TRASCRIZIONE 🔗
WEBVTT
1
00:00:00.000 --> 00:00:03.680
Hey and welcome to another episode of the Human Luke Podcast.
2
00:00:03.680 --> 00:00:08.480
I just had the pleasure of sitting down and talking to Paul from Uki.
3
00:00:08.480 --> 00:00:13.920
Paul runs a, I would say an AI first, bookkeeping business now.
4
00:00:13.920 --> 00:00:20.560
The conversation, it was so rich around topics of how they as a bookkeeping organisation
5
00:00:20.560 --> 00:00:27.040
are enabling their customers and how he's enabling his employees to look at new ways of working
6
00:00:27.040 --> 00:00:32.960
that have been, I would say, supercharged through the use of AI in, in their organisation.
7
00:00:32.960 --> 00:00:39.200
We cover a load of topics around customer value as well as people's, people's time
8
00:00:39.200 --> 00:00:43.200
and how to make the most of, make the most of everyone's time within the organisation.
9
00:00:43.200 --> 00:00:46.240
So tune into this episode to hear more around that conversation.
10
00:00:51.280 --> 00:00:54.560
The Human Luke Podcast, powered by Ethereum.
11
00:00:55.520 --> 00:01:00.720
Mate, we need to start off this conversation with talking about your quiz.
12
00:01:00.720 --> 00:01:01.520
OK.
13
00:01:01.520 --> 00:01:11.680
So I came along to an amazing Uki quiz here in Bournemouth and you killed us.
14
00:01:11.680 --> 00:01:19.280
You killed my team with one three letter word, which was in the wrong place,
15
00:01:19.520 --> 00:01:20.800
which was fast and furious.
16
00:01:20.800 --> 00:01:21.760
Yeah, that wasn't the answer.
17
00:01:21.760 --> 00:01:23.440
And it wasn't the answer.
18
00:01:24.640 --> 00:01:28.160
We lost 20 points for the word the.
19
00:01:28.160 --> 00:01:29.120
Yeah, brilliant.
20
00:01:29.120 --> 00:01:30.800
Well, we created a wipe out round for that reason.
21
00:01:30.800 --> 00:01:34.720
So it was a very specific question, I was very specific about the result.
22
00:01:34.720 --> 00:01:38.240
And yeah, it did cause a bit controversy, but it's all good fun.
23
00:01:38.240 --> 00:01:41.120
So, and you just told me that, you know, you want to sign up for the next one?
24
00:01:41.120 --> 00:01:44.320
Well, yeah, I'm going to, I'm going to, you're not competitive, right?
25
00:01:44.320 --> 00:01:46.720
Not competitive at all, but I did go home.
26
00:01:46.720 --> 00:01:50.000
I did go home and I sat down with my wife and I was like, she's like, how did it go?
27
00:01:50.000 --> 00:01:53.200
And I said, um, yeah, I don't really want to talk about it.
28
00:01:53.200 --> 00:01:56.320
You shouldn't see in the WhatsApp sign cards, it's brilliant.
29
00:01:56.320 --> 00:01:58.320
So that's, you know, people talk about it, right?
30
00:01:58.320 --> 00:01:59.120
That's a good point.
31
00:01:59.120 --> 00:02:00.080
Nicknacks.
32
00:02:00.080 --> 00:02:00.880
Yeah.
33
00:02:00.880 --> 00:02:04.480
The Nicknacks saved the day for you because I wouldn't be coming again if they weren't Nicknacks.
34
00:02:04.480 --> 00:02:06.240
I thought I'd go, yeah, that's another thing we did.
35
00:02:06.240 --> 00:02:12.240
Like, it was just this idea of like just trying to get clients and customers and friends
36
00:02:12.240 --> 00:02:16.800
of Uki together to kind of just just have an evening of a bit of fun, really, a bit of a
37
00:02:16.800 --> 00:02:17.800
pop quiz.
38
00:02:17.800 --> 00:02:22.400
Pop quiz is a great, and a format's great, and it's just getting people having conversations
39
00:02:22.400 --> 00:02:25.040
as well that I've never met before, so all the networking elements.
40
00:02:25.040 --> 00:02:29.520
So, yeah, throwing in stuff like, yeah, pizzas and crisps and all that stuff, it's like,
41
00:02:29.520 --> 00:02:30.520
all good fun.
42
00:02:30.520 --> 00:02:31.520
Yeah, it was good.
43
00:02:31.520 --> 00:02:32.520
Um, Uki.
44
00:02:32.520 --> 00:02:33.520
Yep.
45
00:02:33.520 --> 00:02:34.520
What the bloody hell does that mean?
46
00:02:34.520 --> 00:02:35.520
Yeah.
47
00:02:35.520 --> 00:02:41.160
Okay, so, um, Uki is from the only word in the English language that's got three consecutive
48
00:02:41.160 --> 00:02:44.760
double letters, double A, double K, double E, which is bookkeeping.
49
00:02:44.760 --> 00:02:49.520
And that's kind of the main core of the work that we do for businesses that we work with.
50
00:02:49.520 --> 00:02:50.520
Okay.
51
00:02:50.520 --> 00:02:53.760
That couple of management accounting, but ultimately bookkeeping is our core, and when
52
00:02:53.760 --> 00:02:57.080
we were trying to find a cool, funky name, and I just started writing them all down and
53
00:02:57.080 --> 00:03:00.080
wrote the word bookkeeping down and just circled it, and then wrote another hundred things
54
00:03:00.080 --> 00:03:01.080
and just went straight back to that.
55
00:03:01.080 --> 00:03:06.480
Yeah, I mean, once you see something like that, now I know I'll never forget, although
56
00:03:06.480 --> 00:03:09.960
I, Jade, may have told me, and I have forgotten, I'll never forget it again.
57
00:03:09.960 --> 00:03:10.960
Yeah.
58
00:03:10.960 --> 00:03:13.640
Um, I guess like once you see something like something like that, it stands out, right?
59
00:03:13.640 --> 00:03:14.640
And it's just something different.
60
00:03:14.640 --> 00:03:15.640
And it's a talking point.
61
00:03:15.640 --> 00:03:19.040
And, you know, we're at the center of the bookkeeping and all that stuff, you know, with
62
00:03:19.040 --> 00:03:22.040
the only naming bookkeeping, all that, all that stuff, right?
63
00:03:22.040 --> 00:03:23.040
Oh, okay.
64
00:03:23.040 --> 00:03:26.040
All of the Uki plugs already on the show, done.
65
00:03:26.040 --> 00:03:27.040
Done.
66
00:03:27.040 --> 00:03:28.040
Okay.
67
00:03:28.040 --> 00:03:29.040
Cool.
68
00:03:29.040 --> 00:03:34.880
So you, um, you have like a great business that, um, as you said, bookkeeping management
69
00:03:34.880 --> 00:03:36.880
accounts, etc.
70
00:03:37.880 -->
Load full transcript
00:03:38.880
AI.
71
00:03:38.880 --> 00:03:46.360
So the podcast, human in the loop, okay, AI, AI, um, with ethereo, we've got two, uh,
72
00:03:46.360 --> 00:03:51.720
accountancy practices, customers with ethereo, um, and my first conversation with them was
73
00:03:51.720 --> 00:03:58.920
like, you guys screwed, like with AI, um, and the, the, the, the, the, the first pilot that
74
00:03:58.920 --> 00:04:03.840
we've got with ethereo is a top 100 accounts you've, um, and I led with that thinking then
75
00:04:03.840 --> 00:04:06.320
I'm not going to use ethereo after I lead with that.
76
00:04:06.320 --> 00:04:12.600
And they were like, yes, and no, yes, and no, yes, all of the bottom of the pyramid stuff
77
00:04:12.600 --> 00:04:13.960
that we used to be paid to do.
78
00:04:13.960 --> 00:04:14.960
Yeah.
79
00:04:14.960 --> 00:04:15.960
No.
80
00:04:15.960 --> 00:04:19.360
But actually, this frees us up to be able to add lots of value elsewhere.
81
00:04:19.360 --> 00:04:20.360
Agreed.
82
00:04:20.360 --> 00:04:25.920
So tell me, like tell me where you sit on that side as like a business owner within your,
83
00:04:25.920 --> 00:04:26.920
your industry.
84
00:04:26.920 --> 00:04:32.200
It is just, I think dependent here, you are on dependent industry or in AI is, it's
85
00:04:32.200 --> 00:04:34.760
just a huge opportunity.
86
00:04:34.760 --> 00:04:37.640
You've got to understand its limitations.
87
00:04:37.640 --> 00:04:41.240
You've got to understand how it's going to be delivered and you've got to understand
88
00:04:41.240 --> 00:04:42.240
how it's going to be used.
89
00:04:42.240 --> 00:04:46.840
And I think that's going to be a really big education piece for not only just a business
90
00:04:46.840 --> 00:04:51.720
owner, but the employees to, and even the customers, because they're going to maybe think,
91
00:04:51.720 --> 00:04:56.360
oh, well, AI can do it all now, do it myself, still a bad idea.
92
00:04:56.360 --> 00:04:59.520
We've seen, you know, as I think in this emergence of AI, we've seen where it can
93
00:04:59.520 --> 00:05:00.520
make mistakes.
94
00:05:00.520 --> 00:05:05.000
We've seen how, you know, we need a human in the loop, and it's really important that
95
00:05:05.000 --> 00:05:06.000
that's there.
96
00:05:06.000 --> 00:05:07.000
We've still got to have human connections.
97
00:05:07.000 --> 00:05:09.640
We've still got to have, you know, it's how we base our decisions.
98
00:05:09.640 --> 00:05:14.200
We're not necessarily going to be as based on decisions on the AI, however, the opportunities
99
00:05:14.200 --> 00:05:16.680
are incredible.
100
00:05:16.680 --> 00:05:25.160
We are very lucky as EQ that I'd say we're in the top 1% of tech in our industry for what
101
00:05:25.160 --> 00:05:26.880
we do and how we use it.
102
00:05:26.880 --> 00:05:32.080
And that's only because we have the top people in the industry with regards to the softwares
103
00:05:32.080 --> 00:05:37.840
that we use, people like Sage Zero, Dex, for example, we get to do a lot of stuff behind
104
00:05:37.840 --> 00:05:42.560
the scenes with them, alpha testing, products prior to them being put out of the beta testing.
105
00:05:42.560 --> 00:05:47.560
So we actually have quite a good level of say into what happens, and we feedback quite
106
00:05:47.560 --> 00:05:51.960
a lot to them, but we also get like first dibs and first go with it.
107
00:05:51.960 --> 00:05:54.200
Some of the stuff is incredible.
108
00:05:54.280 --> 00:06:02.280
One example is I'm going to give you the detail, but it just reads the information and then
109
00:06:02.280 --> 00:06:06.800
bases its decision on something we've created as a parameter through an agent.
110
00:06:06.800 --> 00:06:10.200
And that's all bit into one of these pieces of software, and it's saved one of my team
111
00:06:10.200 --> 00:06:12.720
members 9 hours a week, wow.
112
00:06:12.720 --> 00:06:14.080
And that's just like brilliant.
113
00:06:14.080 --> 00:06:17.520
So now with that in mind, you've got lots of things.
114
00:06:17.520 --> 00:06:22.680
So my employees are a bit like, oh, God, well, if I do that, how's that going to look
115
00:06:22.680 --> 00:06:23.680
for me?
116
00:06:23.680 --> 00:06:30.360
And I'm like, if you can leverage your new skill set in building this out to make your
117
00:06:30.360 --> 00:06:34.920
cell five times more efficient, I'll pay you at least double your salary.
118
00:06:34.920 --> 00:06:38.120
You're a purchase ledger clerk on 70 grand then, right?
119
00:06:38.120 --> 00:06:43.440
That's another level, but you're not a purchase ledger clerk, you're an AI agent teacher.
120
00:06:43.440 --> 00:06:48.720
Do you need to teach those agents your skill set, and that's how you leverage that going
121
00:06:48.720 --> 00:06:49.720
forward?
122
00:06:49.720 --> 00:06:56.320
So one very small piece of this puzzle, and I think the opportunities here are incredible.
123
00:06:56.320 --> 00:07:01.320
It's just how you want to use it, and put that into your business.
124
00:07:01.320 --> 00:07:07.680
So yes, there are some places I think where I wouldn't say jobs will go because people
125
00:07:07.680 --> 00:07:12.520
still have skill sets, I think jobs will change a lot, and people just need to have that
126
00:07:12.520 --> 00:07:15.920
resilience to change an adapt with the new tools that we have.
127
00:07:15.920 --> 00:07:20.200
We don't go drilling holes with hand drills anymore, right?
128
00:07:20.200 --> 00:07:24.840
So if you charged per hole you drilled, now we've got an electric drill, okay, in a minute
129
00:07:24.840 --> 00:07:28.200
it might be a robot drilling hole, but still the same thing, we're still involved, we're
130
00:07:28.200 --> 00:07:33.400
still before, we still have those things in place, so yeah, it's changing heat.
131
00:07:33.400 --> 00:07:35.320
There's a lot, I think we can unpick there.
132
00:07:35.320 --> 00:07:42.600
I want to come back to the
purchase ledger, almost being like the coach of the agents,
133
00:07:42.600 --> 00:07:48.360
let's come back to that one, the first bit you said was about, I guess like the risks
134
00:07:48.360 --> 00:07:57.000
or the mistakes, the hallucinations, and that's a commercial risk, so removing the human
135
00:07:57.000 --> 00:08:02.840
from the loop completely and being like great autonomy is fantastic, and I'm going to
136
00:08:02.840 --> 00:08:06.760
share a story in a minute, it's fantastic until it completely screws up and you look
137
00:08:06.760 --> 00:08:12.480
like a dick, which I would argue is as bad in front of your employees as your
138
00:08:12.480 --> 00:08:20.840
customers, like actually if your employees lose faith in your AI strategy, your customers
139
00:08:20.840 --> 00:08:28.600
are going to lose faith, but also they end up double working, I think, I have, so there's
140
00:08:28.600 --> 00:08:38.920
an ex-ITO that I know really well, who had a very successful career with a, we'll call
141
00:08:38.920 --> 00:08:45.360
it an online gift company, let's protect him a bit, an online gift company, all right,
142
00:08:45.360 --> 00:08:51.800
he's ex did now, he's working on little projects, we're mates, got him on WhatsApp, I have
143
00:08:51.800 --> 00:08:56.440
stress tested some ideas for Ethereum on him and we've had a banter on agentic coding
144
00:08:56.440 --> 00:09:03.720
and vibe coding and stuff like that, I woke up two Sundays ago to a WhatsApp from him,
145
00:09:03.720 --> 00:09:08.400
so I woke up in the morning, went downstairs, coffee, WhatsApp, okay, oh great, I read the
146
00:09:08.400 --> 00:09:14.280
WhatsApp, I was like, that doesn't seem like him, but also we stopped our last conversation
147
00:09:14.280 --> 00:09:20.760
like a week ago and this message isn't relevant, but I was a bit, bit like I was waking up,
148
00:09:20.760 --> 00:09:26.560
I had one coffee, I need two before I'm like properly going and so I replied, I replied
149
00:09:26.560 --> 00:09:33.240
back it, there we go, I'm giving you a clue, he replied back straight away and the reply
150
00:09:33.240 --> 00:09:38.080
was again like very persistent about the last question, right, but it was wrong, it was
151
00:09:38.080 --> 00:09:44.680
like wrong, yeah, and I just woke up in that moment and I was like, he's trying to use
152
00:09:44.680 --> 00:09:50.000
a bot, like this is, this is, this is an end hack or something, okay, so I just replied
153
00:09:50.000 --> 00:09:54.200
back in emoji, the robot face emoji, that was it, I was like, I'm done, yeah, so I checked
154
00:09:54.200 --> 00:09:58.760
out, I was like, this is not insert name, I'm done, yeah, all right, left it, went to
155
00:09:58.760 --> 00:10:04.880
the mess about the kids and watched whatever, anyway, later that day I had a message from
156
00:10:04.880 --> 00:10:12.200
him, like, hey Ben, I'm so sorry, I had, I'm building a little experiment, a little side
157
00:10:12.200 --> 00:10:18.320
hustle, a little experiment and I had, I think it was a malt bot, I was using malt bot and
158
00:10:18.320 --> 00:10:25.280
I had integrated it in my WhatsApp because I was using WhatsApp to talk to it, I mistakenly
159
00:10:25.280 --> 00:10:31.920
set up the settings to like communicate all or something and he had over a thousand
160
00:10:31.920 --> 00:10:38.360
people in his WhatsApp's and it was communicating with all of them, yeah, so like the personal
161
00:10:38.360 --> 00:10:44.360
brand damage that that could have done, like could be horrific, I've heard of similar things
162
00:10:44.360 --> 00:10:50.520
with some engineers testing as well, with say Google ads or something and the AI has gone
163
00:10:50.600 --> 00:10:57.480
and sent 11,000 requests to Google, cost them 15 grand, wow, in like a minute, like, excuse
164
00:10:57.480 --> 00:11:01.560
me Google, can we have our money, but didn't, no, of course they didn't, right, it's a transaction,
165
00:11:01.560 --> 00:11:05.720
yeah, so like, I think they're, I think they're like, did done something, we're some credits,
166
00:11:05.720 --> 00:11:11.880
yeah, but like, you know, easily done right and I think, you know, people talk a lot about
167
00:11:11.880 --> 00:11:17.080
guardrails and ring fencing and all that stuff and I think that's really important but and that
168
00:11:17.080 --> 00:11:20.360
needs to happen and a lot of people don't know about that, let so many people haven't got like
169
00:11:20.360 --> 00:11:23.480
any AI policies in their business and stuff like that, those teams can go off and use whatever
170
00:11:23.480 --> 00:11:28.920
they want and upload whatever they want and it's in the learning space, but ultimately I think
171
00:11:30.120 --> 00:11:34.600
and the thing with accounting is we always have to have a reconciliation and a check, so there's
172
00:11:34.600 --> 00:11:39.640
always going to be that manual thing because that manual person has to go and deliver that
173
00:11:39.640 --> 00:11:44.280
report to that other person, yeah, so you know, we've always got those checkpoints and I think
174
00:11:44.280 --> 00:11:49.880
that's the point where we can't really let that go, I think, with any sort of AI, we still need to
175
00:11:49.880 --> 00:11:58.200
have, just, funnily enough, going back to the quiz, I was like, I said it was the round where we
176
00:11:58.200 --> 00:12:05.800
gave you three potential answers, yes, so I put in, one was a chat GPT answer, they were all,
177
00:12:05.800 --> 00:12:09.800
it was all, that little round was done by chat GPT, so I said, give me, give me an example, I said
178
00:12:10.680 --> 00:12:16.280
give them an example of the topic question and it said, it's two treats and a lie wasn't it,
179
00:12:16.280 --> 00:12:22.120
so two lies in a tree, yeah, and it said something like the Eiffel Tower is in Paris,
180
00:12:23.080 --> 00:12:27.160
the Beatles were formed in
Liverpool, I went, the Beatles were formed in Liverpool, like so,
181
00:12:27.160 --> 00:12:31.480
I went, you've got that wrong, it went, oh yeah, I have, sorry, yeah, you're like, I'm doing a
182
00:12:31.480 --> 00:12:36.840
quiz, like I cannot get it wrong, but high stakes, high commercial stakes, but when you just
183
00:12:36.840 --> 00:12:39.880
ask it, you've got it wrong, can you check, and it goes, oh yeah, sorry, what have you done it wrong
184
00:12:39.880 --> 00:12:45.160
in the first place? Yeah, so I think, you know, we have got to be very, very careful with that,
185
00:12:45.160 --> 00:12:48.760
and again, what we're seeing is from a lot of the software manufacturers, they're being,
186
00:12:48.760 --> 00:12:54.520
they're building in the testing phase, which is great, so in this example I gave you about
187
00:12:54.520 --> 00:13:00.120
purchase ledger, first of all what it does is it looks at how you work with it and it makes
188
00:13:00.120 --> 00:13:05.640
suggestions, so you can go with all suggestions, you either confirm the suggestion through your
189
00:13:05.640 --> 00:13:09.480
process, and you go, yep, I'll use that, yep, I'll use that, then you can look back and say,
190
00:13:09.480 --> 00:13:14.200
how many times do we confirm, how many times do we dismiss? If you've got 100% confirmed, turn on,
191
00:13:14.200 --> 00:13:18.920
turn on, you've never got to do it again, right, and that's how it should work, I think it's not
192
00:13:18.920 --> 00:13:24.200
just a case of, because nobody's in expert in prompt writing, they're still I'm in a knowing
193
00:13:24.200 --> 00:13:30.360
how to price for it, like tokens, so token, I think some of the other day, they reckon that's
194
00:13:31.080 --> 00:13:34.680
an amount of characters or something that go into your prompts, then people be shortening their
195
00:13:34.680 --> 00:13:39.160
prompts to try and save money, but the prompt isn't good enough, and then it's definitely
196
00:13:39.160 --> 00:13:44.680
different. And is there a different charge for a voice token versus a written token, just endless,
197
00:13:44.680 --> 00:13:51.080
right, so my, I prefer the charge on my output, that works better for me, because there's lots of
198
00:13:51.080 --> 00:13:56.360
tools that we're going to find stuff to do stuff, but unless it's giving you an output, it's the way
199
00:13:56.360 --> 00:13:59.880
we were, you've got some tangible to charge for, that was why we started the business in the first
200
00:13:59.880 --> 00:14:06.760
base, but it was more about giving something tangible or touching something to be able to charge
201
00:14:06.760 --> 00:14:10.200
for it, and that's I think how it should work, but interesting to see what they come up with.
202
00:14:10.200 --> 00:14:17.720
Yeah, I think so, so my, my CTO, so Dan, in building a theory from like architecture,
203
00:14:17.720 --> 00:14:24.360
through to back in front, and et cetera, like he's fully, this
has been a lived experiment
204
00:14:24.440 --> 00:14:31.400
on, on agentic coding, all right, so theory is end to end, one person, CTO header product,
205
00:14:31.400 --> 00:14:39.560
all develop a swarm of agents, et cetera, and we did an analysis internally, we Dan did an analysis
206
00:14:39.560 --> 00:14:47.240
internally on over the five months, all right, so we built like an enterprise scale app in five
207
00:14:47.240 --> 00:14:51.640
months, yeah, with one person, like to start with that's insane, right, and it's out there,
208
00:14:51.640 --> 00:14:56.680
but 200 managers using it now, it has been stress tested, et cetera, and like it's real,
209
00:14:56.680 --> 00:15:01.640
yeah, and there's a lot of nuances around the behavioral science, et cetera, behind it,
210
00:15:01.640 --> 00:15:07.560
but when we looked at like how much code was committed, like it's massive, like the code it's
211
00:15:07.560 --> 00:15:13.800
been able to, like it's productivity is great, but when you look at what was QA'd and rewritten
212
00:15:13.800 --> 00:15:20.120
and edited, that's also massive, right, so you look at the productivity gains, but also unless you
213
00:15:20.120 --> 00:15:27.080
are creating the environment and the guard rails around it, I mean that's basically, you know,
214
00:15:27.080 --> 00:15:32.440
that's basically like a Claude code, for example, like I'm going, yeah, can I screw that up,
215
00:15:32.440 --> 00:15:37.240
let me fix it, et cetera, but what's really interesting is you can build in self-healing
216
00:15:37.960 --> 00:15:46.200
as well, which is very cool, but I think this comes back to for a theory, without Dan and his
217
00:15:46.200 --> 00:15:52.760
immense like tenure experience and knowledge base to be in that loop, or maybe even overseeing that
218
00:15:52.760 --> 00:15:58.600
loop, me on my own with Replet would have got ourselves into all sorts of trouble, yeah, and so I
219
00:15:58.600 --> 00:16:03.080
think that that that is really, really important, it is important and again it's like what's the
220
00:16:03.080 --> 00:16:09.960
I always look at as well, the like the scalability of it and like what how can it get those sorts of
221
00:16:10.040 --> 00:16:15.560
things if, you know, I'm trying to now train my team to take ownership and control of their own
222
00:16:15.560 --> 00:16:20.920
pieces, as a business we've got the tools, but we've got to train the people to you to tools,
223
00:16:21.560 --> 00:16:26.360
and you know, you've got one guy there doing all the stuff, and I'm like, okay, if I've got
224
00:16:26.360 --> 00:16:29.560
each individual person doing it, and they share it, got to share their experiences, they're sharing
225
00:16:29.560 --> 00:16:33.240
the knowledge between them, it's some absolute one of them, I'm good, like that's, you know, it's got
226
00:16:33.800 --> 00:16:41.320
we're covered, right, but that's also then a problem if, you know, it all starts ending up being
227
00:16:41.320 --> 00:16:46.280
on one person's shoulders and one person's lived experience and all the rest of it, that's where for
228
00:16:46.280 --> 00:16:51.960
me, I'm also like, how are we going to manage that, because that's something that we're looking at
229
00:16:51.960 --> 00:16:57.080
sustainability legacy, what we want to call it down the line, it's like how what's going to happen
230
00:16:57.080 --> 00:17:01.160
down the line, I mean, we're going to get to Supra, I'm going to get to, you know, it's a different
231
00:17:01.160 --> 00:17:03.960
level again, right, the intelligence is just going to get more and more.
232
00:17:05.480 --> 00:17:08.600
Did you see the thing, there's a sort of thing that the day where this person was having a
233
00:17:08.600 --> 00:17:12.200
conversation with somebody, a person to person, but the person was sort of forming like AI,
234
00:17:12.200 --> 00:17:18.680
I think I have seen it, yeah, because it's like, it's got a specific, it has got a specific
235
00:17:18.680 --> 00:17:22.840
way of talking to you, so it has almost got its own personality, yeah, and like you noted,
236
00:17:22.840 --> 00:17:26.680
it's like, you recognise it, because you recognise you went to your friend, and you recognise
237
00:17:26.680 --> 00:17:32.200
its language and what it does, when that can start to really replicate language, how we talk
238
00:17:32.200 --> 00:17:37.560
to each other now, and putting arms and, you know, especially on a voice version, that's when it's
239
00:17:37.560 --> 00:17:45.000
going to get scary. I'm not scaring a bad way, I think scary for my, maybe our, our, just, you know,
240
00:17:45.000 --> 00:17:49.400
how old we are and what we're used to, but really it gets scary for two seconds and then we go,
241
00:17:49.400 --> 00:17:52.920
right, what's the opportunity here, how can we make this thing? And that's it, I think, I think
242
00:17:52.920 --> 00:17:59.400
there are, people will get left behind if they don't, like, super adapt, and there is this,
243
00:17:59.400 --> 00:18:03.160
there's this, like, there's the change curve, etc, that everyone goes through and people go
244
00:18:03.160 --> 00:18:08.440
for it faster, and I think the business is, the employees, the people in work and in life that are
245
00:18:08.440 --> 00:18:13.880
like, I'm up for experimenting with this and feeling uncomfortable, and those are the people that
246
00:18:13.880 --> 00:18:19.400
are going to do really, really well, and the other people, the other people that get left behind,
247
00:18:19.400 --> 00:18:24.520
I think you're almost going to get, you know, it almost be like like a light speed, like an old
248
00:18:24.520 --> 00:18:28.840
school like Star Trek, like, shoots ahead, and everyone's like, oh my goodness, I'm left in space,
249
00:18:28.840 --> 00:18:33.880
everyone else is over there, and we're here, an example of that, was I was speaking to an HR
250
00:18:33.880 -->
00:18:40.920
director of a 600 person law firm, like six, eight months ago, and, you know, Ethereum wasn't a
251
00:18:40.920 --> 00:18:44.040
thing then, we're building it, but I was like, asking the question anyway, I was like, so where are
252
00:18:44.200 --> 00:18:51.000
you at with how you use AI within the law firm, and their answer was Ben, we've just launched
253
00:18:51.000 --> 00:18:56.200
an LMS, like a learning management system, we're probably not ready for AI for like 80 months,
254
00:18:57.080 --> 00:19:03.240
and in my head I was like, you've probably spent, like, for them probably spent 30, 40K on this
255
00:19:04.040 --> 00:19:09.080
content system, and I was, and in my head I was like, I'm getting twitchy because it just makes
256
00:19:09.080 --> 00:19:14.040
you frustrated. Yeah, no, in my head I was like, I guarantee you that there are people in your
257
00:19:14.040 --> 00:19:18.920
business that have amazing little experiments and use cases of how they are using AI to make
258
00:19:18.920 --> 00:19:24.040
there, to be more productive, et cetera, but it's probably outside of the controls or the parameters
259
00:19:24.040 --> 00:19:31.000
or you want. Yeah. But also, you've just invested in something that I pretty much guarantee you
260
00:19:31.000 --> 00:19:38.200
in two years time will be a complete west time. I agree, and I'm a tinkerer, so I'm like, I would say
261
00:19:38.920 --> 00:19:44.120
in comparison to say where our team are now and the workflows and the processes we're using,
262
00:19:44.120 --> 00:19:47.080
it's probably what I was thinking about to, what they're doing now is what I was thinking about
263
00:19:47.080 --> 00:19:52.280
two years ago, what I'm thinking, I'm generally a year to 18 months ahead of where I visualize
264
00:19:52.280 --> 00:19:57.000
the business being, and I'm starting to slowly implement, like, let them catch up basically,
265
00:19:57.000 --> 00:20:03.880
because we can't go that fast. But if they're not thinking about it, like anything to do with
266
00:20:03.880 --> 00:20:09.720
legislative change, legal, anything, AI's got it nailed, like, it's such a simple thing to fix.
267
00:20:10.360 --> 00:20:18.120
So, the switch then is when we're dealing with customers, we've got to keep focus on valley
268
00:20:18.120 --> 00:20:24.040
deliverables, and make sure that they're getting value from it. Now, AI is going to be, you know,
269
00:20:24.040 --> 00:20:32.280
every counting software has now got their AI bot in it. And my thing with this is it needs to
270
00:20:32.280 --> 00:20:37.800
protect the industry and protect the professionalism of the accountants. We aren't accountants,
271
00:20:37.800 --> 00:20:41.560
we don't do year-end, we don't do self-assessment, we don't do any of that stuff, tax, anything
272
00:20:41.560 --> 00:20:45.640
like that, we sit as you're a lot management accountant, we're in your business, and we work with
273
00:20:45.640 --> 00:20:50.920
the accountant or the factual CFA or fractional D, whoever, to deliver them the information. So,
274
00:20:50.920 --> 00:20:59.800
like, when it gets to that point of moving forward into those other spaces, it's really important
275
00:21:00.520 --> 00:21:06.680
that value is in the numbers, because a lot of year-end just becomes like a filing thing that
276
00:21:06.680 --> 00:21:12.200
provided all this information is correct. I mean, my championing is, but keeping is the most
277
00:21:12.200 --> 00:21:17.080
important part of the account industry, because we get it wrong, management accountants gets it wrong,
278
00:21:17.080 --> 00:21:20.520
financial accountants gets it wrong. Everything down straight, we pay the wrong tax, and so on,
279
00:21:20.520 --> 00:21:26.200
right wrong VAT, everything, right, and we see it all the time. And I've been in industry,
280
00:21:26.200 --> 00:21:29.720
I was thinking about the car to say 34 years, right? You look great for it, thanks, man.
281
00:21:31.000 --> 00:21:37.240
VAT, but I think, you know, and we started a year-end of the back of just having a frustration,
282
00:21:37.240 --> 00:21:43.720
not being able to get what we wanted from the numbers, and we, you know, spoke to a lot of businesses,
283
00:21:43.720 --> 00:21:47.880
and I would say, I think the stats are something like, I can't wait, we came from a bank,
284
00:21:47.880 --> 00:21:51.560
I think it's more like 15% of businesses or business owners produce management accounts,
285
00:21:51.640 --> 00:21:56.280
right? I know that 15%, only 5% understand what they're looking at, right? Which you might
286
00:21:56.280 --> 00:22:00.600
think, if you're a numbers person within your business, you've got to comprehend that that doesn't
287
00:22:00.600 --> 00:22:07.000
happen, but believe me, I work for businesses from start-up, who've got no plan, from people that
288
00:22:07.000 --> 00:22:10.360
are amazing at what they do and manage to realise they can enter money from it and start cracking
289
00:22:10.360 --> 00:22:13.640
on, and then they've got a problem because they've got this wake of stuff behind them,
290
00:22:13.640 --> 00:22:19.320
to businesses that have been running over 20 years, we're over 100 staff, and still not really
291
00:22:19.400 --> 00:22:24.200
understanding what things are, and like, I could not believe it, but that's what we find,
292
00:22:24.200 --> 00:22:28.760
what we find out, and all we do is help present the numbers, help them understand what the KPIs
293
00:22:28.760 --> 00:22:32.920
are, the most important parts, but ultimately, that's where the value is, from my perspective,
294
00:22:32.920 --> 00:22:36.840
it's a business owner, I want to know that, and I know that what's going to affect my business,
295
00:22:37.640 --> 00:22:42.680
whereas the AI element can't always do that, okay, it can benchmark, it can look at some other
296
00:22:42.680 --> 00:22:47.480
stuff so it can help us, but everybody's slightly different, everybody's got slightly different goals,
297
00:22:47.560 --> 00:22:51.880
okay, if they can find a way to pump all of that in, unplug it all into the AI, okay,
298
00:22:51.880 --> 00:22:57.000
there, I might be able to make some, create some analysis around that, but ultimately, you have to
299
00:22:57.000 --> 00:23:00.680
have that human communication, what should we do, you feel like you want to go with it, I just
300
00:23:00.680 --> 00:23:03.800
feel a bit poorly today, so I'm not going to go ahead with it for the next few months, you're like,
301
00:23:03.800 --> 00:23:07.800
fine, okay, well, you know, you can help them around it, but all AI is going to do is
302
00:23:09.160 --> 00:23:14.360
give you more options to think about, but you're going to need to discuss, so I think, but when it comes
303
00:23:14.360 --> 00:23:20.680
to like filing tax and filing things or running around rules that are set in stone by somebody that
304
00:23:20.680 --> 00:23:25.800
can be picked up with language and read and understood and converted, that's where it gets done,
305
00:23:25.800 --> 00:23:30.120
and then we'll start some business to do a year end, do they, right, so that's the value in it,
306
00:23:30.120 --> 00:23:33.560
it's like it has to be done, I didn't want to do that, but I have to, but I don't want to know
307
00:23:33.560 --> 00:23:37.560
where I'm business is going, that's more important for me, and that's, and I think the AI can kind of
308
00:23:38.600 --> 00:23:43.720
help in all different areas, and that's where I see accountants with their knowledge helping
309
00:23:44.440 --> 00:23:49.400
that bit, but they can use AI to actually do a lot of their own stuff to it, so it's fun in the
310
00:23:49.400 --> 00:23:56.120
balance. Can I make you cringe a bit? Go for it. Okay, when I set up my first business,
311
00:23:56.920 --> 00:24:04.520
my ethos, because I love Richard Branson, and I heard, I love Richard Branson, I love
312
00:24:04.520 --> 00:24:11.960
distribution, like I'm a sales marketing like person, and obviously he is, and I remember here
313
00:24:11.960 --> 00:24:16.840
in a quote, which was, as long as you sell more than you spend, you've got a great business.
314
00:24:17.720 --> 00:24:21.640
Brilliant. And I was like, oh, I like that, that's the sound I can get on board with, and so,
315
00:24:21.640 --> 00:24:27.240
until my first business, literally, like I would, that was my ethos, like as long as I know,
316
00:24:27.800 --> 00:24:35.720
as long as I look at like invoices out and invoices paid, is more than what I think Bern rate is,
317
00:24:36.280 --> 00:24:41.800
this month we're golden, okay, and I operated like that for years. We're not serious solid.
318
00:24:42.440 --> 00:24:47.400
It is solid, but it's also, it's solid, but you've got to keep an eye on it. And then when you're,
319
00:24:47.400 --> 00:24:51.240
when you get some like product market fit interaction,
you grow, and you're like, great, well,
320
00:24:51.240 --> 00:24:56.840
as long as like we still do this top line number, I can just invest in grow, etc. And that's when
321
00:24:56.840 --> 00:25:02.360
you get sticky and you need systems, and actually, like now with a theory, like our financial model
322
00:25:02.920 --> 00:25:09.160
is like, is well thought through. We are like management accounts. My account in iE is like,
323
00:25:09.160 --> 00:25:15.400
he's awesome, but also he's a pain in the ass because he's awesome. And so you're right,
324
00:25:15.400 --> 00:25:19.720
like end of year, end of year is like, that's what, that's what's happened, but by the time it's done,
325
00:25:19.720 --> 00:25:23.720
it's done. You can't change anything. It's like a scorecard of how you played, without even
326
00:25:23.720 --> 00:25:27.800
like thinking about how you played, whereas like bookkeeping and management accounts. For me,
327
00:25:29.240 --> 00:25:34.280
if I can get as far upstream as I can in understanding like the behaviors and the actions
328
00:25:34.280 --> 00:25:40.040
financially in our business, you just hit the note in your business. Yeah, right. So the context
329
00:25:40.040 --> 00:25:44.520
is super relevant. And then in line with what? Well, in line with where we want to go. Correct.
330
00:25:44.520 --> 00:25:49.960
And the current market appetite, etc. So it's like, for now, now I'm becoming more,
331
00:25:50.920 --> 00:25:55.800
more finance literate, because I've got a bunch of people's money that I have a responsibility
332
00:25:55.800 --> 00:26:01.480
to previously. It was just bootstrapped. So I have a, I have this weird like stewardship value.
333
00:26:01.480 --> 00:26:07.880
100%. So that's really important. And so I completely agree with you. I think the human element,
334
00:26:07.880 --> 00:26:12.120
that we have conversations about like, this is where we're tracking burn rate, this is why,
335
00:26:12.120 --> 00:26:18.840
this is acceptable, etc. And where we're going to look to get to profitable, etc. So I buy into
336
00:26:18.840 --> 00:26:25.880
that fully. And I think you have to, you're going to have to keep the human context relevant
337
00:26:25.960 --> 00:26:33.400
experiential bit. We've got other humans that you're responsible to, right. And also,
338
00:26:33.400 --> 00:26:39.080
they're responsible to and the knock on effect. So I think having that information is great. But
339
00:26:39.640 --> 00:26:44.200
in this instance, you wouldn't have embarked on that without knowing that there's numbers to start
340
00:26:44.200 --> 00:26:49.480
with, whereas a lot of people do. And then they have to play catch up. So I think, you know,
341
00:26:49.480 --> 00:26:53.480
one of the things we ask is, you know, how do you, the question is, how do you measure?
342
00:26:54.200 --> 00:26:57.960
What do you measure for your business? How do you measure success? How do you measure what's good
343
00:26:57.960 --> 00:27:04.600
and what's not good? And a lot of people don't know what to say or they'll say, you know, I'll look
344
00:27:04.600 --> 00:27:08.040
with some of my bank account. Yeah. And it's questioning how much of that is yours. Yeah.
345
00:27:08.040 --> 00:27:12.280
Right. And they go, what do you mean? Well, and things like that. And it's like, well,
346
00:27:12.280 --> 00:27:15.640
how do you measure? And you know, if you don't monitor things and you can't measure them.
347
00:27:15.640 --> 00:27:19.160
And that's the first part of call for any sort of KPIs. But it's really just getting
348
00:27:19.160 --> 00:27:24.440
in to understand because nobody also, I don't know about, I'm terrible at reading. And it sounds,
349
00:27:24.440 --> 00:27:29.160
I'll just bring it into context. I will read a page at five times. I think I must have some sort
350
00:27:29.160 --> 00:27:33.480
of level of ADHD or something because I can't focus on it. I'm thinking about other things
351
00:27:33.480 --> 00:27:37.080
as I'm reading it. So it doesn't go in. If I listen to it, if you tell me something, I won't
352
00:27:37.080 --> 00:27:41.560
forget it. Okay. Right. And if or if I see it or watch it, I won't forget it. So I'm, I have
353
00:27:41.560 --> 00:27:45.320
to listen to a lot of audiobooks and podcasts and stuff like that. That's how I learn. And as I
354
00:27:45.400 --> 00:27:51.640
get kid, I was better at learning from things on TV or whatever. But I think if they can't look
355
00:27:51.640 --> 00:27:56.920
at a page of numbers because they just look at a whole P&L that goes, numbers can't look at it.
356
00:27:56.920 --> 00:28:02.040
So just engage. So we've got it. We're big into creating something that is going to be simple to
357
00:28:02.040 --> 00:28:06.920
read, simply to understand and help them run their business the best way. Call it a dashboard.
358
00:28:06.920 --> 00:28:13.880
I like the newspaper view. So I like headline, bit detail, full detail. Right. So you can get a
359
00:28:13.880 --> 00:28:17.240
really good snapshot because when you sort of go, oh, that doesn't look right. Why? Oh, because
360
00:28:17.240 --> 00:28:20.600
that. Oh, yeah. Remember that. That's fine. Who fixed that? No problem. They can move on
361
00:28:20.600 --> 00:28:24.520
with the day. It probably needs looking at 10 minutes once a month. You know, if there's any
362
00:28:24.520 --> 00:28:28.840
deep dough stuff, obviously they can do more. But ultimately it's about just knowing you're on track,
363
00:28:28.840 --> 00:28:32.920
knowing that all the hard work you're putting in, all the effort, all the people and employees
364
00:28:32.920 --> 00:28:36.840
who are responsible for as a business owner are all going to be looked after and going well,
365
00:28:36.840 --> 00:28:40.200
the business is going well and you're and you're struggling towards those targets and goals.
366
00:28:40.440 --> 00:28:46.120
But people, so many people
once they've been in a business for a long time, don't look to review
367
00:28:46.120 --> 00:28:52.520
it, don't look to do those things, gut feel. I never, I never distrust a business owner's gut feel
368
00:28:52.520 --> 00:28:57.160
because it's your business. Right. And your gut feels true and really bang on what we like to do.
369
00:28:57.160 --> 00:29:00.600
It's back out with numbers. Yeah, let's just stress test. Right. Yeah. Just just get it out there,
370
00:29:00.600 --> 00:29:07.560
but that's that's where the value is for me in doing what we do. So, um, but yeah, pulling
371
00:29:07.560 --> 00:29:13.000
in the AI, I think it's even though the AI does it, you see that sort of need a human to go
372
00:29:13.000 --> 00:29:19.080
we checked it and it's fine. Yeah. I know. So I did I recorded a podcast this week.
373
00:29:19.080 --> 00:29:26.040
Yeah. I have a podcast this week with someone who they are the other end. They are the other end.
374
00:29:26.040 --> 00:29:32.520
And it was a great podcast because I mean, they have a business that supports that viewpoint.
375
00:29:32.520 --> 00:29:36.200
Right. Yeah. Like we all have been we all we're all coming at a different end of the spectrum
376
00:29:36.200 --> 00:29:41.720
based on our reality and our impact, et cetera. And and their viewpoint was that actually,
377
00:29:42.760 --> 00:29:50.200
we we make more emotional errors than something that is just measuring objective data,
378
00:29:50.200 --> 00:29:55.640
analytics, et cetera. And so autonomy is where we're like where we need to go. We need to take
379
00:29:55.640 --> 00:30:02.280
humans outside of the loop and and do our stuff and and my my debate and argument we had
380
00:30:03.000 --> 00:30:08.200
and you can listen to it on our podcast. A little plug for the loop. My my debate, well,
381
00:30:08.200 --> 00:30:10.760
and actually we're going to have a we're going to do a second debate on it because we couldn't
382
00:30:10.760 --> 00:30:18.520
finish it in the time. Okay. My debate was that we are emotional human beings that have egos,
383
00:30:19.240 --> 00:30:24.680
trust is important to us, relationships important to us. And I just don't think that anytime soon
384
00:30:25.960 --> 00:30:31.480
that is going to happen because there's too much friction. There's too much friction in our
385
00:30:32.280 --> 00:30:39.080
our our makeup as tribal animals to to remove wanting to have humans human kind of check and
386
00:30:39.080 --> 00:30:45.320
challenge. I grew on both sides. Okay. There's going to be depending on what it is. Yeah.
387
00:30:45.320 --> 00:30:50.760
There's going to be room for both. And I think this is where like there's no there's no like yes or
388
00:30:50.760 --> 00:30:55.720
no, I think it's and this signs like I'm sitting on a fence. I'm not. There were very some definite
389
00:30:55.720 --> 00:31:00.120
hard knows on one side and hard knows on the other side. But like you know, you can probably agree
390
00:31:00.200 --> 00:31:04.680
that there are definitely some spaces where full autonomy would be great in certain circumstances.
391
00:31:04.680 --> 00:31:08.440
Like I don't give a shit about renewing my insurance every year. But so long as I know,
392
00:31:08.440 --> 00:31:12.600
if I give it a parameter of like make sure it's cheaper or better deologue or break it down for
393
00:31:12.600 --> 00:31:16.760
me so I can make it simple. That's just one example. Like those sorts of companies should be doing
394
00:31:16.760 --> 00:31:21.800
better. Yeah. With the AI to to give us that information so that we can make better decisions.
395
00:31:21.800 --> 00:31:24.840
But like there's do you know what I mean? All the stuff we don't like. So let's take that. Let's
396
00:31:24.840 --> 00:31:29.400
push that. From our side as a customer. Like from our side as a customer, I completely agree. Like
397
00:31:29.480 --> 00:31:34.920
all insurance should just happen. It's like what insurances do you want? This this this this whatever
398
00:31:34.920 --> 00:31:39.880
cover never worry about everything. Exactly. Never worry about it. And that has to be like a third party
399
00:31:39.880 --> 00:31:46.920
rather than yeah, whoever. However, however, the human friction I'm talking about is the insurance
400
00:31:46.920 --> 00:31:55.400
companies. They don't want that because that will that will cut their value ad to say that actually
401
00:31:55.400 --> 00:32:03.720
yeah, we might not be as cheap as X insurer, but we have a 98% 99% payout rate. And so you've got
402
00:32:03.720 --> 00:32:09.480
this friction on the other side of that business saying no, no, no, we can't we can't have autonomy
403
00:32:09.480 --> 00:32:14.760
because our business adds value. And like in some respects, they're kind of right. But in the other
404
00:32:14.760 --> 00:32:19.720
respects, that's that friction that human friction bit I'm talking about. Well, you just identified
405
00:32:19.880 --> 00:32:25.800
there is the fight the fight for value. Yeah. That's what that human interaction is. Yeah,
406
00:32:25.800 --> 00:32:32.440
it's the fight for value, isn't it? So as soon as we go as soon as you realize something isn't real,
407
00:32:33.240 --> 00:32:39.240
you check out. Yeah. So it's like we were talking the other day about what they called avatars,
408
00:32:39.240 --> 00:32:43.800
right? So I was like, you know, my brain instantly goes to cool. People can see more of me, right?
409
00:32:44.760 --> 00:32:48.760
So I was like, cool. So, you know, somebody signs a proposal. That's just something to
410
00:32:48.760 --> 00:32:53.240
propose. Well, you pop up. Yeah, right. It was just, yeah, even pop up. That'd be great.
411
00:32:53.240 --> 00:33:01.960
Yeah. No, nobody wants that. Okay, I'm from literally me popping up. But the idea being
412
00:33:01.960 --> 00:33:05.320
was that, but I thought, do you know what? No, because it's not authentic.
And it's not real.
413
00:33:05.320 --> 00:33:09.080
It's not people would just go, oh, I've signed another one. It's him again doing the same thing.
414
00:33:09.560 --> 00:33:15.160
Sign here. Yeah, it's one of those. So I'm like, no, that's not a rubbish idea. But at the same time,
415
00:33:15.160 --> 00:33:19.480
training, you know, okay, do they need to see, you know, be great to see me? And then I'm thinking,
416
00:33:20.120 --> 00:33:23.800
it's training. They doesn't need to be meeting anyone, anyone. I just want to really dare.
417
00:33:23.800 --> 00:33:28.600
So I start winding it back again to come in completely back off it. And this is I'm talking about
418
00:33:28.600 --> 00:33:32.280
the boring training, not not trying to be, you know, creating a room or anything. But
419
00:33:33.720 --> 00:33:37.640
it does that then make a difference. Well, for me, it would because I learned visually, right?
420
00:33:37.640 --> 00:33:42.280
I'm not going to read a book. So it's kind of that helps me. So I think if it's an aid, great.
421
00:33:42.280 --> 00:33:48.280
And if it's there, great. But I don't necessarily, isn't it me me? That's what I think I went full
422
00:33:48.280 --> 00:33:53.960
circle on. Avatar of me. Yeah, no avatars of me. Exactly. Yeah. And that for me, yeah, I process
423
00:33:53.960 --> 00:33:57.400
that in about five minutes. But that, but that's important. I think that discernment.
424
00:33:58.200 --> 00:34:04.680
Where is the right place for us to use? Absolutely. AI, well, just all tech advancements.
425
00:34:04.680 --> 00:34:08.840
Like, is this the right use case? Can we do it another way? If we could do another way,
426
00:34:08.840 --> 00:34:13.160
let's do another way. Or like what is the best output for this customer or it's an internal
427
00:34:13.160 --> 00:34:17.000
customer or an external customer? And and training is the interesting one. So I want to come back to
428
00:34:17.000 --> 00:34:23.720
what you said earlier about your purchase ledger, right? Your purchase ledger transforms
429
00:34:24.600 --> 00:34:30.760
from doing what they're doing, yeah, to being to having that skill set that they've got tenure
430
00:34:30.760 --> 00:34:34.760
and experience. They know what they they know what the skill set is. Yeah. They know what good looks
431
00:34:34.760 --> 00:34:40.840
like. They know what bad looks like. Yeah. With you're then talking about them training developing
432
00:34:40.840 --> 00:34:47.160
and coaching. Yeah. And by coaching for me, that's a feedback loop to if that if that process,
433
00:34:47.160 --> 00:34:51.480
that agent is is out of sync. Yeah. We need to retweet it or we need to upgrade it or we need
434
00:34:51.480 --> 00:34:56.360
to coach. So coaching isn't just build, train, I'm out of a job. Yeah. It's an ongoing iterative
435
00:34:56.440 --> 00:35:02.920
process. For me, that then that then shifts the complexity of the skills required for that
436
00:35:02.920 -->
00:35:08.920
individual. Correct. In terms of that that job, PM 70K brilliant, because actually they have had
437
00:35:08.920 --> 00:35:15.880
to develop a complimentary skill set. They're not everyone will get, but actually not everyone will
438
00:35:15.880 --> 00:35:21.480
get rapidly. No, everyone get rapidly. So we're pushing back to people like SEMA, AACC, AAT,
439
00:35:21.560 --> 00:35:25.800
all of the training providers, if you like, for the qualifications that are needed to come into
440
00:35:25.800 --> 00:35:30.760
our industry to say you need to have an element of agentic learning, an element that they are
441
00:35:30.760 --> 00:35:36.520
prompting. You need to have all the different that skill set is now very much in our industry.
442
00:35:36.520 --> 00:35:41.480
If you give me someone brand new who hasn't gone that skill, I've got to teach him. Yes.
443
00:35:41.480 --> 00:35:46.200
Right. So, but it's now part of the industry. Yeah. So they're going to need to know it.
444
00:35:46.200 --> 00:35:49.720
They should also in there have in there how to communicate with the business owner how to
445
00:35:49.720 --> 00:35:55.000
understand the business owner's needs. So we've got two divergent skills haven't we? Yeah.
446
00:35:55.000 --> 00:36:01.480
Like one divergent skill is actually we need our employees to be tech literate and also to be
447
00:36:01.480 --> 00:36:07.080
above the line in terms of thinking of continuing to develop that, but also we need them to be off
448
00:36:07.080 --> 00:36:12.600
YouTube and off TikTok and develop human skills. Absolutely. Of how to influence, how to negotiate,
449
00:36:12.600 --> 00:36:17.720
how to communicate. So question to you, business owner. Yep. All right. Leader of a team.
450
00:36:18.840 --> 00:36:26.360
In the next few months, years then now, does that mean that you have a more complex,
451
00:36:26.360 --> 00:36:32.520
you have more complex needs from your team members or it's easier with AI for you as a leader?
452
00:36:33.160 --> 00:36:42.840
You can't sit on the fence. No. It gives us the opportunity. For me, it gives my team the
453
00:36:42.840 --> 00:36:50.360
opportunity to have more human contact because the work is looked after. Their role has changed.
454
00:36:50.360 --> 00:36:54.200
That's a quick example would be not normally, this is going to be really a bit of detail,
455
00:36:54.200 --> 00:36:57.800
but normally with the person more, so you would check it. You want to talk about the person
456
00:36:57.880 --> 00:37:01.720
as you check it. You make sure it's right. When you enter it, you don't have to check it again.
457
00:37:01.720 --> 00:37:09.400
Done. Yeah. Right. The way it works now is that input is like a second, not the amount of time
458
00:37:09.400 --> 00:37:14.280
it took you to do it. Yeah. So save the nine hours. So you save chunk of time. Yeah. What happens is
459
00:37:14.280 --> 00:37:19.800
there is like, there's a lot of software manufactured out there that whatever system they use,
460
00:37:19.800 --> 00:37:23.480
it sends a receipt and an invoice, they look identical. So once it's received, once it's
461
00:37:24.360 --> 00:37:27.800
and they get sent via email and they get flooded into the system. So they're going to come through
462
00:37:27.800 --> 00:37:32.840
as duplicates. Yeah. So one, it can potentially look to see if the duplicate. It's got a different
463
00:37:32.840 --> 00:37:37.240
number than it won't know. It comes all the way through and then we have another system that
464
00:37:37.240 --> 00:37:40.360
checks the duplicates. So we're actually checking it this end now rather than that end because the
465
00:37:40.360 --> 00:37:44.280
amount we have to check is probably less time than having to do all the input. So that works great.
466
00:37:44.280 --> 00:37:47.320
With the agent now, it's though, we're saying like, right, if it says receipt on it flag it and
467
00:37:47.320 --> 00:37:51.160
it says invoice process it. Nice. Right. So that takes another little bit out. So the more and more
468
00:37:51.160 --> 00:37:57.160
this stuff gets developed, then all I can see is my team have more time and I've always said you
469
00:37:57.160 --> 00:38:01.400
know, if you earn your time back, you have it. Right. You've got, I'll give you an amount of work to
470
00:38:01.400 --> 00:38:07.880
do. If you do that amount of work, you're done for the day. And that was working fine until AI
471
00:38:07.880 --> 00:38:11.640
comes along. But then I'm thinking, okay, how can we, how can we leverage this but make it fair?
472
00:38:12.280 --> 00:38:17.080
Okay. So if you can tell it's your work, I'll double triple your pay because I'm charging for
473
00:38:17.080 --> 00:38:22.360
that 10x of work. Yeah. Yeah. So what, but what we've then got to manage is the value. Yeah.
474
00:38:22.360 --> 00:38:26.280
All right. So we still got maintain the value just to fact that because we can do brilliantly
475
00:38:26.280 --> 00:38:31.640
and super like efficient and we've gone to the effort of making this work, that's our win.
476
00:38:32.280 --> 00:38:36.200
Right. We still need to be able to charge what we charge. Yeah. So I think that's the only
477
00:38:36.200 --> 00:38:41.880
element to it is we have to protect the skill sets and the amount of effort that goes into getting
478
00:38:41.880 --> 00:38:48.360
to this point generation long will be different again. So yeah, we've we've always got to be
479
00:38:49.400 --> 00:38:54.600
able to pivot. I think you can't, you can't, you can't stop moving on this. It's like
480
00:38:55.480 --> 00:38:59.000
as much like, hey, yes, it's like a website and it's never ending. It was going up. They
481
00:38:59.000 --> 00:39:02.120
all something to do on it. You know, it's one of these things. So hang on, I've just launched
482
00:39:02.120 --> 00:39:05.640
a website today. Do not tell me I have to change it. It's out of date
tomorrow.
483
00:39:06.280 --> 00:39:10.920
All right. So look, I think let's pull it. Thank you so much for coming on the show. I think,
484
00:39:10.920 --> 00:39:16.440
I think some of the key takeaways from me, they're just there, there's a, it sounds like, you know,
485
00:39:16.440 --> 00:39:21.640
you said you're a tinker yourself and you know where you want to be going with this and then
486
00:39:21.640 --> 00:39:25.640
and then you've got a nice kind of experiment protective time for your team to discern which
487
00:39:25.640 --> 00:39:30.840
bits are appropriate for them. The messaging I take away from this is about like, people have
488
00:39:30.840 --> 00:39:36.200
got to be super adapters. And actually the skill set and the employees and the leaders and
489
00:39:36.200 --> 00:39:42.040
managers of the future have got to be really agile and open to adaptations and experimenting
490
00:39:42.040 --> 00:39:47.080
in new ways of working and then really thinking about the time we are back. Yeah, absolutely.
491
00:39:47.080 --> 00:39:55.320
Time we are back. What do we do valiably with that to increase the customer value? Anything else
492
00:39:55.320 --> 00:40:00.840
you want to share? Yeah, I think, I really, one of the biggest things around protecting jobs and
493
00:40:00.840 --> 00:40:06.920
I think that's protecting jobs, like giving people the opportunity to actually lean into something
494
00:40:06.920 --> 00:40:11.000
that can protect their job. There isn't too much of a stretch and they can learn, like we should
495
00:40:11.000 --> 00:40:17.880
always be learning to improve. I was brought up from a grandparents, but like, they all, they were
496
00:40:17.880 --> 00:40:22.120
like, one of the first people I knew that had a mobile phone and like out of anybody, like
497
00:40:22.200 --> 00:40:26.360
or a VCR or whatever it was at the time, we had it because they just thought we've got to keep up
498
00:40:26.360 --> 00:40:31.080
with the trends and I think that's instilled in me and that's why I'm trying to stay a year to
499
00:40:31.080 --> 00:40:35.240
18 months ahead of the team, trying and testing things, that's going to be an amazing idea. We'll
500
00:40:35.240 --> 00:40:40.440
implement that, right? And all these little things the way along and I think what I'm trying to
501
00:40:40.440 --> 00:40:45.320
teach my team to be is that way is like, keep looking for opportunities, ways to save time.
502
00:40:45.320 --> 00:40:51.000
If you can come to work and do a week's work in a day, I'm stoked for you guys. That's incredible.
503
00:40:51.000 --> 00:40:53.800
I would love that. That would be my favourite thing because I have to still pay. You can do what
504
00:40:53.800 --> 00:40:59.560
you want. All the customers are happy and then we'll get paid, happy days, but there's got to be the
505
00:40:59.560 --> 00:41:03.800
next thing. And the reality is the goal post then shift. So you might have a lag where that time
506
00:41:03.800 --> 00:41:07.720
is brought back, but then the world gets more complex. Customers demands get more, you're able to
507
00:41:07.720 --> 00:41:15.720
offer more, etc. So look, what's happened? What happens to me is that, so look, say 10, 15 years ago,
508
00:41:15.800 --> 00:41:21.640
you're talking top 1% of companies can have a CFO. Look at today, fractional CFOs everywhere,
509
00:41:21.640 --> 00:41:25.800
this is the fastest growing thing in financial industry at the moment, fractional CFOs,
510
00:41:25.800 --> 00:41:30.280
you can get somebody that would be 120 grand a year to work for you for 12 grand a year on a
511
00:41:30.280 --> 00:41:35.960
fractional basis and give you their 120 grand a year skill set. So it means that 50% of businesses
512
00:41:35.960 --> 00:41:40.600
now can have that level of service, which means 50% of businesses need the information we deliver.
513
00:41:41.080 --> 00:41:46.680
So it just means more access. And that's what AI can give us. It gives us more access to more things
514
00:41:47.640 --> 00:41:54.600
at a top top level. And that's why it's for me, it's the most exciting thing. It's way more exciting
515
00:41:54.600 --> 00:41:59.240
internet comes out. I remember when that happens. So that's old. Yeah, I mean, we don't have to wait
516
00:41:59.240 --> 00:42:07.640
for the like, I heard the AOL actually got, got sold recently. But anyway, side, side conversation.
517
00:42:08.360 --> 00:42:12.360
I'm really excited as well. Like you seem really optimistic about the opportunity for
518
00:42:12.360 --> 00:42:16.040
Antiverse. I'm really optimistic about the opportunity for Antiverse. And so I think like,
519
00:42:16.760 --> 00:42:21.000
thank you so much for coming on the show. I think we've got a part two in us. I think there's
520
00:42:21.000 --> 00:42:25.880
more we could talk about than that to be honest. Part three, part four. But yeah, thanks so much for
521
00:42:25.880 --> 00:42:35.880
today. I loved it. Thank you. It's covered me.