1
00:00:08,790 --> 00:00:06,230
[Music]

2
00:00:12,709 --> 00:00:08,800
so let's talk about the hypothesis here

3
00:00:17,910 --> 00:00:14,230
what

4
00:00:19,670 --> 00:00:17,920
dr mosbridge found is that there's a

5
00:00:21,750 --> 00:00:19,680
correlation

6
00:00:23,109 --> 00:00:21,760
or at least there can be it's been found

7
00:00:25,029 --> 00:00:23,119
in some experiments there's a

8
00:00:29,189 --> 00:00:25,039
correlation between

9
00:00:30,950 --> 00:00:29,199
predecision and post-decision data

10
00:00:33,270 --> 00:00:30,960
in an experiment where there's a

11
00:00:35,670 --> 00:00:33,280
decision about how long that experiment

12
00:00:37,430 --> 00:00:35,680
will run so in other words the as you

13
00:00:39,190 --> 00:00:37,440
can see the little x's down here are

14
00:00:40,869 --> 00:00:39,200
data

15
00:00:42,229 --> 00:00:40,879
i can point with my mouse little x's

16
00:00:43,190 --> 00:00:42,239
down here data

17
00:00:45,430 --> 00:00:43,200
and

18
00:00:48,310 --> 00:00:45,440
each time the experiment runs a trial

19
00:00:49,510 --> 00:00:48,320
and it ran about uh 34 000 of these

20
00:00:50,630 --> 00:00:49,520
trials

21
00:00:52,869 --> 00:00:50,640
um

22
00:00:55,590 --> 00:00:52,879
each time it makes a decision where this

23
00:00:57,189 --> 00:00:55,600
vertical line is as to how long the

24
00:00:58,470 --> 00:00:57,199
remaining duration of the experiment

25
00:01:01,430 --> 00:00:58,480
will be

26
00:01:03,029 --> 00:01:01,440
and that remaining duration seems to

27
00:01:04,789 --> 00:01:03,039
affect somehow

28
00:01:07,109 --> 00:01:04,799
the data that

29
00:01:08,390 --> 00:01:07,119
either came before that decision

30
00:01:12,789 --> 00:01:08,400
or

31
00:01:16,630 --> 00:01:12,799
the decision with the data after that

32
00:01:21,590 --> 00:01:16,640
decision but still dependent on how long

33
00:01:25,749 --> 00:01:24,070
in order to make a comparison we have to

34
00:01:28,310 --> 00:01:25,759
have some data after the decision to

35
00:01:31,429 --> 00:01:28,320
compare to the data before the decision

36
00:01:32,469 --> 00:01:31,439
but the remaining duration after that

37
00:01:33,429 --> 00:01:32,479
matters

38
00:01:35,510 --> 00:01:33,439
for

39
00:01:37,109 --> 00:01:35,520
that correlation i'll get into that more

40
00:01:38,550 --> 00:01:37,119
detail and you'll see this slide again

41
00:01:42,069 --> 00:01:38,560
i'm going to bring it up two more times

42
00:01:44,149 --> 00:01:42,079
in this talk in different variations

43
00:01:46,469 --> 00:01:44,159
so let's talk about what julia

44
00:01:47,990 --> 00:01:46,479
mossbridge found

45
00:01:50,069 --> 00:01:48,000
and what her

46
00:01:52,469 --> 00:01:50,079
protocol was and

47
00:01:55,510 --> 00:01:52,479
what she means by what we mean by

48
00:01:56,469 --> 00:01:55,520
causally ambiguous duration sorting

49
00:01:58,950 --> 00:01:56,479
so

50
00:02:01,190 --> 00:01:58,960
um the experiments that she reported on

51
00:02:03,109 --> 00:02:01,200
in 2019 and 2021

52
00:02:05,910 --> 00:02:03,119
all involve these elements on the right

53
00:02:07,429 --> 00:02:05,920
and here's a plot from one of her talks

54
00:02:11,990 --> 00:02:07,439
so

55
00:02:13,830 --> 00:02:12,000
and a detector and photon counting and

56
00:02:16,070 --> 00:02:13,840
the light source and detector are both

57
00:02:19,110 --> 00:02:16,080
turned on and off

58
00:02:20,630 --> 00:02:19,120
as for each trial so when a trial begins

59
00:02:23,030 --> 00:02:20,640
the light source and detector are turned

60
00:02:25,190 --> 00:02:23,040
on and when that trial completes the

61
00:02:27,350 --> 00:02:25,200
light source and detector are turned off

62
00:02:28,949 --> 00:02:27,360
they stay off for three minutes and then

63
00:02:30,309 --> 00:02:28,959
they turn on again for the next trial

64
00:02:33,430 --> 00:02:30,319
which might have a different duration or

65
00:02:35,430 --> 00:02:33,440
the same duration as the previous trial

66
00:02:36,949 --> 00:02:35,440
and then the average photon counts as i

67
00:02:38,790 --> 00:02:36,959
mentioned

68
00:02:39,670 --> 00:02:38,800
show a dependence

69
00:02:41,270 --> 00:02:39,680
on

70
00:02:43,190 --> 00:02:41,280
the future

71
00:02:45,110 --> 00:02:43,200
random

72
00:02:47,350 --> 00:02:45,120
duration

73
00:02:49,670 --> 00:02:47,360
and the time scales involved are seconds

74
00:02:51,430 --> 00:02:49,680
to minutes

75
00:02:53,430 --> 00:02:51,440
and what you can see in this plot here

76
00:02:55,509 --> 00:02:53,440
the reason i'm showing this

77
00:02:58,229 --> 00:02:55,519
is that you can see very clearly here

78
00:03:01,350 --> 00:02:58,239
the yellow curve which is a

79
00:03:04,149 --> 00:03:01,360
average of long duration runs

80
00:03:05,670 --> 00:03:04,159
is distinctly different than the green

81
00:03:07,509 --> 00:03:05,680
blue and you can see a little purple

82
00:03:08,470 --> 00:03:07,519
curve here which are shorter duration

83
00:03:10,710 --> 00:03:08,480
runs

84
00:03:12,630 --> 00:03:10,720
and that difference between those longer

85
00:03:14,070 --> 00:03:12,640
and shorter duration runs

86
00:03:15,030 --> 00:03:14,080
doesn't just appear at the end of the

87
00:03:17,270 --> 00:03:15,040
run

88
00:03:22,149 --> 00:03:17,280
you see a difference in the earlier part

89
00:03:23,430 --> 00:03:22,159
of that of of those runs when averaged

90
00:03:24,420 --> 00:03:23,440
even though

91
00:03:26,149 --> 00:03:24,430
that

92
00:03:28,309 --> 00:03:26,159
[Music]

93
00:03:29,430 --> 00:03:28,319
that decision hasn't really been

94
00:03:31,509 --> 00:03:29,440
implemented

95
00:03:33,430 --> 00:03:31,519
in any difference here and actually even

96
00:03:35,110 --> 00:03:33,440
before the decision is made here the red

97
00:03:37,110 --> 00:03:35,120
line is that decision about how long the

98
00:03:39,830 --> 00:03:37,120
duration will be even before that

99
00:03:42,789 --> 00:03:39,840
decision is made dr mossbridge found a

100
00:03:45,270 --> 00:03:42,799
difference between the longest duration

101
00:03:48,710 --> 00:03:45,280
trials and the shorter ones

102
00:03:52,710 --> 00:03:50,070
so

103
00:03:54,630 --> 00:03:52,720
the current experiment that i've run for

104
00:03:56,550 --> 00:03:54,640
it's actually just completed a two day

105
00:03:58,070 --> 00:03:56,560
two days ago it completed a year running

106
00:04:01,509 --> 00:03:58,080
but the data i'm reporting is just

107
00:04:03,830 --> 00:04:01,519
through february 21st

108
00:04:05,350 --> 00:04:03,840
there were key differences and upgrades

109
00:04:07,509 --> 00:04:05,360
in the hardware

110
00:04:09,670 --> 00:04:07,519
the photon detector

111
00:04:12,710 --> 00:04:09,680
instead of a photomultiplier tube

112
00:04:15,509 --> 00:04:12,720
is a single photon counting module in

113
00:04:17,509 --> 00:04:15,519
other words a reverse bias diode

114
00:04:18,870 --> 00:04:17,519
the pulse counting is done by a field

115
00:04:21,909 --> 00:04:18,880
programmable gate array which is

116
00:04:25,510 --> 00:04:21,919
independent of the computer

117
00:04:27,510 --> 00:04:25,520
there's random number generation um

118
00:04:29,270 --> 00:04:27,520
julia and i are quite excited about that

119
00:04:31,270 --> 00:04:29,280
is a big upgrade and i'll tell you more

120
00:04:32,150 --> 00:04:31,280
about that later

121
00:04:35,430 --> 00:04:32,160
um

122
00:04:36,310 --> 00:04:35,440
i used labview instead of matlab

123
00:04:40,390 --> 00:04:36,320
and

124
00:04:42,790 --> 00:04:40,400
independently written

125
00:04:45,670 --> 00:04:42,800
she had told me her protocol but

126
00:04:48,390 --> 00:04:45,680
i hadn't worked with any of her code

127
00:04:50,790 --> 00:04:48,400
and to ramp up and down the voltage for

128
00:04:53,430 --> 00:04:50,800
the inverse bias of that diode in the

129
00:04:54,150 --> 00:04:53,440
photodetector i used an rc circuit for

130
00:04:57,110 --> 00:04:54,160
that

131
00:05:00,550 --> 00:04:57,120
um there was an rc circuit already in

132
00:05:03,749 --> 00:05:00,560
the in the box that i had for operating

133
00:05:05,749 --> 00:05:03,759
that photodiode and what i did is jumper

134
00:05:07,110 --> 00:05:05,759
in some circuitry so the computer could

135
00:05:08,390 --> 00:05:07,120
turn it on and off instead of a manual

136
00:05:11,110 --> 00:05:08,400
switch

137
00:05:12,950 --> 00:05:11,120
and then um the operate the

138
00:05:15,029 --> 00:05:12,960
experiment was operated continuously 24

139
00:05:17,350 --> 00:05:15,039
7 for over six months

140
00:05:21,590 --> 00:05:17,360
whereas dr moss bridges experiments we

141
00:05:25,830 --> 00:05:23,110
okay so let's look at the optics as i

142
00:05:27,590 --> 00:05:25,840
mentioned we we're photon counting

143
00:05:29,270 --> 00:05:27,600
experiment our photons begin here from

144
00:05:31,749 --> 00:05:29,280
this light emitting diode

145
00:05:33,909 --> 00:05:31,759
this is turned on and off by the

146
00:05:36,230 --> 00:05:33,919
computer software in labview

147
00:05:38,790 --> 00:05:36,240
and then it hits a diffuser so that you

148
00:05:40,070 --> 00:05:38,800
get an even so like even if the led were

149
00:05:41,270 --> 00:05:40,080
to move a little bit it wouldn't affect

150
00:05:42,790 --> 00:05:41,280
the counting that much because it's

151
00:05:44,150 --> 00:05:42,800
being diffused here

152
00:05:45,510 --> 00:05:44,160
then it goes through some miscellaneous

153
00:05:46,950 --> 00:05:45,520
objects that just happen to be on the

154
00:05:49,510 --> 00:05:46,960
table and i didn't

155
00:05:50,950 --> 00:05:49,520
see need to remove them

156
00:05:52,790 --> 00:05:50,960
and then this is important it goes

157
00:05:54,390 --> 00:05:52,800
through this detuned bandpass filter

158
00:05:56,150 --> 00:05:54,400
what i mean by detuned is the bandpass

159
00:05:59,590 --> 00:05:56,160
filter is designed for infrared photons

160
00:06:01,510 --> 00:05:59,600
at 880 nanometers but this led is around

161
00:06:03,270 --> 00:06:01,520
650 so

162
00:06:05,110 --> 00:06:03,280
uh most of the light will be blocked by

163
00:06:07,270 --> 00:06:05,120
this that's important because it gets us

164
00:06:09,670 --> 00:06:07,280
down to a low counting rate

165
00:06:12,309 --> 00:06:09,680
and then the photons that are coupled

166
00:06:13,990 --> 00:06:12,319
into this fiber coupler go to the single

167
00:06:16,629 --> 00:06:14,000
photon county module that reverse bias

168
00:06:19,189 --> 00:06:16,639
diode that i mentioned previously

169
00:06:21,029 --> 00:06:19,199
so it's only about 3 000 counts per

170
00:06:22,469 --> 00:06:21,039
second we're getting here so that's a

171
00:06:25,029 --> 00:06:22,479
low enough rate that it seems

172
00:06:27,350 --> 00:06:25,039
appropriate for this kind of experiment

173
00:06:29,749 --> 00:06:27,360
okay so let's look at the overall

174
00:06:31,350 --> 00:06:29,759
experiment so this is the optics part

175
00:06:33,110 --> 00:06:31,360
and you see the optics part is shown

176
00:06:34,790 --> 00:06:33,120
here this red line

177
00:06:36,550 --> 00:06:34,800
that's representing the light coming

178
00:06:38,710 --> 00:06:36,560
from the light emitting diode to the

179
00:06:39,990 --> 00:06:38,720
single photon counting module here's the

180
00:06:41,029 --> 00:06:40,000
field programmable gate array i

181
00:06:43,510 --> 00:06:41,039
mentioned

182
00:06:45,909 --> 00:06:43,520
it is controlled by the computer but it

183
00:06:48,469 --> 00:06:45,919
counts autonomously and then feeds that

184
00:06:50,550 --> 00:06:48,479
data back to the computer

185
00:06:52,710 --> 00:06:50,560
the single photon county module is also

186
00:06:54,950 --> 00:06:52,720
powered it's powered by this power

187
00:06:57,430 --> 00:06:54,960
supply that i mentioned that has a

188
00:06:59,270 --> 00:06:57,440
resistor capacitor circuit that ramps

189
00:07:01,189 --> 00:06:59,280
the voltage up and down

190
00:07:02,950 --> 00:07:01,199
that's so that the

191
00:07:04,710 --> 00:07:02,960
the inverse

192
00:07:06,629 --> 00:07:04,720
bias diode is a little bit gentle just

193
00:07:08,230 --> 00:07:06,639
like a phototube so you need to ramp the

194
00:07:10,070 --> 00:07:08,240
voltage up and down

195
00:07:11,749 --> 00:07:10,080
and there's an optical isolator there

196
00:07:13,670 --> 00:07:11,759
that's just to avoid ground loops right

197
00:07:14,950 --> 00:07:13,680
because you have a connection here you

198
00:07:16,469 --> 00:07:14,960
have a connection

199
00:07:18,950 --> 00:07:16,479
here

200
00:07:20,390 --> 00:07:18,960
want to have a ground loop so i put an

201
00:07:22,150 --> 00:07:20,400
optical isolator there just to avoid

202
00:07:23,350 --> 00:07:22,160
that

203
00:07:25,430 --> 00:07:23,360
and then it's all controlled by this

204
00:07:28,070 --> 00:07:25,440
labview program it controls the led

205
00:07:31,350 --> 00:07:28,080
controls the counter and it controls the

206
00:07:33,589 --> 00:07:31,360
turning on and off of the detector now

207
00:07:34,469 --> 00:07:33,599
where do i get my random numbers

208
00:07:40,790 --> 00:07:34,479
to

209
00:07:43,029 --> 00:07:40,800
that comes from this whole system down

210
00:07:44,950 --> 00:07:43,039
here which is completely independent

211
00:07:46,390 --> 00:07:44,960
it's not a computer

212
00:07:48,469 --> 00:07:46,400
it is

213
00:07:49,589 --> 00:07:48,479
hardware implemented and

214
00:07:51,189 --> 00:07:49,599
this is the

215
00:07:52,710 --> 00:07:51,199
upgrade in the random number generator

216
00:07:54,230 --> 00:07:52,720
that i said that dr mossbridge and i are

217
00:07:55,029 --> 00:07:54,240
so excited about and i'll tell you about

218
00:07:57,029 --> 00:07:55,039
that

219
00:07:59,430 --> 00:07:57,039
next

220
00:08:00,950 --> 00:07:59,440
so um

221
00:08:03,110 --> 00:08:00,960
actually let me point out before i go to

222
00:08:05,270 --> 00:08:03,120
that next slide that there's

223
00:08:07,589 --> 00:08:05,280
a photo multiplier here

224
00:08:09,189 --> 00:08:07,599
this is independent of the photo

225
00:08:10,950 --> 00:08:09,199
detector here in the single photon

226
00:08:13,029 --> 00:08:10,960
counting module this is an independent

227
00:08:15,110 --> 00:08:13,039
photo multiplier attached to a big tank

228
00:08:16,790 --> 00:08:15,120
of simulator oil this thing is big it's

229
00:08:17,990 --> 00:08:16,800
it spans between two levels in the

230
00:08:19,589 --> 00:08:18,000
building

231
00:08:21,589 --> 00:08:19,599
and then it goes through some amplifiers

232
00:08:22,950 --> 00:08:21,599
and threshold discriminator and then it

233
00:08:24,550 --> 00:08:22,960
hooks up to a pseudo-random number

234
00:08:26,230 --> 00:08:24,560
generator but the numbers that are

235
00:08:27,909 --> 00:08:26,240
generated aren't just pseudorandom it

236
00:08:29,510 --> 00:08:27,919
actually incorporates

237
00:08:31,029 --> 00:08:29,520
the true randomness over here and i'll

238
00:08:33,909 --> 00:08:31,039
explain how that works as i go through

239
00:08:36,230 --> 00:08:33,919
the those next slides

240
00:08:38,070 --> 00:08:36,240
so this is the top end of a photo

241
00:08:39,110 --> 00:08:38,080
multiplier tube it's it's pretty big

242
00:08:42,149 --> 00:08:39,120
it's uh

243
00:08:44,230 --> 00:08:42,159
like kind of the size of a watermelon um

244
00:08:45,910 --> 00:08:44,240
and the photo multiplier tube is down

245
00:08:48,150 --> 00:08:45,920
below and then down below that is tank

246
00:08:50,470 --> 00:08:48,160
scintillator oil supported by these

247
00:08:52,230 --> 00:08:50,480
uh steel beams

248
00:08:53,350 --> 00:08:52,240
and it

249
00:08:54,630 --> 00:08:53,360
when

250
00:08:56,870 --> 00:08:54,640
stuff happens

251
00:08:58,389 --> 00:08:56,880
uh in the atmosphere or nearby you may

252
00:09:00,470 --> 00:08:58,399
get a particle that passes through that

253
00:09:01,829 --> 00:09:00,480
tank you get a flash of light that gives

254
00:09:03,670 --> 00:09:01,839
you a pulse out of the photomultiplier

255
00:09:06,630 --> 00:09:03,680
tube and because it's a pretty big tank

256
00:09:08,870 --> 00:09:06,640
we get about 40 hertz uh pulse rate out

257
00:09:11,030 --> 00:09:08,880
of this

258
00:09:13,590 --> 00:09:11,040
and this provides the true randomness

259
00:09:15,190 --> 00:09:13,600
for our random number generator

260
00:09:18,630 --> 00:09:15,200
then

261
00:09:20,870 --> 00:09:18,640
to get uniform likelihood in the numbers

262
00:09:24,310 --> 00:09:20,880
those true random pulses

263
00:09:25,430 --> 00:09:24,320
are used as the clock inputs on each of

264
00:09:33,350 --> 00:09:25,440
these

265
00:09:35,190 --> 00:09:33,360
bit shift registers

266
00:09:37,350 --> 00:09:35,200
and i can talk more about this when we

267
00:09:39,110 --> 00:09:37,360
get to q a but i enjoyed building this

268
00:09:40,389 --> 00:09:39,120
circuit and

269
00:09:43,030 --> 00:09:40,399
basically

270
00:09:45,350 --> 00:09:43,040
it the pseudorandom numbers are advanced

271
00:09:46,710 --> 00:09:45,360
to the next random number each times

272
00:09:49,269 --> 00:09:46,720
there's a detection in that tank

273
00:09:50,790 --> 00:09:49,279
assimilator oil so when there's a detect

274
00:09:52,550 --> 00:09:50,800
detection in the tank

275
00:09:54,070 --> 00:09:52,560
then you get an advancement to the next

276
00:09:56,389 --> 00:09:54,080
number and because those detections in

277
00:09:58,550 --> 00:09:56,399
the tank are happening at about 40 hertz

278
00:09:59,430 --> 00:09:58,560
about every two seconds you get a whole

279
00:10:03,670 --> 00:09:59,440
fresh

280
00:10:06,630 --> 00:10:03,680
uh 80 bits here in in these random bits

281
00:10:09,590 --> 00:10:08,470
and that's used

282
00:10:11,430 --> 00:10:09,600
right here

283
00:10:14,710 --> 00:10:11,440
two bits of that information are

284
00:10:16,550 --> 00:10:14,720
collected by the labview system and like

285
00:10:18,790 --> 00:10:16,560
i said these are independent so the

286
00:10:21,590 --> 00:10:18,800
labview system doesn't know what's going

287
00:10:24,550 --> 00:10:21,600
on there's just two wires connected from

288
00:10:26,230 --> 00:10:24,560
that hardware circuit to inputs of the

289
00:10:27,430 --> 00:10:26,240
labview system

290
00:10:29,350 --> 00:10:27,440
and

291
00:10:32,389 --> 00:10:29,360
then they're just literally two jumper

292
00:10:34,310 --> 00:10:32,399
wires like that and the labview system

293
00:10:36,870 --> 00:10:34,320
simply says i want to read these two

294
00:10:38,230 --> 00:10:36,880
digital inputs boom and it gets a random

295
00:10:40,550 --> 00:10:38,240
pair of bits

296
00:10:42,870 --> 00:10:40,560
and uses that to determine what the

297
00:10:44,710 --> 00:10:42,880
duration of the trial will be let me go

298
00:10:46,870 --> 00:10:44,720
through the whole trial so

299
00:10:49,350 --> 00:10:46,880
each time a trial occurs

300
00:10:52,150 --> 00:10:49,360
the system let's say is off initially

301
00:10:53,990 --> 00:10:52,160
right when that trial begins

302
00:10:56,310 --> 00:10:54,000
there's a warm-up period

303
00:10:59,030 --> 00:10:56,320
after the photon counter and photon

304
00:11:01,110 --> 00:10:59,040
source the led are turned on

305
00:11:02,949 --> 00:11:01,120
that warm-up period lasts for six

306
00:11:04,790 --> 00:11:02,959
seconds in my experiment that's

307
00:11:06,310 --> 00:11:04,800
different than dr mossberg's experiment

308
00:11:07,670 --> 00:11:06,320
my system requires the six second

309
00:11:10,389 --> 00:11:07,680
warm-up

310
00:11:12,550 --> 00:11:10,399
and then there's three periods of

311
00:11:14,550 --> 00:11:12,560
counting of 11 seconds each for the

312
00:11:17,030 --> 00:11:14,560
photons so that's when you see on those

313
00:11:18,550 --> 00:11:17,040
plots those three dots at the beginning

314
00:11:21,350 --> 00:11:18,560
you'll see that again

315
00:11:23,350 --> 00:11:21,360
that's or the three x's that's um those

316
00:11:25,670 --> 00:11:23,360
three counts that occur in every trial

317
00:11:27,350 --> 00:11:25,680
regardless of the random number then

318
00:11:29,430 --> 00:11:27,360
depending on the random number that's

319
00:11:33,509 --> 00:11:29,440
chosen

320
00:11:37,110 --> 00:11:33,519
you either have 0 20 30 or 60 additional

321
00:11:39,910 --> 00:11:37,120
counts so those would be 20 30

322
00:11:41,910 --> 00:11:39,920
60 or 0 additional 11 second per period

323
00:11:43,110 --> 00:11:41,920
so this whole trial can go on for

324
00:11:45,829 --> 00:11:43,120
several minutes

325
00:11:48,389 --> 00:11:45,839
after which there's a three second off

326
00:11:50,629 --> 00:11:48,399
time where both the detector and the

327
00:11:52,389 --> 00:11:50,639
source are off before beginning again

328
00:11:54,949 --> 00:11:52,399
and that's a key feature of the

329
00:11:56,949 --> 00:11:54,959
experiment that dr mossbridge did and

330
00:11:59,030 --> 00:11:56,959
seems to be key to getting this to work

331
00:12:02,790 --> 00:11:59,040
and so i replicated that in this

332
00:12:05,990 --> 00:12:04,389
so let's look um

333
00:12:08,069 --> 00:12:06,000
this is the second time you've seen this

334
00:12:09,750 --> 00:12:08,079
before i started out this with my first

335
00:12:11,110 --> 00:12:09,760
slide

336
00:12:13,430 --> 00:12:11,120
but i'm pointing to different things

337
00:12:15,990 --> 00:12:13,440
here so i was just talking about how the

338
00:12:18,150 --> 00:12:16,000
random number chooses how long the trial

339
00:12:20,310 --> 00:12:18,160
will go on for and you can see here the

340
00:12:21,910 --> 00:12:20,320
random number shows that the trial would

341
00:12:23,670 --> 00:12:21,920
go on for zero additional counts and you

342
00:12:25,990 --> 00:12:23,680
see there's just those three right it

343
00:12:28,230 --> 00:12:26,000
just did the three initial counts

344
00:12:30,230 --> 00:12:28,240
um you'll see it says here times 10 to

345
00:12:32,150 --> 00:12:30,240
the fourth so this is about 30 000

346
00:12:34,150 --> 00:12:32,160
counts in each one of these

347
00:12:37,190 --> 00:12:34,160
bins right so there's three bins of

348
00:12:38,870 --> 00:12:37,200
counting before uh the random number and

349
00:12:40,310 --> 00:12:38,880
then at this point the random number

350
00:12:41,350 --> 00:12:40,320
said we're not going to go on so okay

351
00:12:43,190 --> 00:12:41,360
fine

352
00:12:45,430 --> 00:12:43,200
here the random number said we're going

353
00:12:47,269 --> 00:12:45,440
to count for 20 more times

354
00:12:49,430 --> 00:12:47,279
and here 30 more here's another 20.

355
00:12:51,350 --> 00:12:49,440
here's another one where it just counted

356
00:12:53,030 --> 00:12:51,360
uh for three and then stopped and here's

357
00:12:55,990 --> 00:12:53,040
one where it went on for 60 additional

358
00:12:58,949 --> 00:12:56,710
so

359
00:13:01,110 --> 00:12:58,959
the trials have a minimum of a 3 and a

360
00:13:05,030 --> 00:13:01,120
maximum of 63

361
00:13:08,389 --> 00:13:07,430
so here's what the data looked like over

362
00:13:10,949 --> 00:13:08,399
the whole

363
00:13:14,470 --> 00:13:10,959
six month plus period of time

364
00:13:18,389 --> 00:13:14,480
we completed 34 454 trials

365
00:13:20,629 --> 00:13:18,399
for a total of 1 million 47 902 of those

366
00:13:21,590 --> 00:13:20,639
11 second measurements

367
00:13:24,310 --> 00:13:21,600
and

368
00:13:27,509 --> 00:13:24,320
so there's a dot on this plot for each

369
00:13:29,509 --> 00:13:27,519
one of those one million measurements

370
00:13:30,870 --> 00:13:29,519
you can see there's

371
00:13:32,230 --> 00:13:30,880
quite a

372
00:13:33,670 --> 00:13:32,240
a lot of interesting features in this

373
00:13:35,829 --> 00:13:33,680
data

374
00:13:38,069 --> 00:13:35,839
there's sort of a general downtrend

375
00:13:39,430 --> 00:13:38,079
there's a daily variation and actually

376
00:13:40,949 --> 00:13:39,440
if you look under that there's actually

377
00:13:43,189 --> 00:13:40,959
twice daily variation you can't really

378
00:13:44,550 --> 00:13:43,199
see here but these little bumps you're

379
00:13:47,110 --> 00:13:44,560
seeing you know

380
00:13:49,030 --> 00:13:47,120
one two three four five those little

381
00:13:53,350 --> 00:13:49,040
bumps are each one day

382
00:13:58,150 --> 00:13:55,910
there's also this sort of larger

383
00:14:00,870 --> 00:13:58,160
movement about every 25 days there's

384
00:14:02,389 --> 00:14:00,880
these kind of larger movements and dips

385
00:14:04,069 --> 00:14:02,399
and these are not particularly

386
00:14:05,829 --> 00:14:04,079
correlated with times that i went in the

387
00:14:06,870 --> 00:14:05,839
room to backup data or anything like

388
00:14:08,550 --> 00:14:06,880
that

389
00:14:09,750 --> 00:14:08,560
this is one continuous run of the

390
00:14:12,150 --> 00:14:09,760
experiment

391
00:14:13,910 --> 00:14:12,160
that's actually still run

392
00:14:15,269 --> 00:14:13,920
was still running until tuesday when i

393
00:14:17,269 --> 00:14:15,279
shut it down

394
00:14:19,590 --> 00:14:17,279
so

395
00:14:21,350 --> 00:14:19,600
the computer was not restart the

396
00:14:22,870 --> 00:14:21,360
software was running continuously all

397
00:14:24,790 --> 00:14:22,880
the hardware was running continuously

398
00:14:26,389 --> 00:14:24,800
there were no power outages

399
00:14:28,310 --> 00:14:26,399
during the whole time

400
00:14:30,389 --> 00:14:28,320
you can also see there's interesting

401
00:14:31,750 --> 00:14:30,399
things going up here and going down here

402
00:14:34,629 --> 00:14:31,760
which

403
00:14:35,990 --> 00:14:34,639
i'll talk about in a couple slides later

404
00:14:38,470 --> 00:14:36,000
i want to point out that there's a

405
00:14:39,670 --> 00:14:38,480
really big change in the data right here

406
00:14:41,110 --> 00:14:39,680
that

407
00:14:43,590 --> 00:14:41,120
up to here

408
00:14:45,350 --> 00:14:43,600
it's kind of sort of a similar pattern

409
00:14:48,310 --> 00:14:45,360
and then the pattern really changes i'll

410
00:14:49,910 --> 00:14:48,320
talk about that in the next slide

411
00:14:51,990 --> 00:14:49,920
so

412
00:14:53,030 --> 00:14:52,000
dr mosbridge noticed

413
00:14:54,710 --> 00:14:53,040
that

414
00:14:55,509 --> 00:14:54,720
you know just looking things up online

415
00:14:57,350 --> 00:14:55,519
like

416
00:14:58,790 --> 00:14:57,360
what happened on the 17th of january and

417
00:15:01,030 --> 00:14:58,800
she looked at some newspaper headlines

418
00:15:03,189 --> 00:15:01,040
and there was this headline about

419
00:15:04,629 --> 00:15:03,199
officials being pegged as uh you know

420
00:15:05,990 --> 00:15:04,639
for stuff

421
00:15:07,509 --> 00:15:06,000
um

422
00:15:08,710 --> 00:15:07,519
that was a major flashpoint that didn't

423
00:15:09,990 --> 00:15:08,720
turn out to be a big flashpoint but

424
00:15:11,430 --> 00:15:10,000
there was a lot of energy a lot of

425
00:15:13,269 --> 00:15:11,440
concern about it

426
00:15:14,629 --> 00:15:13,279
so maybe emotions affect these things

427
00:15:16,629 --> 00:15:14,639
like they affect the global

428
00:15:18,629 --> 00:15:16,639
consciousness project

429
00:15:20,310 --> 00:15:18,639
there's also striking situations where

430
00:15:23,430 --> 00:15:20,320
you don't see anything the here

431
00:15:25,030 --> 00:15:23,440
september 22nd is the

432
00:15:27,670 --> 00:15:25,040
fall equinox and here's the winter

433
00:15:30,949 --> 00:15:27,680
solstice on december 21st and there's no

434
00:15:34,790 --> 00:15:32,870
thing that seems to be happening there

435
00:15:37,509 --> 00:15:34,800
but there's other ones like election day

436
00:15:39,350 --> 00:15:37,519
uh november 3rd appears to be just right

437
00:15:41,430 --> 00:15:39,360
before there's this significant change

438
00:15:43,269 --> 00:15:41,440
in the data and these could just be

439
00:15:45,590 --> 00:15:43,279
coincidences maybe this is when the

440
00:15:47,110 --> 00:15:45,600
weather changed or something in berkeley

441
00:15:49,269 --> 00:15:47,120
so

442
00:15:50,629 --> 00:15:49,279
um this is interesting but it's not the

443
00:15:52,629 --> 00:15:50,639
main focus of this talk so i'm going to

444
00:15:54,470 --> 00:15:52,639
set this aside for now and get back to

445
00:15:56,470 --> 00:15:54,480
the main focus which is looking at the

446
00:16:00,230 --> 00:15:56,480
correlations within each run with it

447
00:16:04,870 --> 00:16:01,030
so

448
00:16:08,870 --> 00:16:07,030
things that are really outlined here

449
00:16:10,230 --> 00:16:08,880
above and below

450
00:16:42,790 --> 00:16:10,240
i

451
00:16:45,670 --> 00:16:42,800
be groups

452
00:16:47,430 --> 00:16:45,680
and so i used that in

453
00:16:49,030 --> 00:16:47,440
noticing what are the outliers and

454
00:16:50,710 --> 00:16:49,040
setting them aside

455
00:16:53,030 --> 00:16:50,720
you'll see later i do the analysis with

456
00:16:55,509 --> 00:16:53,040
and without the outliers and the results

457
00:16:57,509 --> 00:16:55,519
are actually fairly similar but this was

458
00:16:59,670 --> 00:16:57,519
one of the initial things i did was to

459
00:17:01,620 --> 00:16:59,680
identify outliers

460
00:17:03,030 --> 00:17:01,630
and so there's my outliers

461
00:17:04,230 --> 00:17:03,040
[Music]

462
00:17:06,870 --> 00:17:04,240
there are approximately a thousand

463
00:17:09,029 --> 00:17:06,880
outliers in the whole million counts

464
00:17:10,630 --> 00:17:09,039
uh but they cluster in about 64 out of

465
00:17:12,949 --> 00:17:10,640
the 34 000 trials so they're pretty

466
00:17:15,350 --> 00:17:12,959
heavily clustered and so

467
00:17:17,029 --> 00:17:15,360
uh can be fairly um

468
00:17:18,630 --> 00:17:17,039
you just remove those trials and it's

469
00:17:21,350 --> 00:17:18,640
pretty straightforward since we're doing

470
00:17:23,350 --> 00:17:21,360
a trial by trial

471
00:17:25,510 --> 00:17:23,360
approaching the analysis

472
00:17:27,829 --> 00:17:25,520
okay now this is the third time i'm

473
00:17:29,750 --> 00:17:27,839
showing you this slide

474
00:17:31,909 --> 00:17:29,760
i've changed things a little bit

475
00:17:34,150 --> 00:17:31,919
i'm calling attention to the trials that

476
00:17:36,630 --> 00:17:34,160
have data after the decision so remember

477
00:17:38,710 --> 00:17:36,640
there's a decision here by random number

478
00:17:41,350 --> 00:17:38,720
as to how long the trial will go on

479
00:17:44,310 --> 00:17:41,360
whether it's going to go on for 0 20 30

480
00:17:45,909 --> 00:17:44,320
or 60 counts however i'm

481
00:17:48,390 --> 00:17:45,919
paying special attention to the ones

482
00:17:49,430 --> 00:17:48,400
that have at least some counts after the

483
00:17:51,909 --> 00:17:49,440
decision

484
00:17:54,070 --> 00:17:51,919
because an analysis approach that dr

485
00:17:56,390 --> 00:17:54,080
mosbridge came up with utilizes a

486
00:17:59,830 --> 00:17:56,400
correlation between those before and

487
00:18:03,190 --> 00:18:00,789
okay

488
00:18:04,950 --> 00:18:03,200
so here's a schematic of that analysis

489
00:18:07,590 --> 00:18:04,960
that dr mossbridge came up with and that

490
00:18:09,110 --> 00:18:07,600
i implemented in matlab

491
00:18:09,990 --> 00:18:09,120
so

492
00:18:16,150 --> 00:18:10,000
the

493
00:18:17,830 --> 00:18:16,160
collected

494
00:18:20,070 --> 00:18:17,840
and so

495
00:18:22,230 --> 00:18:20,080
the

496
00:18:23,830 --> 00:18:22,240
for each day

497
00:18:25,590 --> 00:18:23,840
we look at the

498
00:18:27,110 --> 00:18:25,600
pre-decision data on the trials and we

499
00:18:28,390 --> 00:18:27,120
look at the post-decision data on the

500
00:18:29,270 --> 00:18:28,400
trials

501
00:18:31,830 --> 00:18:29,280
and

502
00:18:33,430 --> 00:18:31,840
we average over actually just two of the

503
00:18:36,230 --> 00:18:33,440
three pre-decision

504
00:18:38,870 --> 00:18:36,240
data counts and we average over all of

505
00:18:41,350 --> 00:18:38,880
the post-decision date accounts okay

506
00:18:45,430 --> 00:18:41,360
those averages form an x-y pair in some

507
00:18:46,630 --> 00:18:45,440
arbitrary space for each trial okay so

508
00:18:48,789 --> 00:18:46,640
now we do

509
00:18:51,590 --> 00:18:48,799
is

510
00:18:54,710 --> 00:18:51,600
depending on the duration of that trial

511
00:18:56,789 --> 00:18:54,720
we sort those x y values

512
00:18:59,430 --> 00:18:56,799
so that they contribute to either

513
00:19:02,150 --> 00:18:59,440
an average of the 20

514
00:19:04,549 --> 00:19:02,160
so so in other words all the x's

515
00:19:06,710 --> 00:19:04,559
for for the 20 duration trials get

516
00:19:08,870 --> 00:19:06,720
averaged and all the y's for the 20

517
00:19:12,390 --> 00:19:08,880
duration trials get averaged that forms

518
00:19:15,270 --> 00:19:12,400
this x20 y20 pair same way for the 30s

519
00:19:17,510 --> 00:19:15,280
same way for the 60s so we have three xy

520
00:19:19,909 --> 00:19:17,520
pairs and we do a correlation and we get

521
00:19:21,190 --> 00:19:19,919
the row value or we're just calling it r

522
00:19:22,230 --> 00:19:21,200
here

523
00:19:23,029 --> 00:19:22,240
and that

524
00:19:25,590 --> 00:19:23,039
uh

525
00:19:28,150 --> 00:19:25,600
that our value then gets averaged over

526
00:19:29,350 --> 00:19:28,160
all 210 days so we get one r value for

527
00:19:32,549 --> 00:19:29,360
each day

528
00:19:33,430 --> 00:19:32,559
and then we average those 210 r values

529
00:19:34,470 --> 00:19:33,440
and

530
00:19:36,630 --> 00:19:34,480
um

531
00:19:38,950 --> 00:19:36,640
what we find is a rather strong

532
00:19:40,710 --> 00:19:38,960
correlation the average correlation is

533
00:19:42,710 --> 00:19:40,720
quite strong and that's shown by this

534
00:19:45,270 --> 00:19:42,720
line here on the graph

535
00:19:46,390 --> 00:19:45,280
almost 0.75

536
00:19:47,590 --> 00:19:46,400
on the other hand

537
00:19:49,750 --> 00:19:47,600
if we

538
00:19:52,870 --> 00:19:49,760
take each x y pair and contribute it

539
00:19:53,669 --> 00:19:52,880
instead to a random

540
00:19:57,990 --> 00:19:53,679
x

541
00:20:00,150 --> 00:19:58,000
even though the actual duration might be

542
00:20:02,710 --> 00:20:00,160
20 it might be con

543
00:20:06,070 --> 00:20:02,720
that x y average might be contributed to

544
00:20:08,310 --> 00:20:06,080
the x 30 y 30 pair for example okay and

545
00:20:10,390 --> 00:20:08,320
we do the scramble in such a way that we

546
00:20:12,070 --> 00:20:10,400
preserve for each day

547
00:20:16,390 --> 00:20:12,080
the number of

548
00:20:17,990 --> 00:20:16,400
x and y that contribute to the x 20 y 20

549
00:20:19,669 --> 00:20:18,000
and the number that contribute to the x

550
00:20:23,190 --> 00:20:19,679
30 y 30 and the number that contribute

551
00:20:25,590 --> 00:20:23,200
to the x 60 y60 so we preserve how many

552
00:20:28,950 --> 00:20:25,600
trials contribute to each if we do that

553
00:20:30,310 --> 00:20:28,960
correlation we get again 210 r values

554
00:20:33,350 --> 00:20:30,320
one for each day

555
00:20:35,350 --> 00:20:33,360
and then averaging those 210 values we

556
00:20:37,590 --> 00:20:35,360
get a much weaker correlation

557
00:20:40,390 --> 00:20:37,600
and that's demonstrated by doing this

558
00:20:43,110 --> 00:20:40,400
scramble 40 000 times you know 40 000

559
00:20:44,310 --> 00:20:43,120
different random numbers

560
00:20:46,710 --> 00:20:44,320
and then

561
00:20:49,430 --> 00:20:46,720
um that forms this

562
00:20:50,789 --> 00:20:49,440
histogram here shown

563
00:20:53,270 --> 00:20:50,799
which performs

564
00:20:55,830 --> 00:20:53,280
quite a rather nice bell curve

565
00:21:00,470 --> 00:20:58,710
let me move to the next slide

566
00:21:02,149 --> 00:21:00,480
so here

567
00:21:03,590 --> 00:21:02,159
this is that same plot i just showed you

568
00:21:05,990 --> 00:21:03,600
but just bigger

569
00:21:07,270 --> 00:21:06,000
and here i'm calculating

570
00:21:09,190 --> 00:21:07,280
a sigma

571
00:21:10,549 --> 00:21:09,200
for the

572
00:21:11,669 --> 00:21:10,559
original assignment the correct

573
00:21:13,270 --> 00:21:11,679
assignment

574
00:21:14,390 --> 00:21:13,280
of the data

575
00:21:16,310 --> 00:21:14,400
in

576
00:21:17,669 --> 00:21:16,320
by duration versus the scrambled

577
00:21:18,950 --> 00:21:17,679
assignments

578
00:21:21,350 --> 00:21:18,960
and so

579
00:21:23,190 --> 00:21:21,360
uh the way i've done that is to take the

580
00:21:27,190 --> 00:21:23,200
original r value the correct assignment

581
00:21:29,750 --> 00:21:27,200
r value subtract the r value for the

582
00:21:31,990 --> 00:21:29,760
average of the scrambles and then divide

583
00:21:34,950 --> 00:21:32,000
by the standard deviation for those

584
00:21:39,190 --> 00:21:34,960
scrambles and that produces a

585
00:21:44,710 --> 00:21:41,990
i mentioned earlier that this is

586
00:21:46,549 --> 00:21:44,720
pretty robust to outlier removal so if i

587
00:21:48,310 --> 00:21:46,559
put those outliers back in like just

588
00:21:50,710 --> 00:21:48,320
ignore that part of the presentation

589
00:21:53,350 --> 00:21:50,720
where i talked about outliers

590
00:21:57,350 --> 00:21:53,360
then we get a sigma of 5.27 it's not

591
00:21:59,830 --> 00:21:58,470
um

592
00:22:02,710 --> 00:21:59,840
this is

593
00:22:04,149 --> 00:22:02,720
so far the major result that we've come

594
00:22:06,390 --> 00:22:04,159
up with

595
00:22:09,190 --> 00:22:06,400
it's the um

596
00:22:10,230 --> 00:22:09,200
it's the strongest result so far and

597
00:22:13,990 --> 00:22:10,240
we're still

598
00:22:18,230 --> 00:22:14,000
doing a lot of analysis to figure out if

599
00:22:19,990 --> 00:22:18,240
we can come up with um something better

600
00:22:21,750 --> 00:22:20,000
in addition to this

601
00:22:24,310 --> 00:22:21,760
but in the meantime

602
00:22:26,789 --> 00:22:24,320
dr mossbridge has also looked at

603
00:22:27,590 --> 00:22:26,799
some day night variation because

604
00:22:29,750 --> 00:22:27,600
uh

605
00:22:30,870 --> 00:22:29,760
remember the experiment she did she only

606
00:22:32,789 --> 00:22:30,880
measured during the day and i'm

607
00:22:34,310 --> 00:22:32,799
measuring day and night so that gives an

608
00:22:36,149 --> 00:22:34,320
opportunity to compare those so i'll

609
00:22:37,350 --> 00:22:36,159
show that in the next line one of the

610
00:22:38,149 --> 00:22:37,360
things that

611
00:22:40,230 --> 00:22:38,159
um

612
00:22:41,669 --> 00:22:40,240
these are slides generated by dr

613
00:22:42,789 --> 00:22:41,679
mossbridge that i've put into this

614
00:22:43,750 --> 00:22:42,799
presentation

615
00:22:46,070 --> 00:22:43,760
um

616
00:22:48,470 --> 00:22:46,080
one of the things that she noticed is

617
00:22:50,549 --> 00:22:48,480
that there's a significant difference

618
00:22:52,789 --> 00:22:50,559
between day and night data so looking

619
00:22:55,110 --> 00:22:52,799
over the whole

620
00:22:57,590 --> 00:22:55,120
six plus months of data

621
00:22:59,590 --> 00:22:57,600
um and this is sort of

622
00:23:01,510 --> 00:22:59,600
you know it's it's sort of sort of

623
00:23:03,669 --> 00:23:01,520
significant sort of in the noise we're

624
00:23:05,750 --> 00:23:03,679
not claiming the significance of this

625
00:23:06,950 --> 00:23:05,760
uh at this point but you see that

626
00:23:08,789 --> 00:23:06,960
there's this kind of interesting

627
00:23:10,470 --> 00:23:08,799
reversal

628
00:23:12,070 --> 00:23:10,480
that nighttime data has kind of an

629
00:23:13,270 --> 00:23:12,080
opposite character to date time data in

630
00:23:16,310 --> 00:23:13,280
a certain way

631
00:23:18,470 --> 00:23:16,320
in in how the how the data separates by

632
00:23:20,230 --> 00:23:18,480
trial duration

633
00:23:21,110 --> 00:23:20,240
we see that again here

634
00:23:22,470 --> 00:23:21,120
so

635
00:23:24,310 --> 00:23:22,480
horizontally are three different

636
00:23:25,590 --> 00:23:24,320
comparisons here this is

637
00:23:27,669 --> 00:23:25,600
the um

638
00:23:30,470 --> 00:23:27,679
what i call the 20s where there's

639
00:23:31,909 --> 00:23:30,480
20 counts after the decision and this is

640
00:23:33,510 --> 00:23:31,919
the zeros where there's zero counts

641
00:23:35,510 --> 00:23:33,520
after the decision you see the

642
00:23:37,510 --> 00:23:35,520
comparison there and then at night time

643
00:23:39,110 --> 00:23:37,520
you get the reverse of that

644
00:23:41,190 --> 00:23:39,120
well i should say this is

645
00:23:43,750 --> 00:23:41,200
data before the winter solstice so i'm

646
00:23:45,669 --> 00:23:43,760
going to show you two plots two charts

647
00:23:47,190 --> 00:23:45,679
of six plots each before the winter

648
00:23:49,510 --> 00:23:47,200
solstice and after the winter solstice

649
00:23:52,149 --> 00:23:49,520
okay so this is the first of two

650
00:23:53,350 --> 00:23:52,159
slides before the winter solstice

651
00:23:54,149 --> 00:23:53,360
um

652
00:23:56,149 --> 00:23:54,159
and

653
00:23:59,990 --> 00:23:56,159
you can see that there's

654
00:24:02,310 --> 00:24:00,000
this reversal right like and and here

655
00:24:05,029 --> 00:24:02,320
again more showing the reversal

656
00:24:07,029 --> 00:24:05,039
okay now after the winter solstice look

657
00:24:08,630 --> 00:24:07,039
at this

658
00:24:10,310 --> 00:24:08,640
this is in the same direction instead of

659
00:24:12,470 --> 00:24:10,320
reversed

660
00:24:13,909 --> 00:24:12,480
so this is just really a curious thing

661
00:24:16,390 --> 00:24:13,919
like we don't know why what this is

662
00:24:18,390 --> 00:24:16,400
about but it seems like something to

663
00:24:19,830 --> 00:24:18,400
investigate further and here's the other

664
00:24:22,070 --> 00:24:19,840
plot

665
00:24:25,110 --> 00:24:22,080
showing that as well

666
00:24:27,029 --> 00:24:25,120
you can see here maybe there's some

667
00:24:30,630 --> 00:24:27,039
uh it's not quite reversed maybe there's

668
00:24:34,549 --> 00:24:32,310
fluctuation in how these things relate

669
00:24:36,630 --> 00:24:34,559
to each other for some curves relating

670
00:24:38,230 --> 00:24:36,640
than others

671
00:24:39,510 --> 00:24:38,240
so to sum up

672
00:24:41,269 --> 00:24:39,520
um

673
00:24:43,029 --> 00:24:41,279
this has been a replication and

674
00:24:43,909 --> 00:24:43,039
extension

675
00:24:46,630 --> 00:24:43,919
um

676
00:24:47,909 --> 00:24:46,640
the up the apparatus has been upgraded

677
00:24:49,830 --> 00:24:47,919
um

678
00:24:51,350 --> 00:24:49,840
i've run day night collection instead of

679
00:24:52,149 --> 00:24:51,360
just during the days

680
00:24:53,990 --> 00:24:52,159
um

681
00:24:56,310 --> 00:24:54,000
it's possible that there's sensitivity

682
00:24:58,310 --> 00:24:56,320
to astronomical or emotional events that

683
00:25:00,310 --> 00:24:58,320
bears further investigation

684
00:25:03,190 --> 00:25:00,320
um and there's a lot of work to do in

685
00:25:04,789 --> 00:25:03,200
understanding these causally ambiguous

686
00:25:05,909 --> 00:25:04,799
duration sorting and why that's

687
00:25:06,950 --> 00:25:05,919
happening

688
00:25:09,350 --> 00:25:06,960
um

689
00:25:12,870 --> 00:25:09,360
so i'll close there and ask questions

690
00:25:16,549 --> 00:25:14,470
thank you so much for that really

691
00:25:18,950 --> 00:25:16,559
interesting presentation

692
00:25:21,110 --> 00:25:18,960
and we've got quite a lot of buzz in the

693
00:25:22,710 --> 00:25:21,120
chat so i'm going to go through some of

694
00:25:23,909 --> 00:25:22,720
those questions and

695
00:25:25,029 --> 00:25:23,919
um

696
00:25:28,149 --> 00:25:25,039
hopefully we can get through some of

697
00:25:31,350 --> 00:25:28,159
them in this time um let's see so the

698
00:25:34,710 --> 00:25:31,360
first question is about

699
00:25:36,950 --> 00:25:34,720
whether any of the variations in counts

700
00:25:38,870 --> 00:25:36,960
could be correlated with sun activity

701
00:25:40,390 --> 00:25:38,880
like sun spots or flares

702
00:25:42,710 --> 00:25:40,400
um

703
00:25:44,390 --> 00:25:42,720
yeah there's a slide here

704
00:25:47,830 --> 00:25:44,400
about um

705
00:25:48,870 --> 00:25:47,840
geomagnetic uh fluctuations right here

706
00:25:53,029 --> 00:25:48,880
um

707
00:25:54,470 --> 00:25:53,039
there's

708
00:25:57,110 --> 00:25:54,480
that would be a good thing to look at

709
00:25:59,590 --> 00:25:57,120
yes

710
00:26:02,310 --> 00:25:59,600
could i add to that just briefly

711
00:26:04,549 --> 00:26:02,320
yes please um so the donald are there

712
00:26:06,470 --> 00:26:04,559
geometric magnetic or these are yeah

713
00:26:08,390 --> 00:26:06,480
these are geomagnetic field fluctuations

714
00:26:10,710 --> 00:26:08,400
with the seasons you know the earth is

715
00:26:13,269 --> 00:26:10,720
in this little spot in the middle

716
00:26:14,789 --> 00:26:13,279
and so clearly even within each season

717
00:26:17,190 --> 00:26:14,799
well not clearly because you can't see

718
00:26:19,350 --> 00:26:17,200
but the sun is over here

719
00:26:21,350 --> 00:26:19,360
i don't know if you can see my mouse um

720
00:26:23,990 --> 00:26:21,360
no can't see your mouse yeah the sun

721
00:26:26,710 --> 00:26:24,000
would be over here right exactly and so

722
00:26:28,870 --> 00:26:26,720
at night time that's flipped

723
00:26:30,630 --> 00:26:28,880
so this shows the seasonal

724
00:26:32,230 --> 00:26:30,640
and by inference

725
00:26:34,549 --> 00:26:32,240
the day night difference so there could

726
00:26:37,190 --> 00:26:34,559
be a whole geomagnetic effect that could

727
00:26:38,789 --> 00:26:37,200
happen on the electrons in the detector

728
00:26:42,470 --> 00:26:38,799
as potentially as well as the medium

729
00:26:49,190 --> 00:26:43,590
thank you

730
00:26:51,830 --> 00:26:49,200
next question are you planning on

731
00:26:53,909 --> 00:26:51,840
exploring any pre-registered event

732
00:26:55,909 --> 00:26:53,919
correlations with these data patterns as

733
00:26:59,110 --> 00:26:55,919
with the gcp

734
00:27:01,029 --> 00:26:59,120
data that we know more about

735
00:27:02,630 --> 00:27:01,039
um i don't have

736
00:27:03,909 --> 00:27:02,640
i don't know what my plans are with this

737
00:27:07,590 --> 00:27:03,919
next exactly

738
00:27:09,110 --> 00:27:07,600
but that could be a good idea yeah yeah

739
00:27:10,870 --> 00:27:09,120
i would like to say like

740
00:27:12,789 --> 00:27:10,880
this we just discovered those very

741
00:27:15,430 --> 00:27:12,799
recently and that wasn't the main point

742
00:27:17,510 --> 00:27:15,440
but there's a lot of work to do

743
00:27:18,310 --> 00:27:17,520
yeah it's very interesting

744
00:27:19,110 --> 00:27:18,320
um

745
00:27:22,389 --> 00:27:19,120
see

746
00:27:25,269 --> 00:27:22,399
um do you have any information on what

747
00:27:26,230 --> 00:27:25,279
causes the diurnal variation what about

748
00:27:27,990 --> 00:27:26,240
the

749
00:27:29,830 --> 00:27:28,000
steep dips that seem to come more or

750
00:27:33,029 --> 00:27:29,840
less regularly especially in the first

751
00:27:35,110 --> 00:27:33,039
part of the data many days apart

752
00:27:37,750 --> 00:27:35,120
i don't know what the steep tips are

753
00:27:39,510 --> 00:27:37,760
um i i kept track of when i would go in

754
00:27:41,590 --> 00:27:39,520
the room to upload data and stuff like

755
00:27:43,190 --> 00:27:41,600
that or when i had the lights on there

756
00:27:45,269 --> 00:27:43,200
was one time where the lights got

757
00:27:46,950 --> 00:27:45,279
changed and so people other people were

758
00:27:49,669 --> 00:27:46,960
in the room for a little while changing

759
00:27:51,909 --> 00:27:49,679
the lights um and

760
00:27:54,230 --> 00:27:51,919
that's not strong that you know maybe

761
00:27:56,389 --> 00:27:54,240
one of the things is due to that but

762
00:27:58,549 --> 00:27:56,399
that doesn't seem to be a major cause

763
00:28:00,230 --> 00:27:58,559
as far as the diurnal variations could

764
00:28:01,990 --> 00:28:00,240
easily be temperature

765
00:28:03,430 --> 00:28:02,000
you know almost any instrumentation will

766
00:28:04,470 --> 00:28:03,440
vary with temperature

767
00:28:07,190 --> 00:28:04,480
um

768
00:28:08,630 --> 00:28:07,200
for the latter half of the the year

769
00:28:11,510 --> 00:28:08,640
uh i've now been running with a

770
00:28:14,070 --> 00:28:11,520
thermometer in the room uh logging so

771
00:28:17,110 --> 00:28:14,080
we'll see if that is correlated

772
00:28:19,510 --> 00:28:17,120
so watch this space yeah

773
00:28:21,110 --> 00:28:19,520
yeah this is very interesting it's very

774
00:28:21,990 --> 00:28:21,120
hot off the press as well

775
00:28:24,310 --> 00:28:22,000
um

776
00:28:27,029 --> 00:28:24,320
there was a major increase in sunspots

777
00:28:29,029 --> 00:28:27,039
and solar flux in november to december

778
00:28:30,789 --> 00:28:29,039
um in this person says it might be

779
00:28:33,190 --> 00:28:30,799
interesting to look at solar weather

780
00:28:34,950 --> 00:28:33,200
correlations as you go forward

781
00:28:35,990 --> 00:28:34,960
yes i think that's a good thing to look

782
00:28:39,110 --> 00:28:36,000
at

783
00:28:40,549 --> 00:28:39,120
actually because um

784
00:28:43,590 --> 00:28:40,559
who was the one who asked that question

785
00:28:45,110 --> 00:28:43,600
doug that was doug richards yeah doug

786
00:28:47,350 --> 00:28:45,120
talk to us maybe i mean we need someone

787
00:28:48,950 --> 00:28:47,360
who knows what they're doing there

788
00:28:50,389 --> 00:28:48,960
that'd be great

789
00:28:51,909 --> 00:28:50,399
this is the purpose of this conference

790
00:28:53,669 --> 00:28:51,919
to make connections

791
00:28:55,830 --> 00:28:53,679
insight so that's great

792
00:28:57,669 --> 00:28:55,840
um another question from york can you

793
00:28:59,750 --> 00:28:57,679
use the correlation measurement between

794
00:29:02,630 --> 00:28:59,760
the pre-decision measurement and the

795
00:29:04,630 --> 00:29:02,640
first 20 post decision values to predict

796
00:29:07,110 --> 00:29:04,640
whether the remaining duration after 20

797
00:29:09,110 --> 00:29:07,120
trials is an additional 0 10 or 40

798
00:29:12,630 --> 00:29:09,120
trials and with what statistical

799
00:29:14,630 --> 00:29:12,640
confidence can you make this prediction

800
00:29:17,029 --> 00:29:14,640
uh i haven't come up with a way to do

801
00:29:18,789 --> 00:29:17,039
that uh i don't know if dr mossbridge

802
00:29:20,950 --> 00:29:18,799
has

803
00:29:22,789 --> 00:29:20,960
um yeah i'm working on that that's not

804
00:29:25,350 --> 00:29:22,799
in this paper but that's a that's

805
00:29:27,590 --> 00:29:25,360
obviously an interesting topic

806
00:29:29,830 --> 00:29:27,600
yeah lots of lots of fascinating

807
00:29:33,110 --> 00:29:29,840
possibilities here um

808
00:29:34,870 --> 00:29:33,120
can you both say anything about um what

809
00:29:36,549 --> 00:29:34,880
the next steps are then with this

810
00:29:38,830 --> 00:29:36,559
research agenda what what are you

811
00:29:41,269 --> 00:29:38,840
planning

812
00:29:43,750 --> 00:29:41,279
next well

813
00:29:45,350 --> 00:29:43,760
like i said we've got uh more

814
00:29:46,630 --> 00:29:45,360
additional data logging running since

815
00:29:49,350 --> 00:29:46,640
february

816
00:29:50,950 --> 00:29:49,360
we'll see if that tells us anything

817
00:29:52,630 --> 00:29:50,960
um

818
00:29:53,990 --> 00:29:52,640
we've got that whole set of data that i

819
00:29:57,029 --> 00:29:54,000
haven't looked at yet i've just been

820
00:29:58,389 --> 00:29:57,039
uploading it and backing it up

821
00:29:59,510 --> 00:29:58,399
um

822
00:30:01,190 --> 00:29:59,520
there's

823
00:30:03,350 --> 00:30:01,200
uh

824
00:30:06,149 --> 00:30:03,360
lots of different ways we could

825
00:30:07,990 --> 00:30:06,159
set up a photon counting

826
00:30:09,669 --> 00:30:08,000
and it may also be interesting to look

827
00:30:11,830 --> 00:30:09,679
at

828
00:30:14,389 --> 00:30:11,840
fermions instead of bosons

829
00:30:19,029 --> 00:30:14,399
so if we were to count electrons

830
00:30:22,789 --> 00:30:21,350
julia

831
00:30:24,070 --> 00:30:22,799
um i think those are interesting

832
00:30:25,590 --> 00:30:24,080
questions i think there are also

833
00:30:26,870 --> 00:30:25,600
interesting questions about when the

834
00:30:31,110 --> 00:30:26,880
knowledge

835
00:30:32,630 --> 00:30:31,120
the duration occurs um sort of an

836
00:30:34,630 --> 00:30:32,640
ontological

837
00:30:37,029 --> 00:30:34,640
versus epistemological sort of

838
00:30:39,830 --> 00:30:37,039
interpretation of quantum mechanics type

839
00:30:41,190 --> 00:30:39,840
of differentiation um

840
00:30:43,110 --> 00:30:41,200
i think

841
00:30:44,549 --> 00:30:43,120
basically publishing what we have so the

842
00:30:46,389 --> 00:30:44,559
first paper

843
00:30:48,830 --> 00:30:46,399
that i have out there and the conference

844
00:30:50,389 --> 00:30:48,840
proceedings are on like crappy equipment

845
00:30:56,389 --> 00:30:50,399
and

846
00:30:57,669 --> 00:30:56,399
i'm just grateful that i was actually

847
00:30:59,669 --> 00:30:57,679
looking something real and not an

848
00:31:01,750 --> 00:30:59,679
artifact and um

849
00:31:04,230 --> 00:31:01,760
so that's exciting and and sort of

850
00:31:06,549 --> 00:31:04,240
uncovering it more under i i think these

851
00:31:09,350 --> 00:31:06,559
these these seasonal

852
00:31:11,990 --> 00:31:09,360
day night variations are profound

853
00:31:13,430 --> 00:31:12,000
and especially not just the overall

854
00:31:15,110 --> 00:31:13,440
in the overall data but in the actual

855
00:31:16,070 --> 00:31:15,120
relationship between the durations i

856
00:31:16,870 --> 00:31:16,080
think that

857
00:31:20,710 --> 00:31:16,880
the

858
00:31:23,269 --> 00:31:20,720
imagine kind of like different weighted

859
00:31:24,870 --> 00:31:23,279
ropes and oscillating at different

860
00:31:26,950 --> 00:31:24,880
frequencies based on the weight and the

861
00:31:29,269 --> 00:31:26,960
weight is set by the duration sort of

862
00:31:32,149 --> 00:31:29,279
and so i think that that's how i

863
00:31:34,070 --> 00:31:32,159
envision it with no actual

864
00:31:37,110 --> 00:31:34,080
data but just sort of living with this

865
00:31:38,789 --> 00:31:37,120
experiment for a couple of years

866
00:31:41,029 --> 00:31:38,799
that's a good answer

867
00:31:43,909 --> 00:31:41,039
yeah i'll add that

868
00:31:47,190 --> 00:31:43,919
the random numbers in this experiment

869
00:31:48,710 --> 00:31:47,200
are selected once for the the

870
00:31:50,630 --> 00:31:48,720
choice of the four durations it would be

871
00:31:52,149 --> 00:31:50,640
interesting to try

872
00:31:54,789 --> 00:31:52,159
having the first random number just

873
00:31:57,190 --> 00:31:54,799
select whether the experiment concludes

874
00:31:59,669 --> 00:31:57,200
at that moment or not and then draw

875
00:32:01,590 --> 00:31:59,679
another random number to determine

876
00:32:03,830 --> 00:32:01,600
what happens next and whether the

877
00:32:05,990 --> 00:32:03,840
duration goes on longer and then between

878
00:32:08,389 --> 00:32:06,000
the 30 and 60 choice draw another random

879
00:32:10,389 --> 00:32:08,399
number yeah

880
00:32:12,230 --> 00:32:10,399
and there's some indications that if you

881
00:32:14,149 --> 00:32:12,240
don't choose a random duration in other

882
00:32:15,430 --> 00:32:14,159
words at least this is different we

883
00:32:17,430 --> 00:32:15,440
didn't show you this data but i'm just

884
00:32:20,149 --> 00:32:17,440
telling you based on my own observations

885
00:32:22,470 --> 00:32:20,159
um if you don't choose a random duration

886
00:32:24,470 --> 00:32:22,480
you actually might not get this duration

887
00:32:25,909 --> 00:32:24,480
sorting effect at all

888
00:32:27,570 --> 00:32:25,919
so it might be that there has to be a

889
00:32:29,590 --> 00:32:27,580
choice

890
00:32:31,830 --> 00:32:29,600
[Music]

891
00:32:33,590 --> 00:32:31,840
fascinating stuff thank you once again