build the entire airplane from the ground upwards into
199
00:12:47.802 --> 00:12:53.157
software, and everyone globally can work on building a plane digitally before they
200
00:12:53.157 --> 00:12:55.946
even think about going to manufacturing or building it out.
201
00:12:56.412 --> 00:12:58.376
Just kind of reminded me of that, which-
202
00:12:58.392 --> 00:13:04.272
And it's exactly what's starting to happen in HR today because the tools to do
203
00:13:04.333 --> 00:13:07.917
a digital twin team or organization, just like the Boeing example you gave, are
204
00:13:07.917 --> 00:13:09.252
already here. That's exactly-
205
00:13:09.297 --> 00:13:09.297
Yeah
206
00:13:09.297 --> 00:13:11.352
... what task intelligence is starting to unlock.
207
00:13:11.412 --> 00:13:16.062
Wow. This is super interesting. Are companies ready for this, though, in terms of
208
00:13:16.092 --> 00:13:19.722
the way their current tech stack is set up, the way their current data, where their
209
00:13:19.752 --> 00:13:23.172
data is currently stored? Because this sounds incredible, but I feel like you're
210
00:13:23.202 --> 00:13:25.602
going to a company and data's scattered
211
00:13:26.863 --> 00:13:28.241
all over the place.
212
00:13:28.722 --> 00:13:34.303
In terms of the systems that companies have today for people
213
00:13:34.392 --> 00:13:36.762
records, job records, analytics, no.
214
00:13:37.317 --> 00:13:43.196
But in terms of the things you need to actually be able to unlock this,
215
00:13:44.202 --> 00:13:49.962
it's bypassing existing data stores and problems and systems.
216
00:13:50.053 --> 00:13:54.852
There's a great example of, you're probably familiar with HubSpot.
217
00:13:55.092 --> 00:13:55.812
Yeah, of course.
218
00:13:55.962 --> 00:14:00.252
But HubSpot was one of the first companies that connected what's known as an MCP,
219
00:14:00.312 --> 00:14:04.767
which essentially a Model Context Protocol, is a new form of being able to
220
00:14:04.812 --> 00:14:09.522
integrate not just systems like an HRIS and LMS, but essentially AI models.
221
00:14:09.733 --> 00:14:11.682
There's also kind of agent integrations.
222
00:14:11.712 --> 00:14:16.182
And what HubSpot did is connect ChatGPT to HubSpot directly, and it basically said
223
00:14:16.242 --> 00:14:20.142
you don't need to suddenly figure out where do we put in all of this marketing
224
00:14:20.172 --> 00:14:23.892
data. And HubSpot is a lot like a CRM or a marketing tool that you use-
225
00:14:24.072 --> 00:14:24.147
Yeah
226
00:14:24.162 --> 00:14:29.712
... in HR. But it's an example of where they solved the problem of how do you make
227
00:14:29.742 --> 00:14:33.282
the data accessible and relevant by connecting directly into AI models.
228
00:14:33.702 --> 00:14:38.382
And there are similar things starting to happen with topics like tasks and task
229
00:14:38.412 --> 00:14:44.036
intelligence, because tasks are too nuanced to push into a job
230
00:14:44.113 --> 00:14:46.062
profile or an employee profile effectively.
231
00:14:46.077 --> 00:14:49.572
Like, it's tasks linked to groups of skills and automation potential.
232
00:14:49.602 --> 00:14:55.422
There's a lot of data about them, and it may take years to effectively house that
233
00:14:55.483 --> 00:14:59.233
in the way that we typically use HR systems and workforce planning systems.
234
00:14:59.262 --> 00:15:02.053
But suddenly you don't need to in order to be able to do some of these
235
00:15:02.172 --> 00:15:03.553
sophisticated things.
236
00:15:03.567 --> 00:15:03.567
Mm.
237
00:15:03.642 --> 00:15:08.847
And it does today take some engineers with some familiarity about how to do this
238
00:15:08.922 --> 00:15:14.727
within a company. But it's happening in every company outside of HR departments and
239
00:15:14.772 --> 00:15:18.057
how companies enable their salespeople and how companies are interacting with
240
00:15:18.072 --> 00:15:23.292
clients. So I think we're less than 12 months away from the problem of messy data
241
00:15:23.307 --> 00:15:26.472
and where do we put our data not really being a problem-
242
00:15:26.833 --> 00:15:26.892
Wow
243
00:15:26.907 --> 00:15:30.492
... for these types of use cases. It'll still be a problem for governance of you
244
00:15:30.553 --> 00:15:31.392
still need to have-
245
00:15:31.422 --> 00:15:31.542
Sure
246
00:15:31.572 --> 00:15:33.462
... data and payroll and how you do a lot of things.
247
00:15:34.152 --> 00:15:39.132
But for something like analytics and insights, it's becoming very easy to put the
248
00:15:39.192 --> 00:15:43.678
right data into the right places without having to rely on your systems being
249
00:15:43.692 --> 00:15:44.442
perfect.
250
00:15:44.502 --> 00:15:49.632
Yeah. I think my CTO was trying to explain something similar the other day to me,
251
00:15:49.662 --> 00:15:51.613
which you just explained very well.
252
00:15:52.062 --> 00:15:56.022
But for HR in particular, I think that's a really huge opportunity because I think
253
00:15:56.083 --> 00:15:59.472
there's no HR team out there that hasn't at some point been paralyzed by the
254
00:15:59.502 --> 00:16:00.972
difficulty of integrating systems-
255
00:16:01.092 --> 00:16:02.083
100%
256
00:16:02.113 --> 00:16:06.928
... and structure. And so I think the unlock of this being reliable, safe, and easy
257
00:16:07.032 --> 00:16:10.632
to model and generate insights around is significant.
258
00:16:11.172 --> 00:16:16.962
Do you think there's enough trust right now with companies to
259
00:16:17.022 --> 00:16:19.392
take that step, though, even if they can?
260
00:16:19.483 --> 00:16:19.767
Because I-
261
00:16:20.623 --> 00:16:22.422
I think there's a difference between
262
00:16:23.772 --> 00:16:29.757
using insights versus where you start to try to build agents or automate worker
263
00:16:29.803 --> 00:16:35.727
decisions. Because I think the boundary of automating decisions, even where
264
00:16:35.772 --> 00:16:40.632
the decision is safe, is something that should require a high boundary of
265
00:16:40.662 --> 00:16:42.777
validation, trust, people being in the review.
266
00:16:43.542 --> 00:16:47.772
But for insights, if you can improve the quality of labor market insights and make
267
00:16:47.803 --> 00:16:51.972
it more relevant to your company, and even if it's not perfect, but it's
268
00:16:51.988 --> 00:16:55.932
significantly better than what you had before and it's safe to run, I think the
269
00:16:56.053 --> 00:17:00.416
barrier for that to be something people get value from and can safely use is very
270
00:17:00.432 --> 00:17:01.242
low.
271
00:17:01.333 --> 00:17:06.087
Yeah, I love that, the data versus insights, because I think a lot of people are
272
00:17:06.162 --> 00:17:08.202
looking at both of those through the same lens.
273
00:17:09.883 --> 00:17:13.722
What about the one of the big things that we talk about a lot is really how do you
274
00:17:13.782 --> 00:17:16.812
balance the human-centered AI in talent decisions?
275
00:17:18.252 --> 00:17:19.902
That's a tough one right now.
276
00:17:19.932 --> 00:17:25.123
Yeah, for me, it's an interesting question because we started out 12 years ago
277
00:17:25.242 --> 00:17:29.263
building AI to make fairer people decisions.
278
00:17:29.532 --> 00:17:32.892
It's why we started looking at skill inferences 12 years ago.
279
00:17:33.553 --> 00:17:34.932
But back then,
280
00:17:36.312 --> 00:17:39.942
the ideas were simpler, and the AI was simpler.
281
00:17:40.362 --> 00:17:46.182
We were looking at can you compare how unbiased a decision is likely
282
00:17:46.242 --> 00:17:49.722
to be if you look at skills and degrees and all the stuff we now
283
00:17:51.222 --> 00:17:56.638
hear about a lot in HR circles. Today, it's a little bit different because we need
284
00:17:56.682 --> 00:18:02.592
to be very careful around how people actually use
285
00:18:02.622 --> 00:18:07.947
technologies, because making a recommendation to a
286
00:18:07.962 --> 00:18:13.212
person to say, "Hey, take a look at this Take a look at this candidate, take a look
287
00:18:13.242 --> 00:18:16.182
at this recommendation of a learning course, whatever it might be.
288
00:18:17.053 --> 00:18:22.962
Even if you transparently show how the AI works and so forth, it has to be
289
00:18:23.022 --> 00:18:27.897
something that is very carefully monitored for where does it lead people, because
290
00:18:27.942 --> 00:18:33.522
we're becoming used to AI in a way that varies obviously by how much you trust it.
291
00:18:33.582 --> 00:18:38.202
But it creates risk for how people interact with any recommendations, even if the
292
00:18:38.263 --> 00:18:43.362
AI is unbiased in how it's been tested, it doesn't mean people won't end up going
293
00:18:43.392 --> 00:18:46.962
down a path that creates biased outcomes and so on in their own context.
294
00:18:47.053 --> 00:18:52.707
One of the things that we've tried to do is really separate the areas where
295
00:18:52.782 --> 00:18:56.952
AI can do something completely safe, like improve the quality of your data or
296
00:18:56.982 --> 00:19:02.292
improve the quality of how you are nudged to take a look at something.
297
00:19:02.502 --> 00:19:06.882
Things that are without question not introducing any real risks in the decisions
298
00:19:06.912 --> 00:19:11.803
that you're making. And even automation, having something that automates a report
299
00:19:11.816 --> 00:19:14.546
that you look at every day is very different to something that automates something
300
00:19:14.592 --> 00:19:19.888
that happens for you. And I think those boundaries of how people don't just look at
301
00:19:19.962 --> 00:19:22.662
AI for like, is it audited? Is it biased?
302
00:19:22.722 --> 00:19:27.763
But actually, what things is it doing that otherwise we'd be comfortable with
303
00:19:27.792 --> 00:19:33.132
people doing safely, and what tasks do we actually want to automate rather than we
304
00:19:33.162 --> 00:19:36.342
don't want to automate? And so I think those conversations are something that is
305
00:19:36.372 --> 00:19:41.382
probably important for HR teams to help lead rather than for this stuff to just be
306
00:19:41.502 --> 00:19:44.382
a question of technologies and how audited are they.
307
00:19:44.892 --> 00:19:48.582
Yeah. Because I think the struggle we're having right now is maintaining that trust
308
00:19:48.763 --> 00:19:51.972
with employees while scaling workforce innovation, right?
309
00:19:52.242 --> 00:19:54.492
That's kind of the key thing I'm hearing out from them.
310
00:19:54.747 --> 00:19:59.442
It's less of a technology transformation and more of a mindset shift and a cultural
311
00:19:59.502 --> 00:20:00.192
transformation.
312
00:20:00.732 --> 00:20:01.782
Yeah, I feel like the
313
00:20:04.152 --> 00:20:08.772
way that people have been using AI as consumers is obviously falling into
314
00:20:09.882 --> 00:20:14.982
natural extremes. Some people who now live with ChatGPT as a friend and other
315
00:20:15.013 --> 00:20:18.763
people who avoid touching these things.
316
00:20:19.692 --> 00:20:24.912
I find that boundary interesting myself because-- So my wife works in AI governance
317
00:20:25.002 --> 00:20:25.122
and-
318
00:20:25.377 --> 00:20:25.797
Oh, really?
319
00:20:26.217 --> 00:20:29.907
... monitoring AI safety. And so our conversations at home often touch on this sort
320
00:20:29.952 --> 00:20:34.737
of topic. We don't have any devices in our home with listening tools, not because I
321
00:20:34.782 --> 00:20:37.917
assume someone's listening, but because at some point, I think
322
00:20:39.282 --> 00:20:41.997
your personal threshold of what are you comfortable with sort of kick in for how
323
00:20:42.042 --> 00:20:42.942
you use these tools.
324
00:20:42.957 --> 00:20:43.632
Yeah, that's understandable.
325
00:20:43.722 --> 00:20:46.782
So I've never had a smart lock in my house and so on.
326
00:20:46.812 --> 00:20:51.072
But I do think that there are a lot of things that we've assumed we can trust.
327
00:20:51.282 --> 00:20:54.763
Take Google search results before AI without using backlinks.
328
00:20:54.822 --> 00:20:54.837
Mm-hmm.
329
00:20:55.392 --> 00:21:00.372
Was that more reliable or odd than what we're using now, which has reasoning, which
330
00:21:00.432 --> 00:21:06.057
actually fact checks, right? So I think that I am optimistic about the fact that
331
00:21:06.072 --> 00:21:09.912
the way that AI is being developed and progressing, certainly from
332
00:21:11.202 --> 00:21:13.062
a lot of the leading players like Google,
333
00:21:14.292 --> 00:21:19.392
is making it more thoughtful and reliable than the way a lot of technologies worked
334
00:21:19.422 --> 00:21:22.362
before, which has always been algorithms of some kind.
335
00:21:22.782 --> 00:21:27.013
But it doesn't mean that's how people feel about it, and I think that the question
336
00:21:27.072 --> 00:21:31.362
of how do you give people a choice of building their own trust and opting in or out
337
00:21:31.392 --> 00:21:33.642
of these tools is always going to be really important.
338
00:21:34.182 --> 00:21:38.232
Yeah. I mean, the reality is they're going to be using them anyway whether it's in
339
00:21:38.292 --> 00:21:44.202
your organization or outside. And yeah, what I take for granted is how close I
340
00:21:44.263 --> 00:21:47.877
am to the technology. I have ChatGPT built into my headphones.
341
00:21:48.672 --> 00:21:51.822
So I have a pair of headphones where it's literally without my phone, I can access
342
00:21:51.896 --> 00:21:53.803
ChatGPT at any time in my headphones.
343
00:21:53.862 --> 00:21:58.662
It's extreme to a lot of people, and I just talk to my headphones and it's all
344
00:21:58.752 --> 00:21:59.022
there.
345
00:22:00.702 --> 00:22:06.162
But there is a certain element of trust of how
346
00:22:06.852 --> 00:22:09.642
your use of these things is going to impact you.
347
00:22:09.942 --> 00:22:13.962
No matter what call you're in, you're probably having a recording or listening to-
348
00:22:13.977 --> 00:22:14.022
Yes
349
00:22:14.037 --> 00:22:17.172
... even physical devices that people wear into meetings.
350
00:22:17.263 --> 00:22:22.047
And when you consider that all of the things that are being recorded from you, all
351
00:22:22.092 --> 00:22:25.572
the work you're doing is something that's recorded and trained, is also the kind of
352
00:22:25.632 --> 00:22:29.862
stuff that is potentially going to replace what you're good at, right?
353
00:22:30.042 --> 00:22:35.562
AI being trained on-- There was a great case of AI being trained on police forces
354
00:22:35.682 --> 00:22:38.803
that recognized people in crowds.
355
00:22:38.862 --> 00:22:42.252
There was a role known as a super recognizer years ago.
356
00:22:42.537 --> 00:22:42.537
Mm.
357
00:22:42.553 --> 00:22:46.992
One of the highest-paid jobs in security, people were asked to come in and have AI
358
00:22:47.053 --> 00:22:50.428
learn how do you actually recognize faces, and obviously didn't take very long for
359
00:22:50.472 --> 00:22:54.282
suddenly that job to disappear because the AI trained on what they did.
360
00:22:54.792 --> 00:23:00.132
And the same thing is starting to apply for you using recording devices in your
361
00:23:00.192 --> 00:23:05.472
work just comfortably because that's what the AI's doing, and your company could
362
00:23:05.532 --> 00:23:07.527
learn from that and automate some of the tasks you do.
363
00:23:07.572 --> 00:23:11.803
So I think trust isn't just about do you trust the tool, it's do you trust how the
364
00:23:11.832 --> 00:23:14.967
tool is learning and that it's going to be something that's transparent and used in
365
00:23:15.013 --> 00:23:16.872
your best interest, well, what does that mean for you?
366
00:23:16.932 --> 00:23:17.202
Yeah.
367
00:23:17.263 --> 00:23:21.492
And I think that's something that creates a mandate for companies to think about
368
00:23:21.522 --> 00:23:25.242
how do we give employees using our tools visibility into where this stuff is
369
00:23:25.272 --> 00:23:29.053
heading and how this could be for their benefit and where it could be a risk.
370
00:23:29.112 --> 00:23:32.547
Yeah. One of the things I wanted to ask you about, though, is what does this mean
371
00:23:33.162 --> 00:23:38.892
for traditional org charts, annual planning cycles, the way we look
372
00:23:38.922 --> 00:23:42.763
at the role of workforce design, because it's a huge disruptor.
373
00:23:43.303 --> 00:23:49.242
So I'd answer that in two parts. I think what it has started to mean in the short
374
00:23:49.303 --> 00:23:50.667
term has
375
00:23:52.212 --> 00:23:55.332
been things that aren't sustainable and will change.
376
00:23:55.422 --> 00:24:01.152
For example, companies moving from pyramid-shaped orgs to
377
00:24:01.182 --> 00:24:05.803
diamond-shaped orgs because you don't need as many junior people.
378
00:24:05.862 --> 00:24:11.232
And where AI is today, that often makes sense because you have more senior people
379
00:24:11.292 --> 00:24:14.502
in any role starting to be more efficient with copilots and agents, not needing as
380
00:24:14.532 --> 00:24:18.582
many entry-level people. But that is neither sustainable because at some point you
381
00:24:18.612 --> 00:24:19.602
have to feed the-
382
00:24:19.662 --> 00:24:19.812
Yeah
383
00:24:19.932 --> 00:24:20.172
... growth
384
00:24:20.382 --> 00:24:21.372
Mm-hmm.
385
00:24:21.492 --> 00:24:26.622
Nor likely to be how AI is used in a few years' time
386
00:24:26.742 --> 00:24:31.182
because it only is impacting entry-level tasks and work today.
387
00:24:31.212 --> 00:24:34.662
It's going to be impacting all work, and you have to think about actually AI being
388
00:24:34.692 --> 00:24:38.142
used by entry-level people more cost-effectively in a few years' time.
389
00:24:38.592 --> 00:24:43.122
So I think there's a sort of short-term impact of overall how do we design teams
390
00:24:43.152 --> 00:24:47.682
and who's adopting this stuff, which is obviously rapidly evolving, but I think is
391
00:24:47.742 --> 00:24:50.652
not how I would think about the true impact of this stuff when you move beyond the
392
00:24:50.712 --> 00:24:53.562
next sort of 18 months. And so the second thing I think is
393
00:24:54.792 --> 00:25:00.732
much more nuanced to the kind of industry you're in
394
00:25:00.792 --> 00:25:06.372
and what you're working on. But we're starting to see the traditional principles
395
00:25:06.582 --> 00:25:12.537
of your career is built on the size of the orgs you manage and
396
00:25:12.582 --> 00:25:16.662
the size of teams shifting away, right?
397
00:25:17.472 --> 00:25:21.702
We're talking in the tech world about one-person unicorns, and there are now
398
00:25:21.763 --> 00:25:26.517
companies with five or 10 people who are managing hundreds of million in revenue
399
00:25:26.562 --> 00:25:31.002
and so forth. And I think the same principle is how I'd look at what does it mean
400
00:25:31.032 --> 00:25:36.102
for org charts because obviously in some companies you can't just say, "Oh, we'll
401
00:25:36.162 --> 00:25:41.352
all move to flat org charts," or these sort of aspersions around how people
402
00:25:41.382 --> 00:25:46.662
interact. But you can reliably see that the principle of career progression and
403
00:25:46.722 --> 00:25:52.212
size of org going away means that there will be far more people in the position
404
00:25:52.362 --> 00:25:56.143
of being individual contributors who are managing technologies like you would in
405
00:25:56.203 --> 00:26:01.872
org, and then people who are going to be in much more true manager coach
406
00:26:01.902 --> 00:26:04.002
style, managing people who are using tools.
407
00:26:04.692 --> 00:26:09.852
And I do think that the skillsets of what makes a truly good manager will probably
408
00:26:09.882 --> 00:26:13.453
be more valuable than ever, even if what you're doing is managing people who use
409
00:26:13.482 --> 00:26:13.842
tools.
410
00:26:13.902 --> 00:26:14.682
Yeah, I agree.
411
00:26:14.742 --> 00:26:17.997
So I think that's going to create a path for people who do not have to be managers
412
00:26:18.013 --> 00:26:20.802
but be contributors that are scaled.
413
00:26:21.462 --> 00:26:25.302
And that kind of org bifurcation is already starting to happen.
414
00:26:25.393 --> 00:26:29.533
Yeah. No, I love that. Who's managing all these agents?
415
00:26:31.122 --> 00:26:35.563
Interesting question, right? I start by saying that I don't think of agents or
416
00:26:35.622 --> 00:26:37.962
copilots as being in the org chart.
417
00:26:39.582 --> 00:26:45.567
But I do think of the question of who is managing it as an interesting one from
418
00:26:45.643 --> 00:26:50.127
a within-the-company-who-does-this perspective because my view is it should
419
00:26:50.172 --> 00:26:53.203
actually be the new form of HR that does this.
420
00:26:53.322 --> 00:26:54.582
I agree. Yeah.
421
00:26:54.612 --> 00:26:59.757
Today, agents and technologies are brought in within lines of business, like a
422
00:26:59.772 --> 00:27:03.237
sales agent that does X or a marketing agent that does Y, and then the office of
423
00:27:03.237 --> 00:27:06.942
the CIO is typically managing internal technology deployments of pilots.
424
00:27:07.572 --> 00:27:10.783
We are starting to see that almost all companies in the last year have shifted from
425
00:27:10.813 --> 00:27:13.332
internal pilots to buying external technologies.
426
00:27:13.842 --> 00:27:17.712
In 2023, it was like 95% of spend was on internal pilots.
427
00:27:18.222 --> 00:27:23.277
Now it's flipped towards almost all deployments are far more successful by going
428
00:27:23.292 --> 00:27:25.212
through technologies and vendors and so forth.
429
00:27:25.242 --> 00:27:30.328
What that also means is that we need to treat this as a
430
00:27:30.372 --> 00:27:36.303
central decision-making process for how are we redesigning work Because that's what
431
00:27:36.318 --> 00:27:40.503
agents are doing, and that should be an HR question, not a tool question or
432
00:27:41.718 --> 00:27:43.428
just a single-line of business question.
433
00:27:43.458 --> 00:27:47.238
Particularly because these things interact across the company, not just within the
434
00:27:47.268 --> 00:27:53.118
function. Now, a lot of companies are starting to create these hybrid HRIT, HRAI
435
00:27:53.928 --> 00:27:56.748
pockets, like the CHRO of ServiceNow, Jackie-
436
00:27:56.899 --> 00:27:57.558
Right
437
00:27:57.573 --> 00:27:58.399
... and HR and AI.
438
00:27:58.698 --> 00:27:59.029
Yes.
439
00:27:59.358 --> 00:28:03.378
But I think that should be more the norm in every company.
440
00:28:04.339 --> 00:28:06.378
But it's obviously tricky to suddenly do that overnight.
441
00:28:06.438 --> 00:28:11.133
So I think that sort of hybrid of tech and HR teams tackling this problem is
442
00:28:11.478 --> 00:28:12.498
probably the best way.
443
00:28:12.618 --> 00:28:13.938
Yeah. Listen, I could speak to you forever.
444
00:28:13.998 --> 00:28:17.073
Is there anything that we didn't discuss that we should have?
445
00:28:18.139 --> 00:28:18.483
I think there's
446
00:28:19.758 --> 00:28:23.089
important ethical questions. I'm currently co-writing a book.
447
00:28:23.358 --> 00:28:24.183
I've been
448
00:28:25.488 --> 00:28:28.998
close to finishing it for a while, but the book is called "Work, People, Robots,"
449
00:28:29.058 --> 00:28:31.938
which has made it a hard title to ever feel like you've finished.
450
00:28:32.718 --> 00:28:34.218
Yeah, because it's continuous. Yeah.
451
00:28:34.248 --> 00:28:39.798
But one of the reasons I've been co-writing this is because the ethical
452
00:28:39.858 --> 00:28:42.889
dilemma of are we facing mass unemployment
453
00:28:44.208 --> 00:28:49.818
is a real and important one, especially in the next five to 10 years, and the role
454
00:28:49.908 --> 00:28:55.518
that HR leaders and teams can play in stewarding that
455
00:28:55.728 --> 00:28:59.568
thoughtfully rather than reacting to business job changes is also very important.
456
00:29:00.438 --> 00:29:03.228
Some of the things that we've been doing with our clients over the last couple of
457
00:29:03.258 --> 00:29:08.568
years is helping provide insights of how do you retrain people rather than just
458
00:29:08.628 --> 00:29:12.408
automating a function. Just because something can be automated doesn't mean it
459
00:29:12.438 --> 00:29:15.738
should be, but more importantly, it doesn't mean you can't give people the time to
460
00:29:15.768 --> 00:29:19.098
retrain and think about where they go and be intentional with these things.
461
00:29:19.158 --> 00:29:22.907
And it tends to be good for the business and certainly good for the employees.
462
00:29:22.938 --> 00:29:27.589
And it's a societal thing that I think we should probably have more space and
463
00:29:27.618 --> 00:29:32.343
narrative for HR leaders to speak to in terms of what things can you do within your
464
00:29:32.389 --> 00:29:37.248
company that help shape a world where we do minimize how many people are scared
465
00:29:37.279 --> 00:29:38.958
about this stuff or impacted by this stuff.
466
00:29:39.693 --> 00:29:43.654
If you get into a self-driving car, you can't help but think, "Well, what's going
467
00:29:43.654 --> 00:29:47.313
to happen to every taxi driver?" But that's come for the office, it's come for
468
00:29:47.358 --> 00:29:52.788
every job, and it doesn't mean that we won't all have important things to do.
469
00:29:52.878 --> 00:29:56.133
I'm actually an optimist and think that 10, 20 years from now, a lot of work is
470
00:29:56.149 --> 00:29:57.603
going to be more rewarding and interesting.
471
00:29:58.128 --> 00:30:01.203
But I'm not an optimist about the next few years on that front, and certainly not
472
00:30:01.248 --> 00:30:07.218
for all roles. And I don't think there's any group better placed than HR leaders to
473
00:30:07.248 --> 00:30:10.683
help drive that narrative and make sure that the world is being thoughtful about
474
00:30:10.698 --> 00:30:10.878
this.
475
00:30:11.448 --> 00:30:15.258
Yeah. I think that's a very important point, and it's going to be interesting to
476
00:30:15.288 --> 00:30:19.203
see in the next five years how that plays out, and I don't think anyone really
477
00:30:19.279 --> 00:30:23.538
truly knows what's it look like. But when can we expect the book?
478
00:30:23.958 --> 00:30:27.468
I wouldn't dare to commit to timing.
479
00:30:27.529 --> 00:30:29.598
On a live, on a podcast.
480
00:30:30.919 --> 00:30:33.048
I'm writing it without AI, ironically.
481
00:30:33.258 --> 00:30:35.868
That's it. I was thinking that as you said it.
482
00:30:35.899 --> 00:30:37.714
I was like, I wonder if you're writing it with AI, but no.
483
00:30:37.714 --> 00:30:41.839
No, it would've been a much faster effort if I did, but it still felt wrong for
484
00:30:41.899 --> 00:30:42.873
this particular effort.
485
00:30:43.339 --> 00:30:47.448
No, it would be, especially given the context, and I think you definitely need the
486
00:30:47.478 --> 00:30:48.348
critical thinking and
487
00:30:50.029 --> 00:30:53.238
it... You know what, I was saying this to the team the other day, we just wrote our
488
00:30:53.298 --> 00:30:56.658
web copy for our brand-new website we're launching, and I was like, "Imagine how
489
00:30:56.748 --> 00:30:59.628
hard this would've been if we just had a blank piece of paper like we used to do
490
00:30:59.688 --> 00:31:03.498
this." It would take so long, but it does remove...
491
00:31:03.919 --> 00:31:09.589
We went back over the copy without AI and just as a team, and the next day I woke
492
00:31:09.708 --> 00:31:11.913
up, I was like, "Oh, wow, this is really not
493
00:31:13.968 --> 00:31:17.582
what... Last night it looked great, but I woke up with some fresh eyes and read it
494
00:31:17.628 --> 00:31:20.073
again," and I was like, "This is
495
00:31:21.288 --> 00:31:21.738
not good."
496
00:31:22.639 --> 00:31:26.358
But even though I'm writing it without AI, I think that's something we all now do,
497
00:31:26.568 --> 00:31:30.123
even as consumers, but you forget how much of an impact it makes, which is
498
00:31:30.258 --> 00:31:35.118
research, like the speed with which you can find relevant things to research.
499
00:31:35.958 --> 00:31:36.797
Yeah. Obviously, yeah.
500
00:31:36.842 --> 00:31:38.342
Yeah, like finding additional information.
501
00:31:40.188 --> 00:31:42.768
We've become used to it, like you're trying to book travel suddenly it's much
502
00:31:42.798 --> 00:31:46.608
easier to find information. But I think in work, it's everyone's a researcher now
503
00:31:46.639 --> 00:31:49.413
no matter what you're doing, because it's become so much more time efficient to do
504
00:31:49.458 --> 00:31:50.029
it.
505
00:31:50.118 --> 00:31:54.348
Yeah. Well, listen, before I let you go, where can people reach out if they want to
506
00:31:54.408 --> 00:31:58.188
say hi, connect, and where can they learn more about Beamer?
507
00:31:58.818 --> 00:32:02.298
Well, firstly, reach out directly on LinkedIn.
508
00:32:02.598 --> 00:32:07.698
I'm happy for anybody who wants to connect to reach out directly.
509
00:32:08.238 --> 00:32:13.518
And also happy to share more about the book and what I'm writing about.
510
00:32:13.548 --> 00:32:16.863
But I very much welcome any questions, discussion.
511
00:32:16.908 --> 00:32:18.378
And Chris, thanks so much for hosting.
512
00:32:18.468 --> 00:32:20.718
Amazing. Well, now you've put it out into the universe, everyone's going to be
513
00:32:20.779 --> 00:32:24.348
like, "When are you publishing?" So there's no pressure.
514
00:32:24.393 --> 00:32:27.258
It's only going to go out to about half a million HR executives.
515
00:32:27.948 --> 00:32:30.303
They're now going to be waiting on your book as well.
516
00:32:30.318 --> 00:32:34.308
But I appreciate you coming on, and congratulations on the journey so far, and I
517
00:32:34.368 --> 00:32:36.649
wish you and the team all the best until next week. Thanks a lot.
518
00:32:37.128 --> 00:32:37.563
Great. Thanks a lot.
Sultan Saidov, Former CEO & Board Member of Beamery.