1
00:00:00,000 --> 00:00:03,000
foreign

2
00:00:04,960 --> 00:00:08,490
[Music]

3
00:00:18,000 --> 00:00:23,340
I'm here to speak to you today about my

4
00:00:20,699 --> 00:00:25,380
vision for better government

5
00:00:23,340 --> 00:00:27,180
this is about a

6
00:00:25,380 --> 00:00:31,980
type of government which is accountable

7
00:00:27,180 --> 00:00:33,600
is able to demonstrate how all decisions

8
00:00:31,980 --> 00:00:35,520
are made how all actions are taken how

9
00:00:33,600 --> 00:00:38,100
all policy and services and everything

10
00:00:35,520 --> 00:00:40,379
that government administers is done in a

11
00:00:38,100 --> 00:00:42,239
way that is demonstrably lawful

12
00:00:40,379 --> 00:00:43,460
demonstrably

13
00:00:42,239 --> 00:00:46,620
um

14
00:00:43,460 --> 00:00:48,780
aligned to the Mandate and accountable

15
00:00:46,620 --> 00:00:50,760
actions of that government I want to see

16
00:00:48,780 --> 00:00:52,520
a government that is participatory where

17
00:00:50,760 --> 00:00:56,399
the people the public the communities

18
00:00:52,520 --> 00:00:59,160
the families can not just

19
00:00:56,399 --> 00:01:01,320
um deal with policies and services once

20
00:00:59,160 --> 00:01:03,780
they're launched but can be an active

21
00:01:01,320 --> 00:01:06,659
contributor to or even razor of new

22
00:01:03,780 --> 00:01:08,460
ideas of new policies of new things that

23
00:01:06,659 --> 00:01:10,680
they need to live well and can

24
00:01:08,460 --> 00:01:12,240
participate not just in the design of

25
00:01:10,680 --> 00:01:14,340
such policies or services but in the

26
00:01:12,240 --> 00:01:16,320
governance and oversight I'd love to see

27
00:01:14,340 --> 00:01:19,680
a government which is more Humane which

28
00:01:16,320 --> 00:01:21,840
brings our whole person to the job which

29
00:01:19,680 --> 00:01:24,540
brings human measures of success

30
00:01:21,840 --> 00:01:26,520
Community measures of success which

31
00:01:24,540 --> 00:01:28,259
brings values into the way that

32
00:01:26,520 --> 00:01:30,960
everything is done in the public service

33
00:01:28,259 --> 00:01:33,180
how can you have Humane outcomes if you

34
00:01:30,960 --> 00:01:36,479
don't have a Humane workplace and Humane

35
00:01:33,180 --> 00:01:38,280
policy making and Humane ways of working

36
00:01:36,479 --> 00:01:40,439
I want to see a public service and a

37
00:01:38,280 --> 00:01:42,439
government which is adaptive which is

38
00:01:40,439 --> 00:01:44,400
not just sit and forget but is

39
00:01:42,439 --> 00:01:46,259
constantly monitoring for and

40
00:01:44,400 --> 00:01:49,619
understanding and receiving feedback

41
00:01:46,259 --> 00:01:52,380
from the community on the impact of

42
00:01:49,619 --> 00:01:54,119
change so that then we can truly take an

43
00:01:52,380 --> 00:01:55,979
iterative approach that is based on

44
00:01:54,119 --> 00:01:58,320
evidence based on reality based on the

45
00:01:55,979 --> 00:02:00,000
actual effects it's having on people so

46
00:01:58,320 --> 00:02:01,799
that we can constantly guide ourselves

47
00:02:00,000 --> 00:02:03,540
towards a better North Star

48
00:02:01,799 --> 00:02:05,759
and I want to say of course Public

49
00:02:03,540 --> 00:02:07,799
Services which are accessible I want to

50
00:02:05,759 --> 00:02:10,080
see a situation where everyone knows

51
00:02:07,799 --> 00:02:13,260
doesn't just have access according to

52
00:02:10,080 --> 00:02:14,760
their needs and but also has access to

53
00:02:13,260 --> 00:02:17,340
their rights

54
00:02:14,760 --> 00:02:20,520
um where everyone has access to Justice

55
00:02:17,340 --> 00:02:23,520
everyone has access to everything they

56
00:02:20,520 --> 00:02:25,379
need to live well

57
00:02:23,520 --> 00:02:27,840
part of this is about all of us

58
00:02:25,379 --> 00:02:29,580
recognizing that you will always get

59
00:02:27,840 --> 00:02:32,099
something which is lower than your

60
00:02:29,580 --> 00:02:34,140
expectation so if you start with a low

61
00:02:32,099 --> 00:02:36,900
expectation guess what you're going to

62
00:02:34,140 --> 00:02:38,879
get something even lower than that if we

63
00:02:36,900 --> 00:02:41,220
start to raise our expectations of

64
00:02:38,879 --> 00:02:42,440
government of each other of public

65
00:02:41,220 --> 00:02:45,720
servants

66
00:02:42,440 --> 00:02:49,620
if we start from a why isn't it the best

67
00:02:45,720 --> 00:02:51,959
possible action the best possible way of

68
00:02:49,620 --> 00:02:54,360
working when you raise your expectations

69
00:02:51,959 --> 00:02:56,220
you have a chance of changing towards

70
00:02:54,360 --> 00:02:57,959
that raised expectation

71
00:02:56,220 --> 00:03:00,840
when you start from a low expectation

72
00:02:57,959 --> 00:03:03,000
you're going to get what you expected so

73
00:03:00,840 --> 00:03:05,459
I'd like to encourage you all to start

74
00:03:03,000 --> 00:03:09,120
to raise your expectations of government

75
00:03:05,459 --> 00:03:10,860
of public sector and of society so that

76
00:03:09,120 --> 00:03:14,159
we don't end up with a situation where

77
00:03:10,860 --> 00:03:16,560
cynicism desire defines the North Star

78
00:03:14,159 --> 00:03:19,200
policy is this weird word it means

79
00:03:16,560 --> 00:03:21,000
literally pretty much anything for some

80
00:03:19,200 --> 00:03:23,459
people it is a couple of dot points on a

81
00:03:21,000 --> 00:03:24,840
napkin for some people it is the promise

82
00:03:23,459 --> 00:03:26,940
that the government of the day said for

83
00:03:24,840 --> 00:03:28,500
some people it's legislation and the

84
00:03:26,940 --> 00:03:29,400
simple answer is actually all of these

85
00:03:28,500 --> 00:03:31,019
things

86
00:03:29,400 --> 00:03:32,940
the one thing that I would suggest

87
00:03:31,019 --> 00:03:34,920
particularly to a crowd like this is

88
00:03:32,940 --> 00:03:36,239
that policy is like the operating system

89
00:03:34,920 --> 00:03:37,980
for our society

90
00:03:36,239 --> 00:03:39,959
it sort of determines the rules

91
00:03:37,980 --> 00:03:43,159
determines what's available to you it

92
00:03:39,959 --> 00:03:45,480
determines the support mechanisms it is

93
00:03:43,159 --> 00:03:47,159
it's a series of libraries that we all

94
00:03:45,480 --> 00:03:50,220
draw upon every day even if we don't

95
00:03:47,159 --> 00:03:51,659
notice it when we drive our own cars

96
00:03:50,220 --> 00:03:53,340
when we drive public transport when we

97
00:03:51,659 --> 00:03:56,099
use the education system when we use the

98
00:03:53,340 --> 00:03:57,360
health system even just

99
00:03:56,099 --> 00:03:59,580
um you know catching up in a location

100
00:03:57,360 --> 00:04:02,640
like this there are many many policies

101
00:03:59,580 --> 00:04:04,019
that affect us every day and and that's

102
00:04:02,640 --> 00:04:06,659
why it's so critical we get involved in

103
00:04:04,019 --> 00:04:08,640
it so this is my very cheeky way of

104
00:04:06,659 --> 00:04:09,540
describing what policy looks like so for

105
00:04:08,640 --> 00:04:11,580
those

106
00:04:09,540 --> 00:04:13,620
um who can't see the screen I've got a

107
00:04:11,580 --> 00:04:14,939
cartoon here and it basically starts

108
00:04:13,620 --> 00:04:16,859
with the policy team are celebrating

109
00:04:14,939 --> 00:04:18,479
they're saying another great job this is

110
00:04:16,859 --> 00:04:20,100
the current state of how policy is made

111
00:04:18,479 --> 00:04:22,019
in Australia generally and in a lot of

112
00:04:20,100 --> 00:04:23,940
countries great job we got it out the

113
00:04:22,019 --> 00:04:25,199
door now on to the next policy and in

114
00:04:23,940 --> 00:04:26,759
the back of their minds the policy team

115
00:04:25,199 --> 00:04:28,259
are thinking wow I wish we had time to

116
00:04:26,759 --> 00:04:30,360
explore policy Futures to understand

117
00:04:28,259 --> 00:04:31,979
what this looks like even to see what it

118
00:04:30,360 --> 00:04:33,540
means in implementation

119
00:04:31,979 --> 00:04:35,639
but the policy instructions are thrown

120
00:04:33,540 --> 00:04:37,520
into a black hole the black hole of

121
00:04:35,639 --> 00:04:40,320
policy intent and from that black hole

122
00:04:37,520 --> 00:04:41,820
from which purpose is stripped out those

123
00:04:40,320 --> 00:04:43,800
instructions are received by someone to

124
00:04:41,820 --> 00:04:46,020
implement it it might be someone in a

125
00:04:43,800 --> 00:04:47,580
company that has to consume a regulation

126
00:04:46,020 --> 00:04:50,880
it might be someone in a social services

127
00:04:47,580 --> 00:04:52,680
organization that needs to build a new

128
00:04:50,880 --> 00:04:54,240
Public Service it need it might it could

129
00:04:52,680 --> 00:04:56,520
be by anyone who has to consume these

130
00:04:54,240 --> 00:04:59,040
rules who are sort of saying what does

131
00:04:56,520 --> 00:05:00,479
this mean well unfortunately the team

132
00:04:59,040 --> 00:05:02,040
that wrote it on to the next thing and

133
00:05:00,479 --> 00:05:03,540
so let's just do our best so there's a

134
00:05:02,040 --> 00:05:05,220
lot of guessing

135
00:05:03,540 --> 00:05:06,360
and the people actually affected by bad

136
00:05:05,220 --> 00:05:08,820
policy

137
00:05:06,360 --> 00:05:11,100
literally it can ruin people's lives as

138
00:05:08,820 --> 00:05:13,440
we've seen in recent months through

139
00:05:11,100 --> 00:05:15,000
various rural commissions

140
00:05:13,440 --> 00:05:17,100
so what we end up having is this black

141
00:05:15,000 --> 00:05:19,380
hole policy impact

142
00:05:17,100 --> 00:05:21,360
no one knows what the actual impact of

143
00:05:19,380 --> 00:05:23,940
policy is and the people that are

144
00:05:21,360 --> 00:05:25,380
generally measuring policy are looking

145
00:05:23,940 --> 00:05:26,639
to measure the things that they expected

146
00:05:25,380 --> 00:05:28,380
not looking for the unintended

147
00:05:26,639 --> 00:05:29,940
consequences so it can be devastating

148
00:05:28,380 --> 00:05:31,979
and of course eventually you get an

149
00:05:29,940 --> 00:05:33,960
evaluation either an evaluation team or

150
00:05:31,979 --> 00:05:36,240
a royal commission which will eventually

151
00:05:33,960 --> 00:05:38,880
say

152
00:05:36,240 --> 00:05:40,259
there was low intent policy intent or

153
00:05:38,880 --> 00:05:43,380
indeed if there was high policy

154
00:05:40,259 --> 00:05:44,880
realization there was harm created I

155
00:05:43,380 --> 00:05:46,020
wish they'd evaluated Turner but of

156
00:05:44,880 --> 00:05:48,060
course that's always the answer let's

157
00:05:46,020 --> 00:05:49,500
just evaluate sooner and the problem is

158
00:05:48,060 --> 00:05:51,240
that an evaluation is a point in time

159
00:05:49,500 --> 00:05:54,060
but it's not feeding into constant

160
00:05:51,240 --> 00:05:56,160
iteration is it in the tech sector we

161
00:05:54,060 --> 00:05:58,860
all shifted years ago now into

162
00:05:56,160 --> 00:06:00,539
continuous Improvement for systems for

163
00:05:58,860 --> 00:06:03,120
products for services

164
00:06:00,539 --> 00:06:04,919
but there's no equivalent yet in policy

165
00:06:03,120 --> 00:06:06,780
so what we've got is a situation where

166
00:06:04,919 --> 00:06:08,220
there's a Gap a gap between policy and

167
00:06:06,780 --> 00:06:10,139
delivery

168
00:06:08,220 --> 00:06:12,660
um and it is a big gap and just to give

169
00:06:10,139 --> 00:06:14,220
you an example of um the sort of stories

170
00:06:12,660 --> 00:06:16,380
I remember going to a regulatory

171
00:06:14,220 --> 00:06:17,880
conference a Regulators conference and a

172
00:06:16,380 --> 00:06:19,440
regulate tour got up and said there's

173
00:06:17,880 --> 00:06:20,820
nothing wrong with our policy our policy

174
00:06:19,440 --> 00:06:22,500
was perfect it was the implementation

175
00:06:20,820 --> 00:06:24,000
that was the problem

176
00:06:22,500 --> 00:06:25,500
and I got up straight after this person

177
00:06:24,000 --> 00:06:27,600
said has one of your besmirched

178
00:06:25,500 --> 00:06:29,699
implementation people I could have told

179
00:06:27,600 --> 00:06:31,500
you that when you did regulation for

180
00:06:29,699 --> 00:06:33,360
heart monitors which is the regulation

181
00:06:31,500 --> 00:06:34,620
he was talking about you know anyone

182
00:06:33,360 --> 00:06:37,440
with a technical band could have told

183
00:06:34,620 --> 00:06:39,240
you that open wi-fi was a bad idea

184
00:06:37,440 --> 00:06:40,979
um the fact is that every piece of

185
00:06:39,240 --> 00:06:42,960
policy every regulation every

186
00:06:40,979 --> 00:06:44,460
legislation is being implemented in

187
00:06:42,960 --> 00:06:45,720
technology today

188
00:06:44,460 --> 00:06:47,220
which means if you don't have

189
00:06:45,720 --> 00:06:49,020
multi-disciplinary teams if you don't

190
00:06:47,220 --> 00:06:51,060
have technologists together with data

191
00:06:49,020 --> 00:06:53,280
together with social activists together

192
00:06:51,060 --> 00:06:55,680
with um you know highly diverse

193
00:06:53,280 --> 00:06:58,319
perspectives and experiences in the room

194
00:06:55,680 --> 00:06:59,819
when you're developing policy then

195
00:06:58,319 --> 00:07:02,160
chances are first of all it'll probably

196
00:06:59,819 --> 00:07:03,419
miss the mark and chances are it will

197
00:07:02,160 --> 00:07:04,620
probably not be fit for purpose and

198
00:07:03,419 --> 00:07:05,900
chances are it will probably create

199
00:07:04,620 --> 00:07:08,580
unintended

200
00:07:05,900 --> 00:07:10,199
consequences because you're developing

201
00:07:08,580 --> 00:07:12,360
an isolation from those who can feed

202
00:07:10,199 --> 00:07:13,740
into it and of course the Blindfolded

203
00:07:12,360 --> 00:07:15,060
baton passing between policy and

204
00:07:13,740 --> 00:07:17,699
delivery means that the people doing

205
00:07:15,060 --> 00:07:19,680
delivery can't feed back into policy so

206
00:07:17,699 --> 00:07:22,979
you end up with a stagnant artifact very

207
00:07:19,680 --> 00:07:24,720
quickly this is my dream again my future

208
00:07:22,979 --> 00:07:26,880
we're on the way to getting there but

209
00:07:24,720 --> 00:07:29,280
policy options and Futures are

210
00:07:26,880 --> 00:07:30,840
continuously being explored continuously

211
00:07:29,280 --> 00:07:32,580
being raised not just from the policy

212
00:07:30,840 --> 00:07:34,620
team or from the government of the day

213
00:07:32,580 --> 00:07:37,979
but from the public

214
00:07:34,620 --> 00:07:40,139
here's a change I've noticed a bug why

215
00:07:37,979 --> 00:07:42,120
can't we actually have bug fixing of um

216
00:07:40,139 --> 00:07:43,039
of policy of services Etc from the

217
00:07:42,120 --> 00:07:45,180
public

218
00:07:43,039 --> 00:07:47,280
multi-disciplinary disciplinary policy

219
00:07:45,180 --> 00:07:48,780
development where people and policy

220
00:07:47,280 --> 00:07:51,060
teams and implementation teams are

221
00:07:48,780 --> 00:07:52,919
co-designing not just the implementation

222
00:07:51,060 --> 00:07:55,380
but the policy intent the instructions

223
00:07:52,919 --> 00:07:57,840
the success criteria and where you're

224
00:07:55,380 --> 00:08:00,539
actually monitoring and baselining not

225
00:07:57,840 --> 00:08:02,220
just the policy intent but the human

226
00:08:00,539 --> 00:08:03,660
impact from the start

227
00:08:02,220 --> 00:08:05,819
and then you can be constantly

228
00:08:03,660 --> 00:08:07,440
monitoring that proactively managing

229
00:08:05,819 --> 00:08:08,759
that and everyone working from the same

230
00:08:07,440 --> 00:08:10,979
policy infrastructure where you have

231
00:08:08,759 --> 00:08:12,900
policy as code so a policy twin a

232
00:08:10,979 --> 00:08:14,580
digital twin of policy combined with

233
00:08:12,900 --> 00:08:16,560
impact monitoring combined with actual

234
00:08:14,580 --> 00:08:19,319
escalation so when things go wrong which

235
00:08:16,560 --> 00:08:20,819
they will you notice and you can do

236
00:08:19,319 --> 00:08:23,520
something about it and people don't need

237
00:08:20,819 --> 00:08:25,139
to have their lives ruined

238
00:08:23,520 --> 00:08:26,940
so changing things starts with

239
00:08:25,139 --> 00:08:29,220
reimagining good we currently have a

240
00:08:26,940 --> 00:08:30,599
highly manual process where policies are

241
00:08:29,220 --> 00:08:32,760
ideated quite often in political

242
00:08:30,599 --> 00:08:34,459
officers a policy announcement is made

243
00:08:32,760 --> 00:08:37,560
usually in the absence of any actual

244
00:08:34,459 --> 00:08:39,539
evidence then someone goes and develops

245
00:08:37,560 --> 00:08:41,099
the how can we create options to meet

246
00:08:39,539 --> 00:08:43,320
this policy announcements legislative

247
00:08:41,099 --> 00:08:46,140
drafting is very slow and very manual

248
00:08:43,320 --> 00:08:47,940
and and I've literally read legislation

249
00:08:46,140 --> 00:08:50,640
where it's got a triple negative over

250
00:08:47,940 --> 00:08:52,500
the course of 10 pages and years later I

251
00:08:50,640 --> 00:08:55,440
spoke to that drafter and said

252
00:08:52,500 --> 00:08:57,360
why and his answer was fascinating to me

253
00:08:55,440 --> 00:08:59,880
he said I was just having a bad year

254
00:08:57,360 --> 00:09:02,100
like

255
00:08:59,880 --> 00:09:06,360
the fact is

256
00:09:02,100 --> 00:09:08,100
come to write these things is so far

257
00:09:06,360 --> 00:09:09,540
from good it's not even funny so the

258
00:09:08,100 --> 00:09:11,220
Parliamentary process will always be

259
00:09:09,540 --> 00:09:12,660
slow and a lot of debates and you know

260
00:09:11,220 --> 00:09:15,660
we can't really stop politicians being

261
00:09:12,660 --> 00:09:16,980
politicians but publishing is fast but

262
00:09:15,660 --> 00:09:18,660
once you have a new piece of legislation

263
00:09:16,980 --> 00:09:20,160
or regulation art

264
00:09:18,660 --> 00:09:21,660
unless you're the sort of person that

265
00:09:20,160 --> 00:09:23,820
goes looking for and subscribing to

266
00:09:21,660 --> 00:09:25,620
those changes who knows you've got to

267
00:09:23,820 --> 00:09:26,760
discover it you've got to interpret it

268
00:09:25,620 --> 00:09:28,380
no one tells you when you've got to

269
00:09:26,760 --> 00:09:31,920
write only when you got it wrong so you

270
00:09:28,380 --> 00:09:34,800
get Myriad implementations Myriad

271
00:09:31,920 --> 00:09:36,720
um interpretations and implementations

272
00:09:34,800 --> 00:09:38,100
um and um and there's no Assurance of

273
00:09:36,720 --> 00:09:40,920
whether it's the right thing

274
00:09:38,100 --> 00:09:42,839
I see again a situation where policy or

275
00:09:40,920 --> 00:09:44,399
ideation if you actually had a policy

276
00:09:42,839 --> 00:09:46,140
twin where anyone including the

277
00:09:44,399 --> 00:09:48,420
political officers could actually model

278
00:09:46,140 --> 00:09:51,000
and try things out you can actually

279
00:09:48,420 --> 00:09:52,580
draft new regulation legislation as

280
00:09:51,000 --> 00:09:55,920
human and machine readable

281
00:09:52,580 --> 00:09:57,959
simultaneously not one then codify that

282
00:09:55,920 --> 00:10:00,060
into the other but actually implementing

283
00:09:57,959 --> 00:10:01,800
at the same time a reference

284
00:10:00,060 --> 00:10:03,180
implementation so you can feed the

285
00:10:01,800 --> 00:10:04,620
insights of that back into the human

286
00:10:03,180 --> 00:10:06,180
version and then you get something which

287
00:10:04,620 --> 00:10:08,940
is actually good

288
00:10:06,180 --> 00:10:10,200
imagine if um at the point in time where

289
00:10:08,940 --> 00:10:12,839
a new piece of legislation regulation

290
00:10:10,200 --> 00:10:14,700
was passed in the parliament the API was

291
00:10:12,839 --> 00:10:16,200
up like that imagine if at the point of

292
00:10:14,700 --> 00:10:17,760
Consulting on that prior to it even

293
00:10:16,200 --> 00:10:19,380
going to Parliament you could actually

294
00:10:17,760 --> 00:10:21,240
have a reference implementation and test

295
00:10:19,380 --> 00:10:23,540
it there's so much you can do when you

296
00:10:21,240 --> 00:10:26,220
actually start shifting to a

297
00:10:23,540 --> 00:10:27,660
digital and human approach to drafting

298
00:10:26,220 --> 00:10:30,300
these rules and we can start to

299
00:10:27,660 --> 00:10:31,740
dramatically improve the quality and the

300
00:10:30,300 --> 00:10:32,760
integrity and the consistency of how

301
00:10:31,740 --> 00:10:34,380
they're applied

302
00:10:32,760 --> 00:10:36,660
so there's lots of bits to this this is

303
00:10:34,380 --> 00:10:38,399
my mind map it's quite overwhelming I'm

304
00:10:36,660 --> 00:10:40,500
so sorry you're starting to see the

305
00:10:38,399 --> 00:10:42,480
brain of peer culminated over seven

306
00:10:40,500 --> 00:10:44,100
years so sorry

307
00:10:42,480 --> 00:10:46,860
um but here are sort of the enablers for

308
00:10:44,100 --> 00:10:48,300
that better pathway right so I'm not

309
00:10:46,860 --> 00:10:49,500
going to go through all all of these

310
00:10:48,300 --> 00:10:52,740
today in fact I'm only going to touch on

311
00:10:49,500 --> 00:10:54,720
two today but just really briefly if you

312
00:10:52,740 --> 00:10:57,300
want to get persistent impact monitoring

313
00:10:54,720 --> 00:10:59,339
informed and continuous policy reform

314
00:10:57,300 --> 00:11:01,019
based on those impacts you need to be

315
00:10:59,339 --> 00:11:02,579
monitoring you need to have publicly

316
00:11:01,019 --> 00:11:04,200
available modeling tools you need to a

317
00:11:02,579 --> 00:11:05,459
human impact measurements what do you

318
00:11:04,200 --> 00:11:06,540
actually measure what does quality of

319
00:11:05,459 --> 00:11:08,399
life look like and of course that's

320
00:11:06,540 --> 00:11:09,600
going to get very diverse when you start

321
00:11:08,399 --> 00:11:10,800
taking different knowledge systems into

322
00:11:09,600 --> 00:11:13,680
account

323
00:11:10,800 --> 00:11:15,660
actual public engagement and policy my

324
00:11:13,680 --> 00:11:18,680
favorite participatory policy story is

325
00:11:15,660 --> 00:11:20,820
from Taiwan where Uber did their usual

326
00:11:18,680 --> 00:11:22,260
backdoor Brute Force try to get you know

327
00:11:20,820 --> 00:11:23,459
regulation to get them to do whatever

328
00:11:22,260 --> 00:11:24,600
they want

329
00:11:23,459 --> 00:11:26,640
um but the way the government there

330
00:11:24,600 --> 00:11:29,220
works is thank you for your suggestion

331
00:11:26,640 --> 00:11:31,380
we're going to ask people

332
00:11:29,220 --> 00:11:33,720
and then they put out a how would you

333
00:11:31,380 --> 00:11:35,220
feel about Uber coming into Taiwan how

334
00:11:33,720 --> 00:11:37,740
do you feel this is something we're

335
00:11:35,220 --> 00:11:39,120
really terrible at in um in the west uh

336
00:11:37,740 --> 00:11:40,680
we sort of start with what do you think

337
00:11:39,120 --> 00:11:43,260
and the problem is what you think is

338
00:11:40,680 --> 00:11:45,899
people get married to their ideas very

339
00:11:43,260 --> 00:11:47,519
hard it's very hard to shift an idea but

340
00:11:45,899 --> 00:11:49,800
shifting a heart is actually a little

341
00:11:47,519 --> 00:11:51,660
bit different how do you feel draws out

342
00:11:49,800 --> 00:11:53,820
people's values and concerns and hopes

343
00:11:51,660 --> 00:11:55,680
well I feel like it could be a good

344
00:11:53,820 --> 00:11:57,420
opportunity to make some money I I'm

345
00:11:55,680 --> 00:11:59,100
worried that

346
00:11:57,420 --> 00:12:00,600
um you know someone's going to hurt me

347
00:11:59,100 --> 00:12:02,640
and then there's not going to be any way

348
00:12:00,600 --> 00:12:04,260
you know any recount recourse for that

349
00:12:02,640 --> 00:12:05,339
it draws out everyone's hopes and fears

350
00:12:04,260 --> 00:12:06,779
and then you can start to draw out well

351
00:12:05,339 --> 00:12:08,700
what does that mean and then you then

352
00:12:06,779 --> 00:12:11,339
what they did is invited the general

353
00:12:08,700 --> 00:12:13,800
public to participate in live

354
00:12:11,339 --> 00:12:16,860
legislative drafting

355
00:12:13,800 --> 00:12:19,200
live streamed live legislative drafting

356
00:12:16,860 --> 00:12:20,820
how cool does that sound right

357
00:12:19,200 --> 00:12:23,880
um and then and they said to Uber you're

358
00:12:20,820 --> 00:12:25,860
welcome to come or not we don't mind

359
00:12:23,880 --> 00:12:29,940
um and so rather than it being done in

360
00:12:25,860 --> 00:12:31,620
back you know back rooms by experts and

361
00:12:29,940 --> 00:12:33,959
lobbyists it is actually done in the

362
00:12:31,620 --> 00:12:35,820
open this is again a dream

363
00:12:33,959 --> 00:12:37,320
so public engagement

364
00:12:35,820 --> 00:12:39,779
um my favorite

365
00:12:37,320 --> 00:12:42,120
um piece of legislation in the world is

366
00:12:39,779 --> 00:12:44,040
um again total nerd here

367
00:12:42,120 --> 00:12:46,260
um is the right to an explanation is it

368
00:12:44,040 --> 00:12:48,480
well utilized or no no but in New

369
00:12:46,260 --> 00:12:50,160
Zealand in the foi equivalent over the

370
00:12:48,480 --> 00:12:52,980
oia they have a thing called section 23

371
00:12:50,160 --> 00:12:55,079
every citizen every resident in New

372
00:12:52,980 --> 00:12:56,700
Zealand has a legal right to an

373
00:12:55,079 --> 00:12:58,560
explanation about any decision made

374
00:12:56,700 --> 00:13:00,180
about them by the government what a

375
00:12:58,560 --> 00:13:01,079
beautiful piece of law that would

376
00:13:00,180 --> 00:13:02,639
certainly change things quite

377
00:13:01,079 --> 00:13:04,860
substantially here

378
00:13:02,639 --> 00:13:06,420
um human and unique policy and human

379
00:13:04,860 --> 00:13:08,399
indicators you need a public repository

380
00:13:06,420 --> 00:13:10,220
people can share stuff you need

381
00:13:08,399 --> 00:13:13,139
escalation mechanisms test Suites

382
00:13:10,220 --> 00:13:14,880
baselining policy as code and continuous

383
00:13:13,139 --> 00:13:17,100
monitoring so I'm only going to look at

384
00:13:14,880 --> 00:13:18,420
these two together so don't worry we're

385
00:13:17,100 --> 00:13:19,860
going to break this right down but I

386
00:13:18,420 --> 00:13:22,680
wanted to give you if you like the whole

387
00:13:19,860 --> 00:13:24,720
picture and why they um pretty much all

388
00:13:22,680 --> 00:13:26,639
of it does end up coming back to a

389
00:13:24,720 --> 00:13:29,399
Reliance upon policy as code

390
00:13:26,639 --> 00:13:32,760
my origin story in this started at um

391
00:13:29,399 --> 00:13:35,700
oztrack austrack is the

392
00:13:32,760 --> 00:13:37,139
um one of the Australian Regulators of

393
00:13:35,700 --> 00:13:38,940
the financial sector and they look after

394
00:13:37,139 --> 00:13:41,040
things like anti-money laundering and

395
00:13:38,940 --> 00:13:42,660
counterterrorism financing regulation

396
00:13:41,040 --> 00:13:45,180
and it was really interesting going

397
00:13:42,660 --> 00:13:47,040
there because a bunch of problems became

398
00:13:45,180 --> 00:13:49,079
really evident really quickly and that's

399
00:13:47,040 --> 00:13:52,620
why I got onto this whole rules code

400
00:13:49,079 --> 00:13:53,880
Journey one of which was everyone's into

401
00:13:52,620 --> 00:13:55,019
you know has to go and read and

402
00:13:53,880 --> 00:13:56,579
interpret and a lot of these

403
00:13:55,019 --> 00:13:58,500
principles-based regulations people are

404
00:13:56,579 --> 00:14:01,620
like oh what does that mean I don't know

405
00:13:58,500 --> 00:14:03,480
let's try did I get this right I'm not

406
00:14:01,620 --> 00:14:04,380
going to tell you oh we're just going to

407
00:14:03,480 --> 00:14:05,820
find you when you get it wrong so

408
00:14:04,380 --> 00:14:06,860
there's a lot of angst and a lot of

409
00:14:05,820 --> 00:14:09,300
stress

410
00:14:06,860 --> 00:14:11,040
so the seven problems I identified were

411
00:14:09,300 --> 00:14:12,540
first of all we just need better rules

412
00:14:11,040 --> 00:14:14,880
we need rules that are more easy to

413
00:14:12,540 --> 00:14:16,860
implement model Etc because that

414
00:14:14,880 --> 00:14:20,100
translation Gap I spoke about before so

415
00:14:16,860 --> 00:14:22,139
I joke a lot much to my ex-lawyer

416
00:14:20,100 --> 00:14:24,000
husband's annoyance

417
00:14:22,139 --> 00:14:26,220
um I joke a lot that we are currently

418
00:14:24,000 --> 00:14:27,720
using lawyers as modems

419
00:14:26,220 --> 00:14:29,940
we're using lawyers to translate between

420
00:14:27,720 --> 00:14:31,139
analog and digital to sort of say what

421
00:14:29,940 --> 00:14:33,120
does this legislation mean and how

422
00:14:31,139 --> 00:14:35,940
should we implement it we need to adapt

423
00:14:33,120 --> 00:14:37,860
to digital there's a lot of rules like

424
00:14:35,940 --> 00:14:39,540
the example about the heart monitor

425
00:14:37,860 --> 00:14:40,560
where they're just not making the

426
00:14:39,540 --> 00:14:41,760
assumption that things are going to be

427
00:14:40,560 --> 00:14:42,839
done in technology so they don't have

428
00:14:41,760 --> 00:14:45,300
any Geeks in the room they don't have

429
00:14:42,839 --> 00:14:48,120
any technologists informing the um the

430
00:14:45,300 --> 00:14:49,380
policy design so we need to create the

431
00:14:48,120 --> 00:14:50,820
understanding that everything that we're

432
00:14:49,380 --> 00:14:52,260
building is going to go into a digital

433
00:14:50,820 --> 00:14:53,760
system at some point so we need to take

434
00:14:52,260 --> 00:14:56,760
that into account

435
00:14:53,760 --> 00:14:58,920
a lot of regulations assume it's shaping

436
00:14:56,760 --> 00:15:01,620
a human behavior but of course that's

437
00:14:58,920 --> 00:15:03,959
increasingly not the case and a machine

438
00:15:01,620 --> 00:15:06,060
doesn't care about a Legal Financial or

439
00:15:03,959 --> 00:15:09,120
criminal pressure so if you're building

440
00:15:06,060 --> 00:15:11,639
a regulation to shape human behavior and

441
00:15:09,120 --> 00:15:14,519
then the core the main consumer of that

442
00:15:11,639 --> 00:15:17,459
rule is actually machine then you're

443
00:15:14,519 --> 00:15:18,660
probably not going to get your outcome

444
00:15:17,459 --> 00:15:20,279
um a lot of people just building

445
00:15:18,660 --> 00:15:21,480
rebuilding the status quo with shiny new

446
00:15:20,279 --> 00:15:23,399
things they're just speeding things up

447
00:15:21,480 --> 00:15:26,100
and running off a cliff

448
00:15:23,399 --> 00:15:29,399
and there's a lot of disincentives so

449
00:15:26,100 --> 00:15:32,760
they might have a key goal but they're

450
00:15:29,399 --> 00:15:35,399
secondary goals become the primary goal

451
00:15:32,760 --> 00:15:37,199
very quickly and and then they miss the

452
00:15:35,399 --> 00:15:38,459
outcome so adults track one of the key

453
00:15:37,199 --> 00:15:40,199
goals is to strengthen the financial

454
00:15:38,459 --> 00:15:41,760
system against abuse but then the

455
00:15:40,199 --> 00:15:43,560
secondary goals were to detect and to

456
00:15:41,760 --> 00:15:45,360
disrupt Financial crime fine

457
00:15:43,560 --> 00:15:46,980
all of the measures of success when I

458
00:15:45,360 --> 00:15:49,079
was there this is a while ago now were

459
00:15:46,980 --> 00:15:50,820
about how many criminals did we arrest

460
00:15:49,079 --> 00:15:52,620
and how many

461
00:15:50,820 --> 00:15:55,920
um you know non-compliance financial

462
00:15:52,620 --> 00:15:57,420
institutions did we fine and the problem

463
00:15:55,920 --> 00:15:58,740
is that both neither of those are

464
00:15:57,420 --> 00:16:01,139
naturally incentivized to strengthen the

465
00:15:58,740 --> 00:16:02,519
system against abuse both of them are

466
00:16:01,139 --> 00:16:04,199
you know naturally incentivized to have

467
00:16:02,519 --> 00:16:06,480
a weak system so you can actually arrest

468
00:16:04,199 --> 00:16:08,220
more and and find more

469
00:16:06,480 --> 00:16:11,699
so perverse incentives are a problem but

470
00:16:08,220 --> 00:16:14,579
number six is Trust of course government

471
00:16:11,699 --> 00:16:16,920
is a um is not something that should ask

472
00:16:14,579 --> 00:16:18,779
for trust that needs to earn trust and

473
00:16:16,920 --> 00:16:20,880
there's all of that context where people

474
00:16:18,779 --> 00:16:22,440
are saying well you know why people have

475
00:16:20,880 --> 00:16:25,079
happy to share data with the social

476
00:16:22,440 --> 00:16:27,019
media or or um you know big Tech or

477
00:16:25,079 --> 00:16:29,339
whatever and the simple answer

478
00:16:27,019 --> 00:16:30,839
realistically is because those entities

479
00:16:29,339 --> 00:16:33,740
don't have the state Monopoly on

480
00:16:30,839 --> 00:16:35,940
violence like it's fairly simple answer

481
00:16:33,740 --> 00:16:37,139
they can't come and arrest you or take

482
00:16:35,940 --> 00:16:39,839
your children they can do some pretty

483
00:16:37,139 --> 00:16:42,420
awful things but they don't directly

484
00:16:39,839 --> 00:16:43,620
have the powers and stressors the

485
00:16:42,420 --> 00:16:46,019
government does

486
00:16:43,620 --> 00:16:47,820
so and when you lose trust you lose

487
00:16:46,019 --> 00:16:49,320
stability in the whole society so it's a

488
00:16:47,820 --> 00:16:51,779
really critical problem to try and solve

489
00:16:49,320 --> 00:16:53,100
and finally policy impact so I'm

490
00:16:51,779 --> 00:16:54,660
actually helping kick off a research

491
00:16:53,100 --> 00:16:56,160
project at the moment to look at what

492
00:16:54,660 --> 00:16:58,620
the unintended human impacts of public

493
00:16:56,160 --> 00:17:00,899
policy are starting with Robo debt but

494
00:16:58,620 --> 00:17:03,180
also looking at working parents and all

495
00:17:00,899 --> 00:17:05,640
kinds of other social policies to try to

496
00:17:03,180 --> 00:17:07,199
actually say see you can measure and see

497
00:17:05,640 --> 00:17:08,640
when there's an adverse impact on a

498
00:17:07,199 --> 00:17:11,880
community and you should be able to

499
00:17:08,640 --> 00:17:14,579
understand that flag it and escalate it

500
00:17:11,880 --> 00:17:16,199
and mitigate it without having to wait

501
00:17:14,579 --> 00:17:17,579
for a commission

502
00:17:16,199 --> 00:17:19,260
somebody jump in rules this code then

503
00:17:17,579 --> 00:17:21,120
I'm going to give you some examples

504
00:17:19,260 --> 00:17:22,199
what is what sort of rules are we

505
00:17:21,120 --> 00:17:22,799
talking about

506
00:17:22,199 --> 00:17:24,360
um

507
00:17:22,799 --> 00:17:26,280
a lot of legislation and regulation

508
00:17:24,360 --> 00:17:27,780
policy is made up of two types of things

509
00:17:26,280 --> 00:17:29,040
there's prescriptive rules and judgment

510
00:17:27,780 --> 00:17:30,900
based

511
00:17:29,040 --> 00:17:32,580
the prescriptive rules are the sort of

512
00:17:30,900 --> 00:17:33,960
things you want consistently applied so

513
00:17:32,580 --> 00:17:35,880
those are the ones that fit this really

514
00:17:33,960 --> 00:17:38,039
well so things that are eligibility

515
00:17:35,880 --> 00:17:40,380
criteria calculation information things

516
00:17:38,039 --> 00:17:41,340
that are about

517
00:17:40,380 --> 00:17:42,900
um

518
00:17:41,340 --> 00:17:45,000
um you know over ten thousand dollars

519
00:17:42,900 --> 00:17:47,400
needs this action his or this obligation

520
00:17:45,000 --> 00:17:48,720
based on this trigger then they're a

521
00:17:47,400 --> 00:17:50,520
judgment-based ones so you might meet

522
00:17:48,720 --> 00:17:52,140
all of the prescriptive rules for

523
00:17:50,520 --> 00:17:53,580
instance to get citizenship but then the

524
00:17:52,140 --> 00:17:55,860
Judgment based is that you have to pass

525
00:17:53,580 --> 00:17:57,419
a good character test which obviously

526
00:17:55,860 --> 00:18:00,780
changes over time and is is very

527
00:17:57,419 --> 00:18:02,880
judgment based so rules as code really

528
00:18:00,780 --> 00:18:04,440
fits prescriptive rules really well it

529
00:18:02,880 --> 00:18:05,640
shouldn't be used or I don't believe it

530
00:18:04,440 --> 00:18:08,220
should be used for prescriptive rules

531
00:18:05,640 --> 00:18:09,299
for judgment-based rules and that

532
00:18:08,220 --> 00:18:11,039
certainly has played out in all the

533
00:18:09,299 --> 00:18:12,299
places that we've applied it

534
00:18:11,039 --> 00:18:14,039
so it fits really well in Service

535
00:18:12,299 --> 00:18:16,620
delivery it fits really well in social

536
00:18:14,039 --> 00:18:18,299
policy and Taxation policy in a lot of

537
00:18:16,620 --> 00:18:19,559
the the quite prescriptive rules and

538
00:18:18,299 --> 00:18:21,000
regulations that you deal with all the

539
00:18:19,559 --> 00:18:22,500
time

540
00:18:21,000 --> 00:18:24,059
um I've spoken about a little bit the

541
00:18:22,500 --> 00:18:25,860
why

542
00:18:24,059 --> 00:18:27,000
um I have a couple of favorite tools I'm

543
00:18:25,860 --> 00:18:27,720
going to talk about open fiscal in a

544
00:18:27,000 --> 00:18:29,460
second but there's another good one

545
00:18:27,720 --> 00:18:30,720
called blocks that you can check out all

546
00:18:29,460 --> 00:18:33,660
of this is available

547
00:18:30,720 --> 00:18:35,880
what is not rack though this isn't a

548
00:18:33,660 --> 00:18:37,799
source of constant argument

549
00:18:35,880 --> 00:18:40,140
um there are definitely people who are

550
00:18:37,799 --> 00:18:41,820
like oh just you know just create XML

551
00:18:40,140 --> 00:18:43,320
structured content doesn't make it

552
00:18:41,820 --> 00:18:44,120
machine understandable if it's not

553
00:18:43,320 --> 00:18:47,220
machine

554
00:18:44,120 --> 00:18:48,600
executable then um and of course using

555
00:18:47,220 --> 00:18:50,880
the word executable and policy in the

556
00:18:48,600 --> 00:18:53,100
same sentence is deeply uncomfortable

557
00:18:50,880 --> 00:18:54,840
um but structured content is not rules

558
00:18:53,100 --> 00:18:57,600
as code because if a machine can't parse

559
00:18:54,840 --> 00:19:00,179
it then then it's not actually machine

560
00:18:57,600 --> 00:19:02,580
it's not rules as code it's just rules

561
00:19:00,179 --> 00:19:04,500
as structure uh it's not automated law

562
00:19:02,580 --> 00:19:05,940
so anyone who's trying to and there's

563
00:19:04,500 --> 00:19:07,799
plenty of tools that like this who are

564
00:19:05,940 --> 00:19:10,080
like oh it's all right just push your

565
00:19:07,799 --> 00:19:11,520
your rules into our system it'll just

566
00:19:10,080 --> 00:19:13,440
automatically translate them or it'll

567
00:19:11,520 --> 00:19:15,299
just interpret them

568
00:19:13,440 --> 00:19:16,620
um that's not rules as code because of

569
00:19:15,299 --> 00:19:18,720
course you've still got a black box of

570
00:19:16,620 --> 00:19:20,100
decision making it's not a traditional

571
00:19:18,720 --> 00:19:23,340
rules engine traditional rules engine

572
00:19:20,100 --> 00:19:25,380
have business logic and the rules mashed

573
00:19:23,340 --> 00:19:26,940
up which means that any change is very

574
00:19:25,380 --> 00:19:29,039
very complex as anyone that's dealt with

575
00:19:26,940 --> 00:19:30,780
rules engines knows it's not an AI it's

576
00:19:29,039 --> 00:19:33,120
not a website it's been surprising to me

577
00:19:30,780 --> 00:19:34,740
how many people don't and particularly

578
00:19:33,120 --> 00:19:36,360
when you start working with policy teams

579
00:19:34,740 --> 00:19:37,140
they don't understand technology they

580
00:19:36,360 --> 00:19:39,299
don't understand the difference between

581
00:19:37,140 --> 00:19:41,700
a website and rules as code

582
00:19:39,299 --> 00:19:42,900
um and because lots of people are trying

583
00:19:41,700 --> 00:19:44,280
to sell into this place at the moment

584
00:19:42,900 --> 00:19:45,419
into this space at the moment they're

585
00:19:44,280 --> 00:19:46,500
sort of like oh well someone just sold

586
00:19:45,419 --> 00:19:48,360
me this tool last week it's like yeah

587
00:19:46,500 --> 00:19:50,160
but it's still a black box still can't

588
00:19:48,360 --> 00:19:51,480
see the rules still can't trace the

589
00:19:50,160 --> 00:19:53,760
decision back to the actual legislation

590
00:19:51,480 --> 00:19:55,679
and see if it's actually lawful

591
00:19:53,760 --> 00:19:57,059
um so

592
00:19:55,679 --> 00:19:58,919
excuse me for a second okay all right

593
00:19:57,059 --> 00:20:00,179
I'm done okay there's lots of current

594
00:19:58,919 --> 00:20:02,400
use cases I'm going to show you some

595
00:20:00,179 --> 00:20:05,700
examples in a second rather going on but

596
00:20:02,400 --> 00:20:07,559
step one for actually codifying rules is

597
00:20:05,700 --> 00:20:10,980
just understanding it read the

598
00:20:07,559 --> 00:20:13,020
legislation do some concept modeling so

599
00:20:10,980 --> 00:20:14,820
that if and it's been fascinating with

600
00:20:13,020 --> 00:20:16,919
completely brand new regulations like

601
00:20:14,820 --> 00:20:18,299
I've been involved in drafting when you

602
00:20:16,919 --> 00:20:20,220
say look we're not even going to try to

603
00:20:18,299 --> 00:20:21,960
draft it at all yet don't start writing

604
00:20:20,220 --> 00:20:24,660
it we won't start coding anything yet

605
00:20:21,960 --> 00:20:26,280
let's just step back and do a concept

606
00:20:24,660 --> 00:20:27,780
diagram let's just look at what this is

607
00:20:26,280 --> 00:20:29,340
supposed to achieve and even just going

608
00:20:27,780 --> 00:20:30,900
through concept modeling has been able

609
00:20:29,340 --> 00:20:34,200
to flush out a whole bunch of

610
00:20:30,900 --> 00:20:36,179
assumptions use cases oh but if what if

611
00:20:34,200 --> 00:20:37,860
you know test cases that would would

612
00:20:36,179 --> 00:20:40,500
break it so that's been really really

613
00:20:37,860 --> 00:20:42,179
helpful getting into then saying well

614
00:20:40,500 --> 00:20:43,320
what would the human readable version

615
00:20:42,179 --> 00:20:44,940
look like and what might a machine

616
00:20:43,320 --> 00:20:47,220
readable version look like starting to

617
00:20:44,940 --> 00:20:48,960
draw that out gets um actually then

618
00:20:47,220 --> 00:20:51,299
really helpful so you have a what they

619
00:20:48,960 --> 00:20:53,039
call isomorphic drafting

620
00:20:51,299 --> 00:20:54,539
then there's lots of opportunities to do

621
00:20:53,039 --> 00:20:55,740
modeling and all kinds of other tools

622
00:20:54,539 --> 00:20:57,780
I'm actually going to show you a bit of

623
00:20:55,740 --> 00:20:59,340
a demo now

624
00:20:57,780 --> 00:21:00,840
so yeah read the legislation identify

625
00:20:59,340 --> 00:21:02,520
the logic create some test cases and

626
00:21:00,840 --> 00:21:04,860
then start codifying in your favorite

627
00:21:02,520 --> 00:21:05,640
tool so here are some examples I want to

628
00:21:04,860 --> 00:21:06,960
give you

629
00:21:05,640 --> 00:21:09,600
in New Zealand there's a thing called

630
00:21:06,960 --> 00:21:12,120
the rates rebate act rates rebate is

631
00:21:09,600 --> 00:21:14,160
ostensibly very simple if you are a

632
00:21:12,120 --> 00:21:16,260
owner occupier

633
00:21:14,160 --> 00:21:17,460
the the policy was built to try to

634
00:21:16,260 --> 00:21:19,320
encourage

635
00:21:17,460 --> 00:21:21,059
um particularly retirees to stay in

636
00:21:19,320 --> 00:21:22,380
their homes for longer right but no of

637
00:21:21,059 --> 00:21:23,760
course no one was measuring whether that

638
00:21:22,380 --> 00:21:26,160
policy intent was being met whole

639
00:21:23,760 --> 00:21:28,740
different point it's a really simple

640
00:21:26,160 --> 00:21:30,600
piece of legislation as you can see my

641
00:21:28,740 --> 00:21:32,280
favorite bit though is here the actual

642
00:21:30,600 --> 00:21:34,679
calculation

643
00:21:32,280 --> 00:21:35,700
all right payer who was a great payable

644
00:21:34,679 --> 00:21:38,340
residential

645
00:21:35,700 --> 00:21:39,480
is entitled to so much of the rates of

646
00:21:38,340 --> 00:21:41,220
people for that rating year in respect

647
00:21:39,480 --> 00:21:42,240
to the property as represents two-thirds

648
00:21:41,220 --> 00:21:44,460
of the amount by which those rates

649
00:21:42,240 --> 00:21:46,740
exceed 160 reduced by

650
00:21:44,460 --> 00:21:48,419
one dollar for every eight dollars

651
00:21:46,740 --> 00:21:50,400
by which the rate payers income for the

652
00:21:48,419 --> 00:21:52,020
preceding tax year exceeded that amount

653
00:21:50,400 --> 00:21:53,760
the last mentioned amount being

654
00:21:52,020 --> 00:21:54,659
increased by 500 in respect to each

655
00:21:53,760 --> 00:21:56,400
person

656
00:21:54,659 --> 00:21:57,780
who's dependent of the rate payer the

657
00:21:56,400 --> 00:21:59,880
comments on the right thing okay so you

658
00:21:57,780 --> 00:22:02,400
can see even simple regulation gets

659
00:21:59,880 --> 00:22:05,400
quite interestingly annoying

660
00:22:02,400 --> 00:22:07,140
that however because I'm a maths nerd as

661
00:22:05,400 --> 00:22:09,059
well just turned out was a quadratic

662
00:22:07,140 --> 00:22:12,780
equation which is cool so I was able to

663
00:22:09,059 --> 00:22:15,559
just codify that fairly simple sorry

664
00:22:12,780 --> 00:22:15,559
no

665
00:22:15,840 --> 00:22:19,500
um so what we did with that was the

666
00:22:18,600 --> 00:22:20,280
actual

667
00:22:19,500 --> 00:22:22,200
um

668
00:22:20,280 --> 00:22:24,419
coded version we used a tool called open

669
00:22:22,200 --> 00:22:27,059
fisca I love open fisca it is

670
00:22:24,419 --> 00:22:29,159
complicated to get to know but the

671
00:22:27,059 --> 00:22:31,200
reason I like it so much is it doesn't

672
00:22:29,159 --> 00:22:34,440
just provide rules as code for services

673
00:22:31,200 --> 00:22:36,659
it provides the basis of being able to

674
00:22:34,440 --> 00:22:38,280
do a whole of population modeling of

675
00:22:36,659 --> 00:22:40,860
change which we've done now in three

676
00:22:38,280 --> 00:22:42,360
countries where here's a policy if I

677
00:22:40,860 --> 00:22:43,860
make this change to the policy or to the

678
00:22:42,360 --> 00:22:45,960
amount or to the variables or to the

679
00:22:43,860 --> 00:22:47,520
eligibility or to the rates

680
00:22:45,960 --> 00:22:49,200
what impact will that have on the whole

681
00:22:47,520 --> 00:22:51,780
population we've been it's been such a

682
00:22:49,200 --> 00:22:54,179
good tool for exactly that purpose so it

683
00:22:51,780 --> 00:22:56,340
sort of breaks it all down each one of

684
00:22:54,179 --> 00:22:57,480
these is a variable which is then you

685
00:22:56,340 --> 00:23:00,120
know broken down in each of the

686
00:22:57,480 --> 00:23:03,059
variables and the act of that actual

687
00:23:00,120 --> 00:23:05,240
calculation then is right here rates

688
00:23:03,059 --> 00:23:09,240
total minus initial contribution

689
00:23:05,240 --> 00:23:11,520
minus divided by divided by 3 times by 8

690
00:23:09,240 --> 00:23:15,240
plus allowable income yay

691
00:23:11,520 --> 00:23:17,039
um what that then so the sorry I

692
00:23:15,240 --> 00:23:18,840
shouldn't yep no I'm going to run out of

693
00:23:17,039 --> 00:23:20,580
time if I keep going so

694
00:23:18,840 --> 00:23:22,620
um each one of the variables is defined

695
00:23:20,580 --> 00:23:24,600
this is another key thing that I love

696
00:23:22,620 --> 00:23:26,159
about this particular tool is you can

697
00:23:24,600 --> 00:23:29,280
have all of the years

698
00:23:26,159 --> 00:23:31,500
so you can um rather than just having

699
00:23:29,280 --> 00:23:33,240
the current legislation you can keep the

700
00:23:31,500 --> 00:23:35,580
rates over the years you can keep the

701
00:23:33,240 --> 00:23:38,280
history of the change to the regulation

702
00:23:35,580 --> 00:23:41,400
or to the legislation so you can see oh

703
00:23:38,280 --> 00:23:42,840
look at how people are affected today I

704
00:23:41,400 --> 00:23:45,299
can run the same demographics against

705
00:23:42,840 --> 00:23:47,220
people 20 years ago and I can see the

706
00:23:45,299 --> 00:23:48,539
impact look I can now see trends of

707
00:23:47,220 --> 00:23:51,059
who's getting more and less money over

708
00:23:48,539 --> 00:23:52,980
time I can now see Trends about

709
00:23:51,059 --> 00:23:54,780
um who's being affected the size of the

710
00:23:52,980 --> 00:23:57,299
cohort the affected by a policy based on

711
00:23:54,780 --> 00:24:00,179
the eligibility criteria Etc so this

712
00:23:57,299 --> 00:24:02,960
ability to actually maintain history is

713
00:24:00,179 --> 00:24:02,960
actually quite helpful

714
00:24:03,059 --> 00:24:06,840
um and then uh then you build test

715
00:24:05,159 --> 00:24:08,220
Suites this has been the most helpful

716
00:24:06,840 --> 00:24:10,440
thing for engaging with policy people

717
00:24:08,220 --> 00:24:12,419
because if we can say wouldn't give them

718
00:24:10,440 --> 00:24:15,600
this exact thing but if we can say look

719
00:24:12,419 --> 00:24:17,159
here's an input here's the output that

720
00:24:15,600 --> 00:24:18,960
we think we should get is this right

721
00:24:17,159 --> 00:24:20,700
they go and then do their manual

722
00:24:18,960 --> 00:24:22,860
calculation they say okay yes we can

723
00:24:20,700 --> 00:24:25,860
assure that these 10 test cases are

724
00:24:22,860 --> 00:24:28,020
right great so now every time we run the

725
00:24:25,860 --> 00:24:30,240
code and we show them look your 10 test

726
00:24:28,020 --> 00:24:31,559
cases did come out as true are you

727
00:24:30,240 --> 00:24:33,120
comfortable and they're like oh actually

728
00:24:31,559 --> 00:24:35,880
yeah that makes me feel a bit more

729
00:24:33,120 --> 00:24:37,200
comfortable so test it really helps to

730
00:24:35,880 --> 00:24:38,700
get the policy people engaged but it

731
00:24:37,200 --> 00:24:40,440
also helps to flush out with sometimes

732
00:24:38,700 --> 00:24:42,059
where their assumptions are not right

733
00:24:40,440 --> 00:24:43,440
that's then used by a bunch of different

734
00:24:42,059 --> 00:24:46,440
services including this little rates

735
00:24:43,440 --> 00:24:48,179
calculator so you can say how much rates

736
00:24:46,440 --> 00:24:50,539
anyone from New Zealand want to yell at

737
00:24:48,179 --> 00:24:50,539
their rates

738
00:24:50,820 --> 00:24:57,140
cool 3 400 a year perfect how many

739
00:24:53,580 --> 00:24:57,140
dependents same person

740
00:24:57,240 --> 00:25:03,299
same person zero perfect thank you next

741
00:25:00,720 --> 00:25:06,120
question so this has already done a call

742
00:25:03,299 --> 00:25:07,140
to our open fiscal backend because one

743
00:25:06,120 --> 00:25:08,520
of the things we found is a lot of

744
00:25:07,140 --> 00:25:11,100
superannuates don't know how much they

745
00:25:08,520 --> 00:25:12,720
earn exactly and it's actually quite

746
00:25:11,100 --> 00:25:14,580
um undignified for them to engage in the

747
00:25:12,720 --> 00:25:16,140
process so being able to say look if you

748
00:25:14,580 --> 00:25:18,840
earn less than this amount you're going

749
00:25:16,140 --> 00:25:20,400
to get the full amount perfect okay you

750
00:25:18,840 --> 00:25:22,140
get the full estimated amount if you

751
00:25:20,400 --> 00:25:23,220
earn more than this amount you're going

752
00:25:22,140 --> 00:25:24,360
to get zero because we had a lot of

753
00:25:23,220 --> 00:25:25,980
people who were going through all the

754
00:25:24,360 --> 00:25:28,320
pain of applying for this particular

755
00:25:25,980 --> 00:25:29,640
benefit of this particular rebate to

756
00:25:28,320 --> 00:25:31,559
then find out they were eligible for a

757
00:25:29,640 --> 00:25:32,820
dollar fifty kind of thing or somewhere

758
00:25:31,559 --> 00:25:34,980
in the middle

759
00:25:32,820 --> 00:25:36,659
um I won't I won't throw you into it

760
00:25:34,980 --> 00:25:39,299
I'll just oops I'll just pick a number

761
00:25:36,659 --> 00:25:41,460
somewhere between there 43 000.

762
00:25:39,299 --> 00:25:43,559
and and then it says okay that's how

763
00:25:41,460 --> 00:25:44,940
much you will likely get so this

764
00:25:43,559 --> 00:25:48,000
actually just having that calculator

765
00:25:44,940 --> 00:25:50,460
available meant that thousands thousands

766
00:25:48,000 --> 00:25:52,080
of retirees who were it was such a

767
00:25:50,460 --> 00:25:53,520
stressful

768
00:25:52,080 --> 00:25:55,559
um process and they were getting almost

769
00:25:53,520 --> 00:25:57,659
nothing out of the process um could sort

770
00:25:55,559 --> 00:25:59,220
of say oh a this is worth applying for

771
00:25:57,659 --> 00:26:00,299
or it's not and suddenly they could make

772
00:25:59,220 --> 00:26:02,700
a choice in their life this is about

773
00:26:00,299 --> 00:26:03,779
trying to reverse that helpfulness of

774
00:26:02,700 --> 00:26:05,220
government rather than we're going to

775
00:26:03,779 --> 00:26:06,419
make it super hard and if you finally

776
00:26:05,220 --> 00:26:08,039
jump through our Hoops you might get

777
00:26:06,419 --> 00:26:10,200
something it's more about how can I help

778
00:26:08,039 --> 00:26:13,140
you figure out everything that you are

779
00:26:10,200 --> 00:26:14,640
entitled to so as a quick follow-up to

780
00:26:13,140 --> 00:26:16,860
that given I'm almost running out of

781
00:26:14,640 --> 00:26:19,380
time a little project in New Zealand

782
00:26:16,860 --> 00:26:20,880
that a few of us recently did we took a

783
00:26:19,380 --> 00:26:22,260
bunch of benefits and we said well why

784
00:26:20,880 --> 00:26:24,000
don't we just encode all of them and all

785
00:26:22,260 --> 00:26:24,840
of these are available in open Fisker as

786
00:26:24,000 --> 00:26:26,520
well

787
00:26:24,840 --> 00:26:30,419
and we said okay

788
00:26:26,520 --> 00:26:31,860
um I am this I am 24 year 23 years old I

789
00:26:30,419 --> 00:26:34,020
am currently without work I'm this this

790
00:26:31,860 --> 00:26:35,510
is this I have let's say none for the

791
00:26:34,020 --> 00:26:38,980
situation

792
00:26:35,510 --> 00:26:38,980
[Music]

793
00:26:39,799 --> 00:26:46,080
okay that'll do ah I forget postcodes in

794
00:26:43,380 --> 00:26:49,520
New Zealand someone help me out

795
00:26:46,080 --> 00:26:49,520
58 19.

796
00:26:51,380 --> 00:26:56,640
that's all right so long as it's a legit

797
00:26:54,360 --> 00:26:58,380
one then that'll be fine so being able

798
00:26:56,640 --> 00:27:00,600
to then say okay what you've told us and

799
00:26:58,380 --> 00:27:02,520
what you're eligible for and and roughly

800
00:27:00,600 --> 00:27:04,620
how much that's worth for you

801
00:27:02,520 --> 00:27:06,480
these and particularly the accommodation

802
00:27:04,620 --> 00:27:07,380
supplement it is almost impossible to

803
00:27:06,480 --> 00:27:09,000
figure out what your accommodation

804
00:27:07,380 --> 00:27:11,400
supplement is in New Zealand it is

805
00:27:09,000 --> 00:27:12,779
because it's based on where you live on

806
00:27:11,400 --> 00:27:13,740
your amount on your dependence and a

807
00:27:12,779 --> 00:27:15,480
whole bunch of stuff that people find

808
00:27:13,740 --> 00:27:17,100
very hard to figure out so this is

809
00:27:15,480 --> 00:27:18,900
already being used a little project done

810
00:27:17,100 --> 00:27:21,659
by a bunch of just a little citizen

811
00:27:18,900 --> 00:27:23,520
group a bunch of um largely volunteers

812
00:27:21,659 --> 00:27:24,480
working with citizen advice Bureau to

813
00:27:23,520 --> 00:27:25,919
build something so people could just

814
00:27:24,480 --> 00:27:28,020
figure out what they're entitled to and

815
00:27:25,919 --> 00:27:30,000
very cheekily

816
00:27:28,020 --> 00:27:31,500
how can I question a decision to help

817
00:27:30,000 --> 00:27:33,120
people say ah and here's an email

818
00:27:31,500 --> 00:27:35,480
template if you want to challenge a

819
00:27:33,120 --> 00:27:35,480
decision

820
00:27:35,610 --> 00:27:43,679
[Applause]

821
00:27:41,220 --> 00:27:45,600
yeah I'm not planting grenades at all um

822
00:27:43,679 --> 00:27:48,120
so

823
00:27:45,600 --> 00:27:49,440
um there's lots of examples I think

824
00:27:48,120 --> 00:27:51,059
you're already starting to get a bit of

825
00:27:49,440 --> 00:27:53,340
a feel for this in Canada we did a

826
00:27:51,059 --> 00:27:55,320
policy Difference Engine where we were

827
00:27:53,340 --> 00:27:57,360
able to show if you made this change to

828
00:27:55,320 --> 00:27:58,679
this regulation here are all the people

829
00:27:57,360 --> 00:28:00,900
that will do better and all the people

830
00:27:58,679 --> 00:28:03,480
that will do worse

831
00:28:00,900 --> 00:28:05,039
it's so important to be able to make

832
00:28:03,480 --> 00:28:07,020
those kinds of calls and make those cut

833
00:28:05,039 --> 00:28:08,880
and help make those kinds of informed

834
00:28:07,020 --> 00:28:11,279
decisions so and there's lots and lots

835
00:28:08,880 --> 00:28:12,779
of papers I've got them all linked so to

836
00:28:11,279 --> 00:28:13,970
jump back to my slides briefly and then

837
00:28:12,779 --> 00:28:14,580
we'll do some questions

838
00:28:13,970 --> 00:28:15,779
[Music]

839
00:28:14,580 --> 00:28:18,480
um

840
00:28:15,779 --> 00:28:20,580
what you end up with is better rules as

841
00:28:18,480 --> 00:28:22,620
a process this idea of isomorphically

842
00:28:20,580 --> 00:28:24,539
drafting taking a test driven approach

843
00:28:22,620 --> 00:28:26,940
to drafting new policy new regulation

844
00:28:24,539 --> 00:28:29,340
and new legislation specifically

845
00:28:26,940 --> 00:28:32,240
then being able to provide a reference

846
00:28:29,340 --> 00:28:33,960
implementation publicly API dot

847
00:28:32,240 --> 00:28:36,779
legislation.gov.eu dot New South Wales

848
00:28:33,960 --> 00:28:38,640
dot Etc is the dream because anything

849
00:28:36,779 --> 00:28:40,440
that is less than an API is probably not

850
00:28:38,640 --> 00:28:42,480
sufficient

851
00:28:40,440 --> 00:28:44,039
um the whole point is to draw a line in

852
00:28:42,480 --> 00:28:45,600
the sand so that we can start getting

853
00:28:44,039 --> 00:28:47,100
better rules so that we can start

854
00:28:45,600 --> 00:28:49,200
holding up all of our legislation and

855
00:28:47,100 --> 00:28:52,919
regulation in a way which is consumable

856
00:28:49,200 --> 00:28:55,320
actionable appealable understandable by

857
00:28:52,919 --> 00:28:58,559
people more broadly

858
00:28:55,320 --> 00:29:00,720
um and I'm gonna jump past painful

859
00:28:58,559 --> 00:29:01,980
principles but this is a key one as well

860
00:29:00,720 --> 00:29:03,720
AI

861
00:29:01,980 --> 00:29:07,080
there is of course a lot of automation

862
00:29:03,720 --> 00:29:08,760
happening in government and it is

863
00:29:07,080 --> 00:29:09,960
terrifying

864
00:29:08,760 --> 00:29:12,000
um even as a person who works in

865
00:29:09,960 --> 00:29:14,039
government almost all the time it is it

866
00:29:12,000 --> 00:29:16,740
is terrifying so for me digital

867
00:29:14,039 --> 00:29:18,240
legislation and regulation is I

868
00:29:16,740 --> 00:29:20,640
genuinely believe a precondition for

869
00:29:18,240 --> 00:29:22,799
lawful AI because either you have a

870
00:29:20,640 --> 00:29:24,360
system where your software is consuming

871
00:29:22,799 --> 00:29:25,440
the rules which would be good because

872
00:29:24,360 --> 00:29:27,840
there's a whole bunch of you know

873
00:29:25,440 --> 00:29:29,279
rules-based AI That's 30 years old that

874
00:29:27,840 --> 00:29:31,260
works perfectly well it doesn't have to

875
00:29:29,279 --> 00:29:32,820
be a machine learning based AI but even

876
00:29:31,260 --> 00:29:35,399
if you use machine learning based being

877
00:29:32,820 --> 00:29:37,740
able to test the output and the input of

878
00:29:35,399 --> 00:29:40,500
your system against the law is really

879
00:29:37,740 --> 00:29:42,059
important so that it's you know lawful

880
00:29:40,500 --> 00:29:43,580
um the other thing is so I've written

881
00:29:42,059 --> 00:29:47,399
the whole paper about how to create

882
00:29:43,580 --> 00:29:49,980
trustworthy and legitimate use of AI in

883
00:29:47,399 --> 00:29:51,960
government particularly around ADM and

884
00:29:49,980 --> 00:29:53,399
um and the other part of that is rule of

885
00:29:51,960 --> 00:29:55,080
law like people don't think about

886
00:29:53,399 --> 00:29:57,000
administrative law enough which should

887
00:29:55,080 --> 00:29:58,320
explain your decision but rule of law

888
00:29:57,000 --> 00:30:01,260
also holds that you should get a

889
00:29:58,320 --> 00:30:03,440
consistent outcome you know all people

890
00:30:01,260 --> 00:30:05,520
should get the same consistent outcome

891
00:30:03,440 --> 00:30:06,779
and if you've got a machine just

892
00:30:05,520 --> 00:30:08,700
changing the output over time because

893
00:30:06,779 --> 00:30:10,919
it's low learning but it's not a person

894
00:30:08,700 --> 00:30:13,080
accountable for that shift then that can

895
00:30:10,919 --> 00:30:14,700
also get quite dangerous as well so

896
00:30:13,080 --> 00:30:15,960
Human Services need human measures human

897
00:30:14,700 --> 00:30:17,640
Futures and human participation

898
00:30:15,960 --> 00:30:20,159
participatory governance I've given you

899
00:30:17,640 --> 00:30:21,240
the Taiwan example but there's also a

900
00:30:20,159 --> 00:30:23,039
huge amount of opportunity around

901
00:30:21,240 --> 00:30:24,480
drawing on indigenous knowledge systems

902
00:30:23,039 --> 00:30:26,340
so that we're not just planning for the

903
00:30:24,480 --> 00:30:27,539
next financial quarter or the next

904
00:30:26,340 --> 00:30:29,340
electoral cycle but we're actually

905
00:30:27,539 --> 00:30:31,320
trying to plan 50 and 100 years out so

906
00:30:29,340 --> 00:30:34,380
we get values based Futures we're

907
00:30:31,320 --> 00:30:36,419
working towards and the last I'll say is

908
00:30:34,380 --> 00:30:37,799
my epiphanies of over this time seven

909
00:30:36,419 --> 00:30:39,480
years of epiphanies boiled down to one

910
00:30:37,799 --> 00:30:41,279
slide here you go

911
00:30:39,480 --> 00:30:43,320
rules and business logic need to have an

912
00:30:41,279 --> 00:30:44,760
amical separation

913
00:30:43,320 --> 00:30:45,840
every one of you that's worked with

914
00:30:44,760 --> 00:30:47,220
rules engine how many people have worked

915
00:30:45,840 --> 00:30:49,500
with rules engines

916
00:30:47,220 --> 00:30:53,279
not many some of you've been spared that

917
00:30:49,500 --> 00:30:55,020
pain so rules engines are so hard

918
00:30:53,279 --> 00:30:55,679
because someone over here says here's

919
00:30:55,020 --> 00:30:57,899
all the rules in the business

920
00:30:55,679 --> 00:30:59,940
requirements and the business analyst

921
00:30:57,899 --> 00:31:02,279
goes ah here's what it all kind of means

922
00:30:59,940 --> 00:31:03,659
and then someone codes it all up and

923
00:31:02,279 --> 00:31:05,100
then there's a change

924
00:31:03,659 --> 00:31:07,440
and then you keep having to jump through

925
00:31:05,100 --> 00:31:09,720
the cycle of the change coming out

926
00:31:07,440 --> 00:31:11,399
having to refactor all of your business

927
00:31:09,720 --> 00:31:13,919
logic and then re-implement into your

928
00:31:11,399 --> 00:31:15,120
rules if you actually separate your

929
00:31:13,919 --> 00:31:17,820
business logic and keep the business

930
00:31:15,120 --> 00:31:19,740
logic in your applications and just have

931
00:31:17,820 --> 00:31:22,140
the rules as good just rules as code

932
00:31:19,740 --> 00:31:23,279
then just the legislationist code that

933
00:31:22,140 --> 00:31:25,320
you can consume by all these different

934
00:31:23,279 --> 00:31:26,820
applications then suddenly you have a

935
00:31:25,320 --> 00:31:28,919
lot more consistency of How It's applied

936
00:31:26,820 --> 00:31:30,720
but also easy of Mana ease of management

937
00:31:28,919 --> 00:31:32,520
when change happens

938
00:31:30,720 --> 00:31:33,720
I think legislation and regulation is

939
00:31:32,520 --> 00:31:36,539
part of the digital public

940
00:31:33,720 --> 00:31:38,700
infrastructure that we need there is a

941
00:31:36,539 --> 00:31:40,740
lot particularly with any you know

942
00:31:38,700 --> 00:31:42,179
emerging Trend a lot of snake oil sales

943
00:31:40,740 --> 00:31:43,679
people so worth keeping an eye out for

944
00:31:42,179 --> 00:31:45,539
and a lot of people who are not snake or

945
00:31:43,679 --> 00:31:47,640
sales people they're just driven by

946
00:31:45,539 --> 00:31:50,520
their own thing there's a lot of natural

947
00:31:47,640 --> 00:31:52,140
learning people who love this space

948
00:31:50,520 --> 00:31:54,480
because they get to apply all their

949
00:31:52,140 --> 00:31:55,860
skills but if you just build a tool that

950
00:31:54,480 --> 00:31:58,679
just keeps

951
00:31:55,860 --> 00:31:59,820
um parsing badly drafted regulation

952
00:31:58,679 --> 00:32:01,080
legislation in the first place you're

953
00:31:59,820 --> 00:32:02,760
actually propping up a system that

954
00:32:01,080 --> 00:32:04,679
itself needs to change

955
00:32:02,760 --> 00:32:07,080
there is a strong Coalition of the

956
00:32:04,679 --> 00:32:08,700
Willing globally worth getting involved

957
00:32:07,080 --> 00:32:09,899
in and anything short of API that

958
00:32:08,700 --> 00:32:11,760
legislation the government is not enough

959
00:32:09,899 --> 00:32:13,799
people who are like oh technology

960
00:32:11,760 --> 00:32:15,720
doesn't really matter you just need to

961
00:32:13,799 --> 00:32:16,500
have a good you know

962
00:32:15,720 --> 00:32:18,360
um

963
00:32:16,500 --> 00:32:21,059
um you just need to have a good concept

964
00:32:18,360 --> 00:32:22,919
model architecture

965
00:32:21,059 --> 00:32:24,720
so I've got a whole bunch of links here

966
00:32:22,919 --> 00:32:27,179
A whole bunch of materials whole bunch

967
00:32:24,720 --> 00:32:28,559
of case studies I'm going to stop and

968
00:32:27,179 --> 00:32:32,600
take some questions I hope that's been

969
00:32:28,559 --> 00:32:32,600
informative and useful thank you

970
00:32:38,159 --> 00:32:41,940
thank you so much I'm sorry if I rushed

971
00:32:40,320 --> 00:32:44,279
you to a towards me and that was just

972
00:32:41,940 --> 00:32:47,220
absolutely fantastic do we have any

973
00:32:44,279 --> 00:32:49,200
questions to start us off

974
00:32:47,220 --> 00:32:51,600
results

975
00:32:49,200 --> 00:32:52,980
there's one down here I think Russell

976
00:32:51,600 --> 00:32:54,360
are you heading towards anyone Russell

977
00:32:52,980 --> 00:32:56,340
or have we got one here

978
00:32:54,360 --> 00:32:57,659
it's all right I can

979
00:32:56,340 --> 00:32:59,700
I've got two actually we'll start with

980
00:32:57,659 --> 00:33:02,100
you

981
00:32:59,700 --> 00:33:04,140
um fantastic talk I'm trying to work out

982
00:33:02,100 --> 00:33:06,419
how do you engage governments that

983
00:33:04,140 --> 00:33:07,980
aren't already on the bandwagon like

984
00:33:06,419 --> 00:33:09,240
you've shown some examples with the

985
00:33:07,980 --> 00:33:11,220
rates calculator and whatnot there's

986
00:33:09,240 --> 00:33:13,740
evidently some engagement with

987
00:33:11,220 --> 00:33:16,559
government to actually use this process

988
00:33:13,740 --> 00:33:18,480
to get there so how how do you bootstrap

989
00:33:16,559 --> 00:33:21,179
that like if you are interested in

990
00:33:18,480 --> 00:33:24,000
having better government how do you get

991
00:33:21,179 --> 00:33:26,279
them to do this cool so I'll give you

992
00:33:24,000 --> 00:33:26,880
three strategies just real quick

993
00:33:26,279 --> 00:33:29,880
um

994
00:33:26,880 --> 00:33:31,860
because policy and delivery are

995
00:33:29,880 --> 00:33:33,059
physically separated between a lot of

996
00:33:31,860 --> 00:33:36,120
departments if you're dealing with an

997
00:33:33,059 --> 00:33:38,100
implementation agency rules is code is a

998
00:33:36,120 --> 00:33:39,960
great way for them to save money so for

999
00:33:38,100 --> 00:33:42,720
an implementation agency

1000
00:33:39,960 --> 00:33:44,960
the cost of change for traditional rules

1001
00:33:42,720 --> 00:33:47,760
models is rules engines is

1002
00:33:44,960 --> 00:33:49,679
extraordinarily high they feel the pain

1003
00:33:47,760 --> 00:33:52,019
of poorly drafted policy they feel the

1004
00:33:49,679 --> 00:33:53,880
pain of when you said this what did you

1005
00:33:52,019 --> 00:33:55,620
mean because by the way how I interpret

1006
00:33:53,880 --> 00:33:57,059
that could affect thousands or hundreds

1007
00:33:55,620 --> 00:33:59,120
of thousands of people right so for

1008
00:33:57,059 --> 00:34:01,320
implementation agencies saying

1009
00:33:59,120 --> 00:34:03,899
regardless of how it's drafted you could

1010
00:34:01,320 --> 00:34:06,179
run your legislation as code as a way to

1011
00:34:03,899 --> 00:34:07,860
actually minimize your own complexity to

1012
00:34:06,179 --> 00:34:10,260
simplify your own Service delivery to

1013
00:34:07,860 --> 00:34:11,639
improve change management and to

1014
00:34:10,260 --> 00:34:12,899
actually test the outcomes of your

1015
00:34:11,639 --> 00:34:14,460
systems against the law to make sure

1016
00:34:12,899 --> 00:34:16,980
it's lawful before it you know pumps out

1017
00:34:14,460 --> 00:34:18,300
to a person for a policy agency they're

1018
00:34:16,980 --> 00:34:19,639
only starting to realize that they're

1019
00:34:18,300 --> 00:34:21,899
part of the problem

1020
00:34:19,639 --> 00:34:23,220
so it's a different type of conversation

1021
00:34:21,899 --> 00:34:26,099
but that conversation is already

1022
00:34:23,220 --> 00:34:27,480
starting to happen my personal hack that

1023
00:34:26,099 --> 00:34:30,240
I'm going to try to do this year and I

1024
00:34:27,480 --> 00:34:31,679
mean that in the Civic hack way is I'm

1025
00:34:30,240 --> 00:34:33,119
working with the person who's one of the

1026
00:34:31,679 --> 00:34:34,500
authors of the Australian policy

1027
00:34:33,119 --> 00:34:36,659
handbook and the Australian policy cycle

1028
00:34:34,500 --> 00:34:38,820
to try to rewrite that for instance to

1029
00:34:36,659 --> 00:34:41,099
actually make this a normal process

1030
00:34:38,820 --> 00:34:43,080
um so working with policy agencies and

1031
00:34:41,099 --> 00:34:44,339
showing them look if you isomorphically

1032
00:34:43,080 --> 00:34:45,960
draft you can take a test driven

1033
00:34:44,339 --> 00:34:47,520
approach why not have a couple of people

1034
00:34:45,960 --> 00:34:49,080
that have consume this Rule and what I

1035
00:34:47,520 --> 00:34:51,480
found is every policy team that has

1036
00:34:49,080 --> 00:34:52,740
experienced having someone in the room

1037
00:34:51,480 --> 00:34:54,300
that's going to naturally consume the

1038
00:34:52,740 --> 00:34:56,399
rule and then getting a better policy

1039
00:34:54,300 --> 00:34:58,140
getting something which is implemented

1040
00:34:56,399 --> 00:34:59,580
closer to what they imagined in the

1041
00:34:58,140 --> 00:35:02,040
first place every time they experience

1042
00:34:59,580 --> 00:35:03,060
that they they want to change so that's

1043
00:35:02,040 --> 00:35:05,580
how to deal with implementation that's

1044
00:35:03,060 --> 00:35:07,080
how I do with policy agencies with the

1045
00:35:05,580 --> 00:35:08,220
Parliamentary Council Office which is

1046
00:35:07,080 --> 00:35:11,160
the key one to change the problem is

1047
00:35:08,220 --> 00:35:13,980
they don't have any Tech capability

1048
00:35:11,160 --> 00:35:15,900
so a little call out we need great

1049
00:35:13,980 --> 00:35:17,160
techie people with Innovative minds and

1050
00:35:15,900 --> 00:35:20,520
open source mentalities to work in

1051
00:35:17,160 --> 00:35:22,260
government come play please

1052
00:35:20,520 --> 00:35:25,260
that will be my third strategy come come

1053
00:35:22,260 --> 00:35:27,920
and join the workforce and help

1054
00:35:25,260 --> 00:35:27,920
thank you

1055
00:35:28,920 --> 00:35:32,460
cool thank you once again g'day

1056
00:35:31,680 --> 00:35:35,060
um

1057
00:35:32,460 --> 00:35:37,980
Dr Andrew Lee

1058
00:35:35,060 --> 00:35:40,800
member for fenno in Canberra

1059
00:35:37,980 --> 00:35:43,320
um really likes randomized controlled

1060
00:35:40,800 --> 00:35:47,339
trials yeah and I'm wondering where that

1061
00:35:43,320 --> 00:35:49,160
fits in are they the tests are they the

1062
00:35:47,339 --> 00:35:52,380
the the

1063
00:35:49,160 --> 00:35:55,140
feedback that go to policy makers is

1064
00:35:52,380 --> 00:35:57,540
that part of the plan so the when you

1065
00:35:55,140 --> 00:36:00,240
take a better rules process in the first

1066
00:35:57,540 --> 00:36:02,220
instance and there are some agencies who

1067
00:36:00,240 --> 00:36:05,579
have started doing more let's call it

1068
00:36:02,220 --> 00:36:06,720
design LED approach to policy

1069
00:36:05,579 --> 00:36:08,579
um so

1070
00:36:06,720 --> 00:36:10,619
people that just want to do a b testing

1071
00:36:08,579 --> 00:36:13,260
the challenge is that you may you can

1072
00:36:10,619 --> 00:36:14,880
maybe test your way to a spoon from a

1073
00:36:13,260 --> 00:36:16,800
rock over a thousand iterations right

1074
00:36:14,880 --> 00:36:17,760
but if you actually ask a person what

1075
00:36:16,800 --> 00:36:19,200
they're trying to do and they say and

1076
00:36:17,760 --> 00:36:22,140
trying to drink some soup you might give

1077
00:36:19,200 --> 00:36:24,599
them a spoon to start so I think that a

1078
00:36:22,140 --> 00:36:25,980
b testing and randomized testing has a

1079
00:36:24,599 --> 00:36:27,420
role to play but it shouldn't be the

1080
00:36:25,980 --> 00:36:28,500
start of the process if you see what I

1081
00:36:27,420 --> 00:36:30,660
mean like the start of the process

1082
00:36:28,500 --> 00:36:31,740
should be that concept modeling that

1083
00:36:30,660 --> 00:36:32,940
what are we trying to achieve and what

1084
00:36:31,740 --> 00:36:34,500
are the different ways that we might

1085
00:36:32,940 --> 00:36:35,640
achieve that it might be a grant it

1086
00:36:34,500 --> 00:36:37,320
might be a piece of Regulation it might

1087
00:36:35,640 --> 00:36:39,780
be a service it might be tweaking

1088
00:36:37,320 --> 00:36:41,220
something that already exists but it

1089
00:36:39,780 --> 00:36:44,040
certainly should involve the people

1090
00:36:41,220 --> 00:36:46,079
affected by that policy to be part of

1091
00:36:44,040 --> 00:36:48,119
that ideation process before they even

1092
00:36:46,079 --> 00:36:49,260
commit to one way or the other as it

1093
00:36:48,119 --> 00:36:51,540
were

1094
00:36:49,260 --> 00:36:54,119
um so but yeah then actually monitoring

1095
00:36:51,540 --> 00:36:57,420
for impact is really critical the other

1096
00:36:54,119 --> 00:36:59,280
problem with um randomized testing and a

1097
00:36:57,420 --> 00:37:02,579
b testing and policy is if it's

1098
00:36:59,280 --> 00:37:05,880
something as critical as Social Services

1099
00:37:02,579 --> 00:37:09,060
what happens to the bees as it were so

1100
00:37:05,880 --> 00:37:10,800
so some of this has to be done in in um

1101
00:37:09,060 --> 00:37:13,160
not just collaboration with people but

1102
00:37:10,800 --> 00:37:15,300
in micro testing before you push it out

1103
00:37:13,160 --> 00:37:16,500
just because of the absolute critical

1104
00:37:15,300 --> 00:37:18,300
nature of some of some of the things

1105
00:37:16,500 --> 00:37:20,040
that government provides

1106
00:37:18,300 --> 00:37:22,220
so participatory

1107
00:37:20,040 --> 00:37:25,980
design LED from the start

1108
00:37:22,220 --> 00:37:27,180
and and then micro testing before things

1109
00:37:25,980 --> 00:37:29,820
are rolled out I think are probably

1110
00:37:27,180 --> 00:37:32,660
really key

1111
00:37:29,820 --> 00:37:32,660
well it's a few

1112
00:37:33,359 --> 00:37:37,440
thanks

1113
00:37:34,920 --> 00:37:40,140
um so I just want to start by saying I

1114
00:37:37,440 --> 00:37:42,119
love this stuff so you had me at hello

1115
00:37:40,140 --> 00:37:45,300
um but I do have a question

1116
00:37:42,119 --> 00:37:47,220
um from the inside it seems that uh the

1117
00:37:45,300 --> 00:37:48,780
the biggest barrier we have to overcome

1118
00:37:47,220 --> 00:37:50,880
in getting all of these things to come

1119
00:37:48,780 --> 00:37:53,160
together and and produce these great

1120
00:37:50,880 --> 00:37:55,200
outcomes is that that none of the

1121
00:37:53,160 --> 00:37:56,940
players trust each other and I'm

1122
00:37:55,200 --> 00:37:59,420
wondering if there are things that

1123
00:37:56,940 --> 00:38:02,780
you've seen in practice that can help

1124
00:37:59,420 --> 00:38:06,119
say the the policy

1125
00:38:02,780 --> 00:38:08,220
the policy experts and the the people

1126
00:38:06,119 --> 00:38:10,200
who have to do the implementation and

1127
00:38:08,220 --> 00:38:12,540
the people who lived experience and the

1128
00:38:10,200 --> 00:38:14,460
broader Community who aren't personally

1129
00:38:12,540 --> 00:38:16,740
going to be impacted but have a real

1130
00:38:14,460 --> 00:38:19,859
interest in seeing this work well in

1131
00:38:16,740 --> 00:38:21,720
society how do we how can we start to

1132
00:38:19,859 --> 00:38:24,480
build that trust between them so that

1133
00:38:21,720 --> 00:38:25,859
they can come to the table and co-design

1134
00:38:24,480 --> 00:38:30,180
for sure

1135
00:38:25,859 --> 00:38:33,540
um I I remember in Canada I was talking

1136
00:38:30,180 --> 00:38:35,160
at a policy I was talking to some policy

1137
00:38:33,540 --> 00:38:36,180
lead people and we were talking about

1138
00:38:35,160 --> 00:38:37,920
running an event called policy

1139
00:38:36,180 --> 00:38:40,260
transformation and they're a bit

1140
00:38:37,920 --> 00:38:41,640
bit funny about it and I and they

1141
00:38:40,260 --> 00:38:43,140
finally said well Peter I just want to

1142
00:38:41,640 --> 00:38:44,400
change the name because policy

1143
00:38:43,140 --> 00:38:46,440
transformation sounds like you actually

1144
00:38:44,400 --> 00:38:49,400
want to change how we do things

1145
00:38:46,440 --> 00:38:51,660
I'm like yes

1146
00:38:49,400 --> 00:38:52,920
and so then that led to a different sort

1147
00:38:51,660 --> 00:38:55,800
of conversation

1148
00:38:52,920 --> 00:38:56,940
um I I think that partly um I think

1149
00:38:55,800 --> 00:39:00,119
there's a couple of things first of all

1150
00:38:56,940 --> 00:39:03,599
some of this happens naturally in tiny

1151
00:39:00,119 --> 00:39:05,420
Pockets so weirdly enough my observation

1152
00:39:03,599 --> 00:39:08,460
is that almost every tax department

1153
00:39:05,420 --> 00:39:10,079
globally that I've dealt with does a

1154
00:39:08,460 --> 00:39:12,180
micro aversion of this they do have

1155
00:39:10,079 --> 00:39:14,280
policy people and implementation people

1156
00:39:12,180 --> 00:39:16,680
in the same room rapidly iterating

1157
00:39:14,280 --> 00:39:18,000
policy regularly so there's a there's a

1158
00:39:16,680 --> 00:39:19,619
micro example of that in every

1159
00:39:18,000 --> 00:39:22,800
jurisdiction that I've worked in for

1160
00:39:19,619 --> 00:39:24,960
sure and several that I've seen where

1161
00:39:22,800 --> 00:39:27,599
for whatever reason and it's probably

1162
00:39:24,960 --> 00:39:29,640
because of you know people governments

1163
00:39:27,599 --> 00:39:31,800
have adapted to be very responsive to

1164
00:39:29,640 --> 00:39:33,660
business needs just less so citizens but

1165
00:39:31,800 --> 00:39:35,099
you know but we can use the patent to

1166
00:39:33,660 --> 00:39:36,660
being able to say look we already do it

1167
00:39:35,099 --> 00:39:39,780
well here so why can't we just extend it

1168
00:39:36,660 --> 00:39:42,000
to other sorts of portfolios so

1169
00:39:39,780 --> 00:39:44,040
amplifying patterns is helpful

1170
00:39:42,000 --> 00:39:45,359
starting from the the actual outcome

1171
00:39:44,040 --> 00:39:47,160
because everyone's sort of focused on

1172
00:39:45,359 --> 00:39:48,480
they have too much work and Source too

1173
00:39:47,160 --> 00:39:49,920
stressful but if you say well hold on

1174
00:39:48,480 --> 00:39:52,079
what's the impact you're trying to have

1175
00:39:49,920 --> 00:39:54,480
what's the shared goal that we have

1176
00:39:52,079 --> 00:39:55,859
and if you can identify a shared goal

1177
00:39:54,480 --> 00:39:58,320
you have a better chance of people being

1178
00:39:55,859 --> 00:40:00,660
able to come together because if even if

1179
00:39:58,320 --> 00:40:02,700
they even across disciplines you know

1180
00:40:00,660 --> 00:40:04,560
and identifying discipline friction as a

1181
00:40:02,700 --> 00:40:06,060
thing you know if you get your data

1182
00:40:04,560 --> 00:40:08,040
scientist and your designer in the same

1183
00:40:06,060 --> 00:40:10,560
room they will kill each other

1184
00:40:08,040 --> 00:40:12,599
um unless you can say we will only get a

1185
00:40:10,560 --> 00:40:14,640
better outcome because you're both in

1186
00:40:12,599 --> 00:40:16,619
the room and by the way it's not that

1187
00:40:14,640 --> 00:40:17,579
each either of you is better or worse

1188
00:40:16,619 --> 00:40:19,260
than the other it's that you're

1189
00:40:17,579 --> 00:40:20,460
different let's identify that you have

1190
00:40:19,260 --> 00:40:23,339
different lexicon different methods

1191
00:40:20,460 --> 00:40:25,200
different like different techniques and

1192
00:40:23,339 --> 00:40:26,640
let's Embrace that diversity as part of

1193
00:40:25,200 --> 00:40:28,800
the strength of getting to where we need

1194
00:40:26,640 --> 00:40:31,380
to go identifying discipline friction as

1195
00:40:28,800 --> 00:40:34,859
a thing and building that into a culture

1196
00:40:31,380 --> 00:40:36,300
that can be multi-disciplinary by choice

1197
00:40:34,859 --> 00:40:37,800
um helps build that trust because some

1198
00:40:36,300 --> 00:40:39,119
of that trust is like from my

1199
00:40:37,800 --> 00:40:40,500
observation anyway as pure discipline

1200
00:40:39,119 --> 00:40:41,099
friction

1201
00:40:40,500 --> 00:40:42,839
um

1202
00:40:41,099 --> 00:40:46,079
and then of course actually starting to

1203
00:40:42,839 --> 00:40:48,359
shift to outcomes-based structures so 20

1204
00:40:46,079 --> 00:40:50,040
30 years ago as some of you again are

1205
00:40:48,359 --> 00:40:51,660
noticed me to know but most of you

1206
00:40:50,040 --> 00:40:52,859
probably won't about 30 years ago they

1207
00:40:51,660 --> 00:40:54,000
made a big shift to governments all

1208
00:40:52,859 --> 00:40:56,460
around the world where they decided to

1209
00:40:54,000 --> 00:40:59,760
turn them into businesses so all of the

1210
00:40:56,460 --> 00:41:01,560
departments were structurally changed to

1211
00:40:59,760 --> 00:41:03,359
have functional segmentation

1212
00:41:01,560 --> 00:41:05,099
that structural segmentation means that

1213
00:41:03,359 --> 00:41:06,720
every every discipline has a different

1214
00:41:05,099 --> 00:41:09,300
cost center and can't talk to each other

1215
00:41:06,720 --> 00:41:10,740
unless they can give them some money or

1216
00:41:09,300 --> 00:41:12,960
do some cost recovery or cite a cost

1217
00:41:10,740 --> 00:41:14,280
center if we can shift the structures of

1218
00:41:12,960 --> 00:41:16,500
our departments back to what they were

1219
00:41:14,280 --> 00:41:18,060
which was based around outcomes you need

1220
00:41:16,500 --> 00:41:19,740
to have an outcomes based cost center

1221
00:41:18,060 --> 00:41:22,619
and outcomes-based structure so that you

1222
00:41:19,740 --> 00:41:24,540
can hire and maintain the

1223
00:41:22,619 --> 00:41:26,099
cross-disciplinary set of skills and

1224
00:41:24,540 --> 00:41:28,380
experience that you need to drive that

1225
00:41:26,099 --> 00:41:31,020
outcome whereas with everyone just

1226
00:41:28,380 --> 00:41:33,060
trying to Blind Baton Pass between their

1227
00:41:31,020 --> 00:41:34,320
functions one group writing business

1228
00:41:33,060 --> 00:41:35,700
requirements the other group

1229
00:41:34,320 --> 00:41:36,780
implementing the first group saying

1230
00:41:35,700 --> 00:41:38,099
that's not what I meant well that's what

1231
00:41:36,780 --> 00:41:39,839
you wrote well I meant this well why

1232
00:41:38,099 --> 00:41:41,880
didn't you say that

1233
00:41:39,839 --> 00:41:43,200
um every time you hear the word business

1234
00:41:41,880 --> 00:41:45,380
requirements they're probably doing it

1235
00:41:43,200 --> 00:41:45,380
wrong

1236
00:41:45,560 --> 00:41:50,579
and yet so so we need to identify these

1237
00:41:48,420 --> 00:41:51,900
patterns and help people remember that

1238
00:41:50,579 --> 00:41:54,320
they've got more in common that they

1239
00:41:51,900 --> 00:41:54,320
haven't difference

1240
00:41:55,079 --> 00:41:59,579
a good approach we have one minute left

1241
00:41:57,420 --> 00:42:02,119
but I think we can fit one more sneaky

1242
00:41:59,579 --> 00:42:02,119
question and

1243
00:42:03,540 --> 00:42:07,380
just one second for a lovely microphone

1244
00:42:05,160 --> 00:42:09,900
for you thank you

1245
00:42:07,380 --> 00:42:13,980
um I'm just wondering if you could um

1246
00:42:09,900 --> 00:42:16,500
outline or give any examples of once

1247
00:42:13,980 --> 00:42:19,740
these um once this legislation or

1248
00:42:16,500 --> 00:42:22,020
regulation has been written of um either

1249
00:42:19,740 --> 00:42:24,060
citizen groups like you were showing in

1250
00:42:22,020 --> 00:42:26,700
New Zealand or other organizations

1251
00:42:24,060 --> 00:42:30,720
actually consuming them

1252
00:42:26,700 --> 00:42:32,280
of other organizations using them yes

1253
00:42:30,720 --> 00:42:34,020
um

1254
00:42:32,280 --> 00:42:36,720
so that's starting to happen all around

1255
00:42:34,020 --> 00:42:39,000
the world the um so

1256
00:42:36,720 --> 00:42:41,520
just from my experience

1257
00:42:39,000 --> 00:42:43,140
um so the the New Zealand I'm pretty

1258
00:42:41,520 --> 00:42:44,520
pleased with because

1259
00:42:43,140 --> 00:42:46,619
um there's lots of

1260
00:42:44,520 --> 00:42:48,000
um there's lots of departments starting

1261
00:42:46,619 --> 00:42:50,820
to say okay yeah actually this

1262
00:42:48,000 --> 00:42:52,500
separation of rules and um of

1263
00:42:50,820 --> 00:42:54,359
legislation and Business Systems is

1264
00:42:52,500 --> 00:42:57,000
really actually helpful to what we're

1265
00:42:54,359 --> 00:43:00,180
trying to do so it's definitely starting

1266
00:42:57,000 --> 00:43:01,619
to be adopted in government itself in

1267
00:43:00,180 --> 00:43:02,400
New Zealand the group that we pulled

1268
00:43:01,619 --> 00:43:04,260
together

1269
00:43:02,400 --> 00:43:06,780
um was sort of loosely based on the gov

1270
00:43:04,260 --> 00:43:09,540
zero lessons out of Taiwan

1271
00:43:06,780 --> 00:43:11,700
um but it is I think the world's first

1272
00:43:09,540 --> 00:43:14,640
Citizen Community repository of

1273
00:43:11,700 --> 00:43:16,980
legislation as code which is pretty cool

1274
00:43:14,640 --> 00:43:18,540
um and so it's continuing to be built

1275
00:43:16,980 --> 00:43:20,819
upon they're finding new funding to get

1276
00:43:18,540 --> 00:43:22,680
more benefits put in and the goal is to

1277
00:43:20,819 --> 00:43:25,140
actually have a community managed

1278
00:43:22,680 --> 00:43:27,480
repository which is the next call for

1279
00:43:25,140 --> 00:43:28,980
all of you I want to build a community

1280
00:43:27,480 --> 00:43:30,180
repository of legislations code here

1281
00:43:28,980 --> 00:43:32,280
that was actually the last slide which I

1282
00:43:30,180 --> 00:43:34,619
mentioned I get to that was really bad

1283
00:43:32,280 --> 00:43:37,020
um I want to build a community

1284
00:43:34,619 --> 00:43:40,020
repository here something that we can

1285
00:43:37,020 --> 00:43:42,119
use to test new regulation Test new

1286
00:43:40,020 --> 00:43:44,640
legislation something we can use to say

1287
00:43:42,119 --> 00:43:46,440
oh same as benefit me here's what the

1288
00:43:44,640 --> 00:43:48,300
law says you're eligible for and if the

1289
00:43:46,440 --> 00:43:50,040
department gives you a different answer

1290
00:43:48,300 --> 00:43:51,240
and you say well here's what the law

1291
00:43:50,040 --> 00:43:52,619
says I'm eligible for why are you giving

1292
00:43:51,240 --> 00:43:53,579
me a different answer and when they say

1293
00:43:52,619 --> 00:43:56,460
well you don't have any operational

1294
00:43:53,579 --> 00:43:57,660
policy in your tool we can say are you

1295
00:43:56,460 --> 00:43:59,520
suggesting your operational policy

1296
00:43:57,660 --> 00:44:02,099
conflicts with the law

1297
00:43:59,520 --> 00:44:04,140
question mark

1298
00:44:02,099 --> 00:44:05,819
um you know you're going to shake up the

1299
00:44:04,140 --> 00:44:07,200
power differential otherwise uh

1300
00:44:05,819 --> 00:44:09,180
otherwise it you know we all take it too

1301
00:44:07,200 --> 00:44:10,859
seriously um so I want to build one of

1302
00:44:09,180 --> 00:44:12,900
those in Australia that we can all rely

1303
00:44:10,859 --> 00:44:15,599
upon every single Think Tank every

1304
00:44:12,900 --> 00:44:17,520
single research institution Gratton here

1305
00:44:15,599 --> 00:44:19,020
in Melbourne case in point keeps

1306
00:44:17,520 --> 00:44:20,400
building out these models these

1307
00:44:19,020 --> 00:44:21,660
legislation and regulations code models

1308
00:44:20,400 --> 00:44:23,400
to do their own thing why wouldn't we

1309
00:44:21,660 --> 00:44:25,500
have a shared repository why can't we

1310
00:44:23,400 --> 00:44:27,720
build policy infrastructure for our

1311
00:44:25,500 --> 00:44:30,440
nation so that would be my suggestion

1312
00:44:27,720 --> 00:44:30,440
anyone yeah

1313
00:44:31,460 --> 00:44:34,520
thank you