1
00:00:01,760 --> 00:00:03,679
Join us as we gather around the hedge,

2
00:00:03,679 --> 00:00:06,660
where we dig into technology, business, and culture

3
00:00:06,719 --> 00:00:09,139
with the finest minds in computer networking.

4
00:00:20,175 --> 00:00:22,355
Hello, Tom. Hello, Tom's plant.

5
00:00:23,454 --> 00:00:24,515
Good morning, Russ.

6
00:00:24,894 --> 00:00:26,250
The plant says good morning.

7
00:00:28,089 --> 00:00:30,109
We need a we need a plant interpreter.

8
00:00:30,570 --> 00:00:32,190
You need to put, like, a big,

9
00:00:32,729 --> 00:00:35,530
like, a big device around it and say,

10
00:00:35,530 --> 00:00:37,689
this is the plant interpreter. I can put

11
00:00:37,689 --> 00:00:39,369
it up here so we can hear what

12
00:00:39,369 --> 00:00:39,770
the plant

13
00:00:40,329 --> 00:00:42,994
it would look very, you know, sci fi

14
00:00:43,134 --> 00:00:45,854
looking. Yeah. Yeah. Well and we I'd like

15
00:00:45,854 --> 00:00:47,155
to announce that Yvonne,

16
00:00:47,454 --> 00:00:49,054
for anyone who didn't listen to the last

17
00:00:49,054 --> 00:00:51,534
episode, Yvonne has named the plant. The plant

18
00:00:51,534 --> 00:00:53,395
shall be called Audrey henceforth.

19
00:00:55,229 --> 00:00:58,510
Audrey. I like Audrey. I think that's really

20
00:00:58,510 --> 00:01:00,929
great. Do you know where it's from, George?

21
00:01:02,350 --> 00:01:03,869
Well, it could be from a lot of

22
00:01:03,869 --> 00:01:05,489
places. It's from a movie.

23
00:01:07,390 --> 00:01:09,170
I'm sure Audrey's in a movie somewhere.

24
00:01:09,694 --> 00:01:11,215
It's in the it's the name of the

25
00:01:11,215 --> 00:01:12,754
plant in Little Shop of Horrors.

26
00:01:13,295 --> 00:01:15,215
Oh, is that the oh, perfect. That's what

27
00:01:15,215 --> 00:01:16,354
it is. Yeah.

28
00:01:18,094 --> 00:01:20,415
I I think the plant interpreter wouldn't be

29
00:01:20,415 --> 00:01:22,974
some amazing high-tech Gizmo. It'd be like two

30
00:01:22,974 --> 00:01:23,474
wires,

31
00:01:24,210 --> 00:01:26,609
crinkly wires with crocodile clips, and one of

32
00:01:26,609 --> 00:01:28,689
them's had to be twisted on. You know?

33
00:01:28,689 --> 00:01:30,870
And it goes through a plant o matic

34
00:01:30,930 --> 00:01:32,150
meter that goes,

35
00:01:32,530 --> 00:01:34,870
yes, no, or good, bad.

36
00:01:35,329 --> 00:01:37,409
I I don't think plants have a higher

37
00:01:37,409 --> 00:01:39,170
logic going on here with a lot of

38
00:01:39,170 --> 00:01:39,670
interpreting.

39
00:01:45,465 --> 00:01:48,525
Water. Yeah. That that could be. And

40
00:01:48,905 --> 00:01:50,365
hello. Good morning, George.

41
00:01:50,825 --> 00:01:51,564
Good morning.

42
00:01:52,260 --> 00:01:54,579
Oh my goodness. Yeah. We got George Johnson.

43
00:01:54,979 --> 00:01:56,840
Here. No plants here?

44
00:01:57,540 --> 00:01:59,540
Well, if they were there, they'd be hanging

45
00:01:59,540 --> 00:02:01,879
from the ceiling in our perspective. They would.

46
00:02:02,340 --> 00:02:03,159
They would.

47
00:02:04,474 --> 00:02:06,715
Do not ask about the shark attacks. Why

48
00:02:06,715 --> 00:02:09,034
do we get three in one week? I

49
00:02:09,034 --> 00:02:10,974
mean, that is just insane.

50
00:02:11,514 --> 00:02:12,175
You know?

51
00:02:12,634 --> 00:02:16,254
Oh, interesting. Anyway Not good. Alright. So

52
00:02:16,900 --> 00:02:18,659
let's talk about what comes after five g.

53
00:02:18,659 --> 00:02:19,860
Now, we're gonna have a couple of shows

54
00:02:19,860 --> 00:02:22,020
here on five g because and what comes

55
00:02:22,020 --> 00:02:23,879
after because it's an interesting topic,

56
00:02:24,259 --> 00:02:26,039
and it does truly impact

57
00:02:26,500 --> 00:02:27,400
network engineering,

58
00:02:27,780 --> 00:02:29,719
particularly when you live in a

59
00:02:30,224 --> 00:02:31,125
resource constrained

60
00:02:31,425 --> 00:02:31,925
environment

61
00:02:32,705 --> 00:02:33,205
where

62
00:02:33,504 --> 00:02:34,245
some people

63
00:02:34,705 --> 00:02:36,305
may not be able to build the Wi

64
00:02:36,305 --> 00:02:38,224
Fi network they want. They may have to

65
00:02:38,224 --> 00:02:39,685
use five g or,

66
00:02:40,305 --> 00:02:41,044
god forbid,

67
00:02:41,425 --> 00:02:42,245
six g

68
00:02:42,860 --> 00:02:44,159
or 18 g

69
00:02:44,459 --> 00:02:47,259
or 99 g He gave away he gave

70
00:02:47,259 --> 00:02:48,879
away the whole show.

71
00:02:51,019 --> 00:02:52,159
What are you doing?

72
00:02:53,900 --> 00:02:56,234
So, George, talk to us. What happens after

73
00:02:56,234 --> 00:02:58,314
five g? And and and and do we

74
00:02:58,314 --> 00:02:58,814
care?

75
00:03:00,715 --> 00:03:01,215
So

76
00:03:01,514 --> 00:03:04,174
what happens after five g is the inductive

77
00:03:04,234 --> 00:03:07,275
reasoning from asking the question, what happens after

78
00:03:07,275 --> 00:03:08,094
two g?

79
00:03:09,150 --> 00:03:11,169
And the do we care part

80
00:03:11,550 --> 00:03:12,050
is

81
00:03:12,909 --> 00:03:13,409
yes

82
00:03:13,949 --> 00:03:17,229
and probably no. And so I kinda wanna

83
00:03:17,229 --> 00:03:19,629
roll I wanna roll the clock back. You

84
00:03:19,629 --> 00:03:20,129
know?

85
00:03:20,965 --> 00:03:23,685
Image of the calendar not peeling away, image

86
00:03:23,685 --> 00:03:26,324
of the calendar growing up as we put

87
00:03:26,324 --> 00:03:28,185
the leaves back on the sheet

88
00:03:28,884 --> 00:03:31,465
and go back to the heady days of

89
00:03:32,324 --> 00:03:32,824
AMPS,

90
00:03:33,290 --> 00:03:33,950
the predigital

91
00:03:34,650 --> 00:03:37,230
mobile telephony world. Oh my goodness.

92
00:03:38,250 --> 00:03:40,990
You remember those days? Oh my goodness. Yes.

93
00:03:41,210 --> 00:03:41,710
Continue.

94
00:03:42,569 --> 00:03:45,290
They they were actually pretty good pretty good

95
00:03:45,290 --> 00:03:47,469
days. And if you wanna talk about constraints,

96
00:03:48,525 --> 00:03:51,004
think about the constraint of the size of

97
00:03:51,004 --> 00:03:53,164
the handbag you need to put one of

98
00:03:53,164 --> 00:03:54,944
those funky phones in.

99
00:03:55,245 --> 00:03:57,025
I mean, they were big. Right?

100
00:03:57,324 --> 00:04:00,284
So the actual constraint in those days was

101
00:04:00,284 --> 00:04:01,264
really essentially

102
00:04:02,000 --> 00:04:05,439
space. These devices were large and powerful. Yeah.

103
00:04:05,439 --> 00:04:06,419
Yeah. Yeah. And lifetime.

104
00:04:06,799 --> 00:04:08,259
Yeah. And lifetime. And Yeah.

105
00:04:09,039 --> 00:04:09,859
This technology

106
00:04:10,400 --> 00:04:12,579
was about as straightforward

107
00:04:12,879 --> 00:04:15,205
as you can get. You take a microphone

108
00:04:15,985 --> 00:04:17,985
and a speaker, and you bang them onto

109
00:04:17,985 --> 00:04:20,625
some circuitry that plugs them out on the

110
00:04:20,625 --> 00:04:21,125
radio,

111
00:04:21,985 --> 00:04:22,485
and

112
00:04:23,024 --> 00:04:24,404
that's kind of it, guys.

113
00:04:24,705 --> 00:04:26,464
Yeah. In fact a mic, you attach it

114
00:04:26,464 --> 00:04:27,584
to a front end to a front end

115
00:04:27,745 --> 00:04:29,665
not even a front end processor. You attach

116
00:04:29,665 --> 00:04:30,350
it to

117
00:04:30,889 --> 00:04:31,389
basically

118
00:04:32,170 --> 00:04:33,230
a a tank circuit

119
00:04:33,930 --> 00:04:36,270
that modulates a carrier frequency,

120
00:04:37,129 --> 00:04:40,189
and off you go. That's the whole system.

121
00:04:40,889 --> 00:04:43,149
And if you're prince Charles and his girlfriend

122
00:04:43,290 --> 00:04:46,004
Dujour and you wanna take talk sweet nothings

123
00:04:46,004 --> 00:04:48,084
to each other over a mobile network, hey.

124
00:04:48,084 --> 00:04:50,084
You can do that. And the guy sitting

125
00:04:50,084 --> 00:04:52,245
in a car two blocks away with a

126
00:04:52,245 --> 00:04:55,064
journalist with him can tune in and listen.

127
00:04:55,524 --> 00:04:57,464
It it wasn't a private conversation.

128
00:04:58,245 --> 00:04:59,785
It was just radio, man.

129
00:05:00,220 --> 00:05:00,720
So

130
00:05:02,139 --> 00:05:04,779
what we have been engaged in from those

131
00:05:04,779 --> 00:05:05,600
heady days

132
00:05:06,220 --> 00:05:06,959
is essentially

133
00:05:07,500 --> 00:05:08,720
stepping up ladders.

134
00:05:09,339 --> 00:05:11,100
We can do better. Right? We can do

135
00:05:11,100 --> 00:05:12,620
better than this. Why are we doing this

136
00:05:12,620 --> 00:05:14,779
analog stuff when we could be doing digital

137
00:05:14,779 --> 00:05:17,865
encoding and have constant strength radio instead of

138
00:05:17,865 --> 00:05:19,324
all this amplitude modification

139
00:05:19,625 --> 00:05:21,084
nonsense, when we could have

140
00:05:21,384 --> 00:05:24,185
encrypted contents, when we could packetize this thing,

141
00:05:24,185 --> 00:05:25,884
when we could do all these additions?

142
00:05:26,345 --> 00:05:27,324
And the g's,

143
00:05:28,120 --> 00:05:30,199
two g. Interesting it kicks off with two

144
00:05:30,199 --> 00:05:32,199
g. Why did nobody wanna go with one

145
00:05:32,199 --> 00:05:33,259
g as the branding?

146
00:05:33,879 --> 00:05:36,360
And it's because it's the second generation. Right?

147
00:05:36,360 --> 00:05:37,259
The first generation,

148
00:05:37,959 --> 00:05:39,399
you don't have a name for it. It's

149
00:05:39,399 --> 00:05:41,274
just what it is. But when you're doing

150
00:05:41,274 --> 00:05:44,394
version two, you're in the second generation. So

151
00:05:44,394 --> 00:05:46,794
two g is kind of the entry point

152
00:05:46,794 --> 00:05:48,794
for a world where we were thinking, maybe

153
00:05:48,794 --> 00:05:51,435
it's not about people speaking. Maybe it's actually

154
00:05:51,435 --> 00:05:52,334
about data.

155
00:05:54,470 --> 00:05:56,069
And I don't know about you, but I

156
00:05:56,069 --> 00:05:58,169
found two g pretty unsatisfying.

157
00:05:59,589 --> 00:06:01,669
Do you remember an ETF when,

158
00:06:02,789 --> 00:06:05,910
the WAP protocol was gonna be shown off

159
00:06:05,910 --> 00:06:08,044
to us, and a guy came on stage

160
00:06:08,044 --> 00:06:11,504
to demonstrate the amazing technology that was gonna

161
00:06:11,644 --> 00:06:13,805
take over the world and be revolutionary. And

162
00:06:13,805 --> 00:06:15,964
there had been so much energy put into

163
00:06:15,964 --> 00:06:17,345
it called WAP.

164
00:06:17,725 --> 00:06:20,990
And straight after him, a lovely Japanese engineer

165
00:06:20,990 --> 00:06:22,589
got up and said, hey. We've been working

166
00:06:22,589 --> 00:06:24,829
on this idea in Japan where we're just

167
00:06:24,829 --> 00:06:26,610
gonna put IP in the phones.

168
00:06:26,990 --> 00:06:29,970
It's IP. There's no magic. It's just IP.

169
00:06:30,269 --> 00:06:32,930
And the room got up and went, yes.

170
00:06:35,404 --> 00:06:37,324
And the WAP engineer is like, oh my

171
00:06:37,324 --> 00:06:39,345
god. How am I gonna tell my bosses

172
00:06:39,485 --> 00:06:41,964
that they don't want what we've just sold

173
00:06:41,964 --> 00:06:42,464
them?

174
00:06:43,165 --> 00:06:44,464
WAP was a disaster.

175
00:06:45,165 --> 00:06:46,545
So two gs,

176
00:06:46,949 --> 00:06:49,529
kind of interesting, cut over to full digital

177
00:06:49,750 --> 00:06:50,250
ish,

178
00:06:51,110 --> 00:06:51,610
virtually

179
00:06:51,910 --> 00:06:52,410
zero

180
00:06:52,870 --> 00:06:55,849
real energy behind making this technology interesting.

181
00:06:56,629 --> 00:06:57,129
And

182
00:06:58,389 --> 00:07:00,789
I don't wanna go there this strongly, but,

183
00:07:00,789 --> 00:07:01,289
essentially,

184
00:07:01,985 --> 00:07:03,524
it's all about patents.

185
00:07:03,824 --> 00:07:05,845
It's totally about

186
00:07:06,305 --> 00:07:08,865
patents. We are used to working in a

187
00:07:08,865 --> 00:07:11,584
technology space where people are doing patent free

188
00:07:11,584 --> 00:07:12,805
technology design,

189
00:07:13,264 --> 00:07:15,044
and it's available for

190
00:07:15,829 --> 00:07:17,050
round, reasonable,

191
00:07:18,310 --> 00:07:18,970
and neutral

192
00:07:19,509 --> 00:07:21,990
Yeah. Technology. Right. Yeah. It's like a level

193
00:07:21,990 --> 00:07:23,689
playing field. Well, three g

194
00:07:24,550 --> 00:07:25,050
is

195
00:07:25,509 --> 00:07:27,129
the worldwide telephony

196
00:07:27,675 --> 00:07:30,794
community coming together saying, man, it would just

197
00:07:30,794 --> 00:07:32,954
totally suck if we wound up deep in

198
00:07:32,954 --> 00:07:35,194
a patent war. Wouldn't it be better if

199
00:07:35,194 --> 00:07:36,875
we all agreed that we were gonna work

200
00:07:36,875 --> 00:07:37,935
to own the patents

201
00:07:38,555 --> 00:07:40,794
and agree a technology base to do this?

202
00:07:40,794 --> 00:07:43,560
Because this market I mean, we've sold, like,

203
00:07:43,560 --> 00:07:44,060
what,

204
00:07:44,759 --> 00:07:47,399
one telephone for every 20 people? This is

205
00:07:47,399 --> 00:07:47,899
insane.

206
00:07:48,279 --> 00:07:50,279
We have so much growth ahead of us.

207
00:07:50,279 --> 00:07:52,680
Let's build a bigger market, guys. Let's actually

208
00:07:52,680 --> 00:07:55,639
make this technology fly a bit. So that

209
00:07:55,639 --> 00:07:57,659
cutover, two g to three g,

210
00:07:57,964 --> 00:08:01,404
that was really interesting because it's like, wow.

211
00:08:01,404 --> 00:08:03,964
People want bits. They don't actually wanna talk

212
00:08:03,964 --> 00:08:06,125
to each other. We can do bits. We've

213
00:08:06,125 --> 00:08:07,425
been doing bits forever.

214
00:08:07,884 --> 00:08:09,025
So three g

215
00:08:09,884 --> 00:08:12,764
was like this amazing moment where people went,

216
00:08:12,764 --> 00:08:14,449
okay. It's gonna be digital.

217
00:08:14,910 --> 00:08:17,650
Let's really make digital do something interesting.

218
00:08:18,029 --> 00:08:20,210
It was like for an old timer technology

219
00:08:20,509 --> 00:08:22,370
hand converting from slip

220
00:08:22,670 --> 00:08:25,170
as the protocol to manage bits on wire

221
00:08:25,230 --> 00:08:25,889
to PPP.

222
00:08:26,509 --> 00:08:27,009
Right?

223
00:08:27,365 --> 00:08:29,285
We actually need a frame here. We need

224
00:08:29,285 --> 00:08:32,164
a frame structure that says something. We need

225
00:08:32,164 --> 00:08:33,845
a way to say this packet is from

226
00:08:33,845 --> 00:08:36,085
a foreign network and I'm transiting it. We

227
00:08:36,085 --> 00:08:37,445
need a way to say that it's got

228
00:08:37,445 --> 00:08:40,404
these behaviors. We've got interesting codecs. We need

229
00:08:40,404 --> 00:08:42,089
a way to signal the codecs. Do you

230
00:08:42,089 --> 00:08:44,409
remember in three g when you'd have a

231
00:08:44,409 --> 00:08:46,809
phone call and one sender used one codec

232
00:08:46,809 --> 00:08:49,049
and the other guy used another codec, and

233
00:08:49,049 --> 00:08:51,289
the two wouldn't talk to each other? And

234
00:08:51,289 --> 00:08:53,389
you'd have these weird things going on.

235
00:08:53,690 --> 00:08:56,029
I mean, people were having fun. Right?

236
00:08:56,544 --> 00:08:58,565
And they were exploring things like,

237
00:08:59,264 --> 00:09:01,445
how small can we make the base station?

238
00:09:01,585 --> 00:09:03,504
How small can we make the power burden

239
00:09:03,504 --> 00:09:06,225
here? How do we make this legacy idea

240
00:09:06,225 --> 00:09:09,585
of text messages, which was only an add

241
00:09:09,585 --> 00:09:10,725
on using

242
00:09:11,309 --> 00:09:13,789
spare bits in the framing from when they'd

243
00:09:13,789 --> 00:09:16,910
originally designed this technology and brought paging into

244
00:09:16,910 --> 00:09:19,230
the modern era. How do we design this

245
00:09:19,230 --> 00:09:19,730
stuff?

246
00:09:20,750 --> 00:09:21,649
The three g,

247
00:09:22,990 --> 00:09:26,115
you couldn't do multiple protocols well. It was

248
00:09:26,115 --> 00:09:29,214
actually pretty difficult to live in a multiprotocol

249
00:09:29,235 --> 00:09:32,035
world, and we were emerging into this dual

250
00:09:32,035 --> 00:09:35,634
stack Internet world. We needed framing that could

251
00:09:35,634 --> 00:09:36,134
actually

252
00:09:36,995 --> 00:09:39,634
handle four and six, and we needed to

253
00:09:39,634 --> 00:09:40,375
talk about

254
00:09:40,754 --> 00:09:43,470
using a complex substrate of digital data.

255
00:09:43,850 --> 00:09:46,889
So three g actually wound up pretty early

256
00:09:46,889 --> 00:09:47,709
on being

257
00:09:48,809 --> 00:09:50,029
not quite good enough.

258
00:09:50,409 --> 00:09:52,829
Plus there's a second thing going on here.

259
00:09:53,784 --> 00:09:57,065
When you start selling phones at volume to

260
00:09:57,065 --> 00:09:58,044
the whole world,

261
00:09:58,664 --> 00:10:01,544
you need to think really carefully about how

262
00:10:01,544 --> 00:10:04,184
much capacity you are gonna need inside the

263
00:10:04,184 --> 00:10:05,565
radio waves. Right?

264
00:10:06,039 --> 00:10:07,960
I mean, you gotta plan for the future.

265
00:10:07,960 --> 00:10:10,360
You have to behave as if you're gonna

266
00:10:10,360 --> 00:10:11,960
be in a world when you've doubled the

267
00:10:11,960 --> 00:10:14,600
market share and then doubled the market share

268
00:10:14,600 --> 00:10:17,480
and doubled it again. How many bands do

269
00:10:17,480 --> 00:10:20,360
you need to have people using a device

270
00:10:20,360 --> 00:10:20,985
like that?

271
00:10:21,544 --> 00:10:24,184
So old school telephony, it was based on

272
00:10:24,184 --> 00:10:26,504
the idea of scarcity. Oh, you don't wanna

273
00:10:26,504 --> 00:10:28,745
be sending those bits, man. They are so

274
00:10:28,745 --> 00:10:30,985
expensive. So we're gonna send bits, and then

275
00:10:30,985 --> 00:10:32,524
we'll stop. And

276
00:10:33,379 --> 00:10:34,980
if you lived in a world where you

277
00:10:34,980 --> 00:10:36,899
were in the real world and you didn't

278
00:10:36,899 --> 00:10:39,399
care about data, that was just peachy.

279
00:10:39,940 --> 00:10:42,179
Now you start to make data so good

280
00:10:42,179 --> 00:10:43,940
that people start saying, oh, I think I

281
00:10:43,940 --> 00:10:45,720
might leave it on like, you know,

282
00:10:46,034 --> 00:10:47,914
stations tuned to country and western. I'm on

283
00:10:47,914 --> 00:10:50,195
a long drive along the road. People are

284
00:10:50,195 --> 00:10:52,434
leaving their data on. Are you using it?

285
00:10:52,434 --> 00:10:54,054
No. But I'm I'm on.

286
00:10:54,754 --> 00:10:57,815
So those bandwidths that you planned the scaling

287
00:10:57,954 --> 00:11:00,195
function that you'd have enough bandwidth for one

288
00:11:00,195 --> 00:11:03,850
phone conversation per person per day. No.

289
00:11:04,389 --> 00:11:07,029
That bandwidth is being used permanently. You need

290
00:11:07,029 --> 00:11:09,190
to make those bands smaller. You need a

291
00:11:09,190 --> 00:11:11,590
way to get more data down. So what's

292
00:11:11,590 --> 00:11:13,690
the classic way in the radio spectrum

293
00:11:13,990 --> 00:11:15,690
that you get more data

294
00:11:16,125 --> 00:11:18,465
into more bands. You got two tricks.

295
00:11:19,325 --> 00:11:20,865
You make the bands thinner,

296
00:11:21,245 --> 00:11:23,665
and you have to start using smarter encoding

297
00:11:23,804 --> 00:11:25,904
so that they don't collide with each other.

298
00:11:26,365 --> 00:11:28,759
And you move up the frequency map because

299
00:11:28,759 --> 00:11:29,980
the higher the frequency,

300
00:11:30,679 --> 00:11:32,620
the higher the information density

301
00:11:33,000 --> 00:11:34,299
inside that coding.

302
00:11:34,840 --> 00:11:37,080
Higher frequency just means that I can actually

303
00:11:37,080 --> 00:11:39,419
put more bits or more symbols

304
00:11:39,879 --> 00:11:40,620
per second

305
00:11:41,504 --> 00:11:43,904
Per Per per second per wavelength because my

306
00:11:43,985 --> 00:11:45,845
yep. Because my wavelength is shorter.

307
00:11:46,545 --> 00:11:49,365
And it also means my sidebands are narrower.

308
00:11:50,144 --> 00:11:52,884
Yeah. Okay. So the overlaps and the boundaries

309
00:11:53,105 --> 00:11:56,065
between things and you shave everything down, the

310
00:11:56,065 --> 00:11:58,709
engineering gets precision. We're getting a bit formula

311
00:11:58,709 --> 00:12:00,169
one here. Oh,

312
00:12:00,549 --> 00:12:03,029
hang on a moment. But but yours but

313
00:12:03,029 --> 00:12:03,850
your length,

314
00:12:04,549 --> 00:12:07,449
the if you lose three dB every wavelength,

315
00:12:08,309 --> 00:12:10,809
your power ratio goes way down,

316
00:12:11,134 --> 00:12:12,034
and you can't penetrate

317
00:12:12,414 --> 00:12:14,654
solid objects as well with low or higher

318
00:12:14,654 --> 00:12:16,034
bandwidth. What are we gonna do

319
00:12:16,654 --> 00:12:19,455
if there aren't enough mathematicians in the world

320
00:12:19,455 --> 00:12:22,095
to do smart coding to get data into

321
00:12:22,095 --> 00:12:24,335
narrower and narrower bandwidth? We need to start

322
00:12:24,335 --> 00:12:25,475
training more mathematicians.

323
00:12:26,230 --> 00:12:28,149
So you gotta plan long term when you

324
00:12:28,149 --> 00:12:29,129
start doing this.

325
00:12:29,750 --> 00:12:30,250
So

326
00:12:30,870 --> 00:12:32,250
three g to four g.

327
00:12:34,629 --> 00:12:36,789
What about all those handsets we sold that

328
00:12:36,789 --> 00:12:38,789
were doing two g? Do they still live

329
00:12:38,789 --> 00:12:40,330
in this world? Well,

330
00:12:40,804 --> 00:12:43,204
no. Because those two g handsets, you turn

331
00:12:43,204 --> 00:12:45,704
one of those babies on, and it's spreading

332
00:12:45,845 --> 00:12:48,024
data out to the edges of the signal.

333
00:12:48,325 --> 00:12:51,445
It's flooding the zone with bands. That's not

334
00:12:51,445 --> 00:12:53,830
good use of everyone's time. How about we

335
00:12:53,830 --> 00:12:55,669
tell everyone that two g is gonna get

336
00:12:55,669 --> 00:12:57,190
turned down and that you've got to be

337
00:12:57,190 --> 00:12:59,210
either three g or four g now?

338
00:12:59,590 --> 00:13:00,330
And then

339
00:13:00,710 --> 00:13:03,350
we could reuse those two g signal spaces

340
00:13:03,350 --> 00:13:04,950
to do things now that we know how

341
00:13:04,950 --> 00:13:06,730
to use tighter bands,

342
00:13:07,154 --> 00:13:09,714
narrower gaps between them, guardrails can be thinner.

343
00:13:09,714 --> 00:13:11,954
We could still use that because there's this

344
00:13:11,954 --> 00:13:13,574
horrible story emerging.

345
00:13:14,034 --> 00:13:15,975
Higher the frequency, more data,

346
00:13:16,355 --> 00:13:17,174
higher the frequency,

347
00:13:17,794 --> 00:13:20,214
less penetration through bricks and mortar.

348
00:13:20,589 --> 00:13:22,589
And higher power required to make the same

349
00:13:22,589 --> 00:13:23,809
distance. Higher power.

350
00:13:24,190 --> 00:13:26,370
So four g comes out,

351
00:13:26,750 --> 00:13:29,230
and we're starting to say it's digital. So

352
00:13:29,230 --> 00:13:30,529
the voice is encoded

353
00:13:31,230 --> 00:13:32,610
into digital bits.

354
00:13:33,225 --> 00:13:33,725
Now

355
00:13:34,345 --> 00:13:36,985
when four g was on the horizon, I

356
00:13:36,985 --> 00:13:39,384
went to some meetings where engineers were talking

357
00:13:39,384 --> 00:13:41,725
about the brave new world of four g,

358
00:13:42,264 --> 00:13:43,725
and they were telling us

359
00:13:44,184 --> 00:13:45,964
four g is gonna solve

360
00:13:46,309 --> 00:13:48,809
every problem you've got. It's the housewives

361
00:13:49,269 --> 00:13:51,929
housewives dream. It's the microwave cooking

362
00:13:52,549 --> 00:13:53,529
of our era.

363
00:13:53,830 --> 00:13:56,309
No more coal fires. You're gonna love four

364
00:13:56,309 --> 00:13:59,110
g. It's clean. It's new. It cooks your

365
00:13:59,110 --> 00:14:01,209
chicken in thirty seconds with radar.

366
00:14:02,964 --> 00:14:05,445
And their version of that was telling us

367
00:14:05,445 --> 00:14:07,865
that we were gonna have this amazing ability

368
00:14:07,924 --> 00:14:11,284
to roam between Wi Fi and cellular, and

369
00:14:11,284 --> 00:14:14,084
we'd carry our conversations with us. And it

370
00:14:14,084 --> 00:14:17,160
would be doing this thing. You've never heard

371
00:14:17,160 --> 00:14:19,259
of this thing. It's called beamforming.

372
00:14:20,040 --> 00:14:22,540
What we do is we put trained ants

373
00:14:22,600 --> 00:14:25,960
inside every telephone and base station, and they

374
00:14:25,960 --> 00:14:27,100
swing the dishes

375
00:14:27,705 --> 00:14:30,045
to make your phone point at the tower

376
00:14:30,264 --> 00:14:32,665
so that the beams go just to you

377
00:14:32,665 --> 00:14:35,485
and we can have more customers with beams.

378
00:14:35,945 --> 00:14:37,245
Four g is beams.

379
00:14:37,865 --> 00:14:39,165
Four g is mobile

380
00:14:39,940 --> 00:14:40,839
voice circuitry,

381
00:14:41,459 --> 00:14:44,199
four g is cutting over to Wi Fi.

382
00:14:44,419 --> 00:14:46,519
So But it didn't solve one problem, George.

383
00:14:47,139 --> 00:14:48,120
What was that?

384
00:14:48,740 --> 00:14:50,579
Being able to sell the new generation of

385
00:14:50,579 --> 00:14:52,820
phones after the four g phones were all

386
00:14:52,820 --> 00:14:53,320
sold.

387
00:14:56,235 --> 00:14:57,534
I've got a plan here.

388
00:14:57,914 --> 00:15:00,095
So we've done two g to three g,

389
00:15:00,154 --> 00:15:01,914
and we've just done three g to four

390
00:15:01,914 --> 00:15:05,355
g. Let's start the hype machine and make

391
00:15:05,355 --> 00:15:07,774
people be interested in five g

392
00:15:08,450 --> 00:15:10,529
While we're putting four g out, we'll tell

393
00:15:10,529 --> 00:15:13,330
them there's a bright new future. Over the

394
00:15:13,330 --> 00:15:15,009
hill, when you get to the crest of

395
00:15:15,009 --> 00:15:17,490
hill, you will catch glimpses of this new

396
00:15:17,490 --> 00:15:20,529
technology. We call it five g. So four

397
00:15:20,529 --> 00:15:23,475
g goes wide. Four g has full dual

398
00:15:23,475 --> 00:15:23,975
stack,

399
00:15:24,355 --> 00:15:26,674
good framing. We're not in slip anymore, guys.

400
00:15:26,674 --> 00:15:29,714
It's full PPP. It's got voice over IP,

401
00:15:29,714 --> 00:15:32,375
and it is crisp and clean and beautiful.

402
00:15:32,835 --> 00:15:35,500
It's got voice over Wi Fi, and people

403
00:15:35,500 --> 00:15:37,820
see it flicker on in the indicator on

404
00:15:37,820 --> 00:15:40,879
their phone. Nobody knows what it actually means.

405
00:15:41,500 --> 00:15:42,399
It says

406
00:15:42,700 --> 00:15:43,600
VO Wi

407
00:15:44,059 --> 00:15:44,720
Fi enabled.

408
00:15:45,579 --> 00:15:46,079
And

409
00:15:47,125 --> 00:15:48,105
there's this hint

410
00:15:48,485 --> 00:15:50,404
that you could buy a thing if the

411
00:15:50,404 --> 00:15:52,325
carrier wants to sell it to you called

412
00:15:52,325 --> 00:15:53,225
a femtocell,

413
00:15:54,165 --> 00:15:56,085
which you can stick in your own home,

414
00:15:56,085 --> 00:15:57,144
which in principle

415
00:15:57,925 --> 00:16:00,725
would allow you to have stronger signal because

416
00:16:00,725 --> 00:16:03,419
people are starting to recognize now that this

417
00:16:03,419 --> 00:16:06,080
four g, him no go very far.

418
00:16:06,539 --> 00:16:09,039
If you're not close to a tower,

419
00:16:09,500 --> 00:16:11,919
four g is kind of off the cliff.

420
00:16:12,220 --> 00:16:15,100
So suddenly, the telephone companies are what we

421
00:16:15,100 --> 00:16:17,360
would call hoist on their own petard.

422
00:16:17,834 --> 00:16:19,774
They've sold this amazing technology

423
00:16:20,154 --> 00:16:22,475
that for people living in rural and remote

424
00:16:22,475 --> 00:16:25,774
areas is actually not better than two g,

425
00:16:26,074 --> 00:16:28,875
let alone three g. So there's a few

426
00:16:28,875 --> 00:16:31,115
hints that this three g g to four

427
00:16:31,115 --> 00:16:32,014
g transition

428
00:16:32,759 --> 00:16:35,899
didn't work well for some communities of people.

429
00:16:35,960 --> 00:16:37,100
This power budget,

430
00:16:37,480 --> 00:16:39,179
who's paying for all this power?

431
00:16:39,720 --> 00:16:42,519
This tap cell tower problem. You mean that

432
00:16:42,519 --> 00:16:45,160
cell towers have, like, eminent domain and can

433
00:16:45,160 --> 00:16:47,800
just say we're sticking a tower on your

434
00:16:47,800 --> 00:16:48,300
territory

435
00:16:48,975 --> 00:16:50,335
and I have to put up with this

436
00:16:50,335 --> 00:16:52,414
thing? Oh, that's okay. We're gonna make them

437
00:16:52,414 --> 00:16:53,875
look like pine trees.

438
00:16:55,054 --> 00:16:57,375
But we don't have pine trees here. We

439
00:16:57,375 --> 00:16:59,695
have maple trees. I'm sorry. We don't have

440
00:16:59,695 --> 00:17:02,335
maple tree shaped poles. You get a pine

441
00:17:02,335 --> 00:17:03,714
tree if you wanna

442
00:17:04,039 --> 00:17:06,059
hide your four g mask.

443
00:17:06,599 --> 00:17:07,099
So

444
00:17:07,720 --> 00:17:08,539
five g,

445
00:17:09,159 --> 00:17:11,720
what's five g gonna deliver that four g

446
00:17:11,720 --> 00:17:12,539
didn't deliver?

447
00:17:13,000 --> 00:17:13,500
More,

448
00:17:14,919 --> 00:17:15,419
more.

449
00:17:16,285 --> 00:17:16,785
Yep.

450
00:17:17,404 --> 00:17:20,625
And it's gonna have beam forming. Yes. Beams.

451
00:17:20,765 --> 00:17:22,464
It's gonna have beam forming.

452
00:17:22,765 --> 00:17:25,585
Five g, the ants can steer these dishes.

453
00:17:26,125 --> 00:17:26,625
So

454
00:17:27,164 --> 00:17:28,799
five g comes out.

455
00:17:30,720 --> 00:17:33,619
I don't think I have ever cared

456
00:17:34,080 --> 00:17:36,400
that my hand said could or could not

457
00:17:36,400 --> 00:17:37,619
do five g.

458
00:17:38,320 --> 00:17:40,815
I cared about two g to three g,

459
00:17:40,894 --> 00:17:42,815
and I had to care about three g

460
00:17:42,815 --> 00:17:44,335
to four g because I had to go

461
00:17:44,335 --> 00:17:45,954
out and buy a new handset.

462
00:17:46,335 --> 00:17:47,394
But the thing is,

463
00:17:47,934 --> 00:17:49,954
every single g generation

464
00:17:50,575 --> 00:17:51,075
subsumes,

465
00:17:51,534 --> 00:17:52,274
it engulfs

466
00:17:52,575 --> 00:17:56,269
its predecessors. And so all four g phones

467
00:17:56,269 --> 00:17:59,089
would also talk three g and two g,

468
00:17:59,309 --> 00:18:02,210
and all five g phones will all automatically

469
00:18:02,669 --> 00:18:04,609
talk four g and three g.

470
00:18:05,069 --> 00:18:07,835
And so an awful lot of the world

471
00:18:07,835 --> 00:18:10,075
is living in three g and four g

472
00:18:10,075 --> 00:18:12,414
radio space with a five g capability

473
00:18:13,115 --> 00:18:15,515
that they are not aware they're using if

474
00:18:15,515 --> 00:18:16,494
they use it.

475
00:18:17,274 --> 00:18:19,835
And I do know people who are using

476
00:18:19,835 --> 00:18:21,769
five g who live

477
00:18:22,869 --> 00:18:23,529
in constrained

478
00:18:23,830 --> 00:18:24,330
environments,

479
00:18:24,950 --> 00:18:28,250
like cities with planning regulations that restrict

480
00:18:28,630 --> 00:18:30,869
the amount of cell towers that can be

481
00:18:30,869 --> 00:18:32,330
deployed because they're heritage.

482
00:18:32,710 --> 00:18:33,210
And

483
00:18:33,724 --> 00:18:34,545
the chatter

484
00:18:35,085 --> 00:18:36,865
amongst those people is,

485
00:18:37,164 --> 00:18:39,404
cell sucks now. I used to be able

486
00:18:39,404 --> 00:18:41,244
to pick up cell no matter where I

487
00:18:41,244 --> 00:18:44,144
was inside the store. Now it's five g.

488
00:18:44,445 --> 00:18:45,744
I can't get anything.

489
00:18:46,090 --> 00:18:47,950
I was reading a story the other day

490
00:18:48,250 --> 00:18:50,809
about the new Cross River Rail system in

491
00:18:50,809 --> 00:18:51,309
London

492
00:18:51,690 --> 00:18:53,850
and how they were determined to make all

493
00:18:53,850 --> 00:18:56,350
telephones work inside that environment.

494
00:18:57,049 --> 00:18:59,595
And what they did is they built a

495
00:18:59,595 --> 00:19:00,095
gigantic

496
00:19:00,474 --> 00:19:03,775
single common carrier in the underground of London.

497
00:19:04,795 --> 00:19:06,315
The way they do it is that they

498
00:19:06,315 --> 00:19:09,515
put this cable throughout the tube system where

499
00:19:09,515 --> 00:19:10,255
they deliberately

500
00:19:10,634 --> 00:19:12,335
break holes in the insulation,

501
00:19:12,634 --> 00:19:14,970
and the thing becomes what's called the leaky

502
00:19:14,970 --> 00:19:15,470
pipe.

503
00:19:15,769 --> 00:19:16,750
It's radiating

504
00:19:17,609 --> 00:19:19,150
miniature radio signal

505
00:19:19,529 --> 00:19:22,170
out of holes every 50 meters down this

506
00:19:22,170 --> 00:19:23,869
thing, and they function

507
00:19:24,250 --> 00:19:27,210
like tiny cells. So as you go past

508
00:19:27,210 --> 00:19:28,829
them in a train, you can go.

509
00:19:30,255 --> 00:19:32,575
You flip between the cells going down there,

510
00:19:32,575 --> 00:19:33,875
and people are going,

511
00:19:34,335 --> 00:19:34,835
seriously?

512
00:19:35,214 --> 00:19:35,955
They're saying,

513
00:19:36,335 --> 00:19:38,654
I get better five g reception in the

514
00:19:38,654 --> 00:19:39,154
underground

515
00:19:39,775 --> 00:19:41,955
than I do up on the street.

516
00:19:42,430 --> 00:19:44,990
So a network engineer would be going, well,

517
00:19:44,990 --> 00:19:47,950
yeah, because, a, it's a single carrier network,

518
00:19:47,950 --> 00:19:50,109
and we're doing, like, tenancies for all of

519
00:19:50,109 --> 00:19:53,070
the phones, and, b, there's an antenna every

520
00:19:53,070 --> 00:19:55,070
50 feet away from you. What did you

521
00:19:55,070 --> 00:19:57,890
expect? Mhmm. And this is starting to mean,

522
00:19:58,535 --> 00:20:00,375
so if we want five g to work

523
00:20:00,375 --> 00:20:01,835
in the general space,

524
00:20:02,134 --> 00:20:03,734
we have to agree to be able to

525
00:20:03,734 --> 00:20:04,234
see

526
00:20:04,694 --> 00:20:05,194
multiple

527
00:20:05,815 --> 00:20:09,095
endpoints, cellular radio points. Yep. And that's like,

528
00:20:09,095 --> 00:20:11,755
yeah. Those ads need a lot of places

529
00:20:11,815 --> 00:20:13,769
to direct the beams to because it's all

530
00:20:13,769 --> 00:20:14,670
about beamforming.

531
00:20:15,930 --> 00:20:18,170
Yeah. Well, it's a big boat, by the

532
00:20:18,170 --> 00:20:19,690
way. A big boat, a ship. What we

533
00:20:19,690 --> 00:20:20,990
would call a ship. Right?

534
00:20:21,289 --> 00:20:23,710
Cruise ships, you don't realize how many

535
00:20:24,170 --> 00:20:26,970
Wi Fi access points there are or little

536
00:20:26,970 --> 00:20:27,470
femtocells

537
00:20:27,849 --> 00:20:29,855
there are on a cruise

538
00:20:30,394 --> 00:20:32,894
ship. That those boats, they are stuffed

539
00:20:33,355 --> 00:20:34,734
with wireless gear

540
00:20:35,035 --> 00:20:37,595
because bulkheads by code have to be a

541
00:20:37,595 --> 00:20:39,994
certain depth, and you just can't get five

542
00:20:40,154 --> 00:20:42,634
You can't punch a signal through them. Yeah.

543
00:20:42,634 --> 00:20:44,659
So there so there aren't just like, there's

544
00:20:44,659 --> 00:20:47,059
not just, like, one Wi Fi or one

545
00:20:47,059 --> 00:20:47,559
femtocell

546
00:20:48,339 --> 00:20:50,839
per there's, like, one Wi Fi or femtocell

547
00:20:51,059 --> 00:20:51,880
per stateroom.

548
00:20:52,339 --> 00:20:54,419
Like, it's Oh, yeah. Crazy. It's how many

549
00:20:54,419 --> 00:20:56,519
of the rooms there are. So

550
00:20:57,704 --> 00:20:59,085
let's come back a bit.

551
00:20:59,704 --> 00:21:02,125
These things are now running at frequencies

552
00:21:03,144 --> 00:21:04,044
that are

553
00:21:04,664 --> 00:21:05,164
everywhere.

554
00:21:06,265 --> 00:21:08,585
There are frequency bands that are right up

555
00:21:08,585 --> 00:21:10,744
the top of the radio spectrum we're used

556
00:21:10,744 --> 00:21:11,404
to using.

557
00:21:12,345 --> 00:21:12,845
They're

558
00:21:13,180 --> 00:21:14,000
they're fringing

559
00:21:14,700 --> 00:21:17,500
into the region that we call terahertz. And

560
00:21:17,500 --> 00:21:19,660
there's a whole conversation to be had about

561
00:21:19,660 --> 00:21:22,720
what these labels mean, but they are astronomically

562
00:21:23,500 --> 00:21:24,240
high frequency

563
00:21:24,859 --> 00:21:25,680
data rates

564
00:21:26,244 --> 00:21:28,664
up at the top, but they also include

565
00:21:28,724 --> 00:21:31,284
bands that go all the way down into

566
00:21:31,284 --> 00:21:33,784
the spaces we used to use for VHF

567
00:21:34,005 --> 00:21:34,505
TV

568
00:21:35,284 --> 00:21:38,085
that we loved because they would punch I

569
00:21:38,085 --> 00:21:38,904
Love Lucy

570
00:21:39,309 --> 00:21:41,710
through bricks and mortar to get to your

571
00:21:41,710 --> 00:21:44,830
TV with rabbit antennas sitting in home. And

572
00:21:44,830 --> 00:21:45,490
we have

573
00:21:45,869 --> 00:21:46,369
stolen

574
00:21:46,830 --> 00:21:48,450
all of that radio spectrum

575
00:21:49,150 --> 00:21:52,130
to use it to provide this telephone service.

576
00:21:52,190 --> 00:21:53,250
Now the thing is,

577
00:21:53,549 --> 00:21:54,529
you can't actually

578
00:21:55,444 --> 00:21:58,644
steal radio because other people put their hand

579
00:21:58,644 --> 00:22:00,884
up to say they need it. The police,

580
00:22:00,884 --> 00:22:04,585
the military, the emergency response teams, local government,

581
00:22:04,724 --> 00:22:06,984
engineers who are out in the field, truckers,

582
00:22:07,045 --> 00:22:08,424
ten four, big buddy,

583
00:22:08,805 --> 00:22:10,025
microwave ovens.

584
00:22:10,890 --> 00:22:11,630
Wi Fi.

585
00:22:12,330 --> 00:22:14,910
Everybody wants a chunk of radio

586
00:22:15,210 --> 00:22:17,210
they can use to do some of the

587
00:22:17,210 --> 00:22:18,670
things they need to do.

588
00:22:19,210 --> 00:22:22,490
And it's interesting to say Wi Fi because

589
00:22:22,490 --> 00:22:24,670
an awful lot of these cell bands

590
00:22:25,125 --> 00:22:28,085
fringe into the area that is kind of

591
00:22:28,085 --> 00:22:30,565
the neighbor next door to the Wi Fi

592
00:22:30,565 --> 00:22:31,065
bands.

593
00:22:32,085 --> 00:22:32,585
Right?

594
00:22:32,884 --> 00:22:33,384
Now

595
00:22:34,005 --> 00:22:35,224
you're making chips

596
00:22:35,684 --> 00:22:38,585
that are able to consume radio in

597
00:22:38,884 --> 00:22:39,384
slots,

598
00:22:40,230 --> 00:22:42,549
and you have an upper limit and lower

599
00:22:42,549 --> 00:22:45,029
limit that is kind of defined that you

600
00:22:45,029 --> 00:22:45,529
can't

601
00:22:46,150 --> 00:22:47,289
break out of.

602
00:22:47,990 --> 00:22:50,309
You are constrained at the low and the

603
00:22:50,309 --> 00:22:52,490
high end. They are like fence lines

604
00:22:52,815 --> 00:22:55,214
you just cannot get out of. And that

605
00:22:55,214 --> 00:22:57,454
essentially is the problem that hit two g

606
00:22:57,454 --> 00:22:59,214
to three g and three g to four

607
00:22:59,214 --> 00:23:02,255
g. Four g had to have wider fence

608
00:23:02,255 --> 00:23:04,434
lines to get all these wonderful high frequencies,

609
00:23:04,974 --> 00:23:06,815
and they had to fringe down into the

610
00:23:06,815 --> 00:23:08,960
low ones to use them. The two g

611
00:23:08,960 --> 00:23:11,059
phones could only share bandwidth

612
00:23:11,599 --> 00:23:12,900
in that lower segment.

613
00:23:13,680 --> 00:23:16,019
So five g has got this

614
00:23:16,320 --> 00:23:18,019
enormous spread of footprint

615
00:23:18,400 --> 00:23:20,820
that fringes into the Wi Fi region,

616
00:23:21,335 --> 00:23:23,174
And people are starting to look at Wi

617
00:23:23,174 --> 00:23:23,914
Fi going,

618
00:23:25,494 --> 00:23:28,474
awfully nice bandwidth you've got there in the

619
00:23:28,535 --> 00:23:29,035
unregulated

620
00:23:29,654 --> 00:23:32,214
bandwidth space. It would be a shame if

621
00:23:32,214 --> 00:23:35,349
somebody fringed into that and started using it

622
00:23:35,349 --> 00:23:36,169
for commercial

623
00:23:36,549 --> 00:23:37,049
services,

624
00:23:37,509 --> 00:23:38,329
wouldn't it?

625
00:23:38,789 --> 00:23:40,730
So behind the scenes,

626
00:23:41,429 --> 00:23:43,910
in the three g p p, the forum

627
00:23:43,910 --> 00:23:46,309
where people do spec design on this work,

628
00:23:46,309 --> 00:23:47,369
and in the IEEE

629
00:23:48,065 --> 00:23:50,724
where people do the engineering work to constrain

630
00:23:50,785 --> 00:23:52,325
chips and make them behave.

631
00:23:52,704 --> 00:23:55,345
There is a lot of engineering horse trading

632
00:23:55,345 --> 00:23:57,045
going off where people are saying,

633
00:23:58,144 --> 00:24:00,304
same chip. I could I could sell you

634
00:24:00,304 --> 00:24:02,085
one chip that does this stuff.

635
00:24:02,400 --> 00:24:04,080
Yeah. I can I can lock out that

636
00:24:04,080 --> 00:24:06,480
region that's Wi Fi, and there's other people

637
00:24:06,480 --> 00:24:06,980
muttering,

638
00:24:07,359 --> 00:24:09,279
why don't we allow these phones to just

639
00:24:09,279 --> 00:24:11,539
use the Wi Fi and do stuff there?

640
00:24:11,920 --> 00:24:13,840
If we if we ever tried that, have

641
00:24:13,840 --> 00:24:16,559
we ever tried just sharing Wi Fi and

642
00:24:16,559 --> 00:24:19,625
using it? Why, yes. We have. There's a

643
00:24:19,625 --> 00:24:21,964
technology system called FON,

644
00:24:22,265 --> 00:24:23,945
f o n, that a friend of mine

645
00:24:23,945 --> 00:24:27,144
was working on engineering, where if you agreed

646
00:24:27,144 --> 00:24:29,305
to allow a percentage of your Wi Fi

647
00:24:29,305 --> 00:24:31,910
bandwidth to be used for general

648
00:24:32,369 --> 00:24:34,690
purpose, you got a rebate on your bill.

649
00:24:34,690 --> 00:24:37,350
And while I was walking down the street,

650
00:24:37,570 --> 00:24:40,390
I could fringe into different people's Wi Fi

651
00:24:40,450 --> 00:24:41,590
and send messages.

652
00:24:42,210 --> 00:24:42,710
And

653
00:24:43,170 --> 00:24:44,710
I'm looking at that going,

654
00:24:46,045 --> 00:24:48,765
okay. That was FON was kind of when

655
00:24:48,765 --> 00:24:51,164
home Wi Fi was operating at, like, 50

656
00:24:51,164 --> 00:24:51,664
megabits,

657
00:24:52,045 --> 00:24:54,845
and my home network now is maybe 300

658
00:24:54,845 --> 00:24:57,105
megabits, and my home fiber connection

659
00:24:57,565 --> 00:25:00,609
is 500 megabits in and maybe a 150

660
00:25:00,609 --> 00:25:01,910
to 200 out,

661
00:25:02,369 --> 00:25:03,829
and I'm using, like,

662
00:25:04,289 --> 00:25:05,670
2% of that,

663
00:25:06,369 --> 00:25:08,950
I I could be offering that radio bandwidth

664
00:25:09,089 --> 00:25:11,910
to hundreds of people and barely notice.

665
00:25:12,404 --> 00:25:14,585
If we organize this commons

666
00:25:15,205 --> 00:25:16,424
of radio frequency

667
00:25:16,725 --> 00:25:18,965
better, we could be doing this stuff. This

668
00:25:18,965 --> 00:25:21,225
stuff could be too cheap to meter.

669
00:25:22,244 --> 00:25:23,384
It's so ubiquitously

670
00:25:24,085 --> 00:25:24,585
fantastic,

671
00:25:25,159 --> 00:25:27,640
and it's pretty much the same chipsets at

672
00:25:27,640 --> 00:25:28,940
the radio layer

673
00:25:29,319 --> 00:25:30,380
that could do it.

674
00:25:31,159 --> 00:25:32,839
Could make it work. We could have an

675
00:25:32,839 --> 00:25:35,240
interesting world. The problem is it's kinda hard

676
00:25:35,240 --> 00:25:37,319
to charge for it, so it's probably not

677
00:25:37,319 --> 00:25:38,059
gonna happen.

678
00:25:38,865 --> 00:25:42,724
So what comes next after five g?

679
00:25:44,305 --> 00:25:47,184
Well, inductive logic tells us it's gonna be

680
00:25:47,184 --> 00:25:47,684
called

681
00:25:48,065 --> 00:25:48,964
six g.

682
00:25:49,265 --> 00:25:50,085
Right. And

683
00:25:50,799 --> 00:25:54,160
inductive reasoning tells us that somebody's gonna say

684
00:25:54,160 --> 00:25:56,980
it involves significant amounts of beamforming.

685
00:25:57,599 --> 00:25:59,039
I was gonna say, yeah. It's gonna be

686
00:25:59,039 --> 00:26:00,339
it's gonna be beamforming.

687
00:26:01,519 --> 00:26:02,259
So whatever.

688
00:26:03,154 --> 00:26:05,634
But there is something that has been going

689
00:26:05,634 --> 00:26:09,414
on in telephony lately that actually has potential

690
00:26:09,474 --> 00:26:11,714
to be very interesting, and that is the

691
00:26:11,714 --> 00:26:14,694
fact that although these incredibly high frequencies

692
00:26:16,100 --> 00:26:19,320
don't have good signal strength in some ways,

693
00:26:19,859 --> 00:26:23,400
it is actually possible to punch this stuff

694
00:26:23,619 --> 00:26:24,119
upwards

695
00:26:24,740 --> 00:26:26,039
in a fairly unconstrained

696
00:26:26,420 --> 00:26:29,539
manner. And so the big deal that's been

697
00:26:29,539 --> 00:26:30,359
going on

698
00:26:30,715 --> 00:26:33,035
in the four g, five g space that's

699
00:26:33,035 --> 00:26:35,434
kind of been bubbling under the counter and

700
00:26:35,434 --> 00:26:36,815
not really getting sold

701
00:26:37,195 --> 00:26:38,815
is cellular telephone

702
00:26:39,115 --> 00:26:40,975
to low earth orbit satellite.

703
00:26:41,674 --> 00:26:44,255
It's called d to c, direct to cell,

704
00:26:44,359 --> 00:26:46,920
or d to d, direct to device. Which

705
00:26:46,920 --> 00:26:50,200
we're seeing with, like, Starlink and Kupia and

706
00:26:50,200 --> 00:26:52,140
stuff like that. We're starting to see

707
00:26:52,599 --> 00:26:53,580
direct Yeah.

708
00:26:54,759 --> 00:26:55,259
And

709
00:26:56,039 --> 00:26:56,779
the world,

710
00:26:57,160 --> 00:26:59,259
as in the engineers who run

711
00:27:00,234 --> 00:27:02,015
cellular communication systems,

712
00:27:02,634 --> 00:27:05,115
are looking at that cost problem of power

713
00:27:05,115 --> 00:27:08,494
budget and available sites to put antenna

714
00:27:08,875 --> 00:27:11,515
and the need for more antenna, and they're

715
00:27:11,515 --> 00:27:13,054
looking at the availability

716
00:27:13,869 --> 00:27:16,529
of cellular space in a worldwide framework.

717
00:27:16,990 --> 00:27:20,029
And they're looking at things like that leaky

718
00:27:20,029 --> 00:27:22,829
pipe model where you construct a common carrier

719
00:27:22,829 --> 00:27:25,149
bus, and then people just rent bandwidth on

720
00:27:25,149 --> 00:27:26,909
top of it. And they're looking at the

721
00:27:26,909 --> 00:27:29,169
low Earth orbit satellites thinking,

722
00:27:29,815 --> 00:27:31,115
maybe this is the model

723
00:27:31,575 --> 00:27:33,994
we have to go to. So five g

724
00:27:34,295 --> 00:27:36,474
introduced this thing called NTN,

725
00:27:36,775 --> 00:27:38,075
the non terrestrial

726
00:27:38,615 --> 00:27:39,115
network,

727
00:27:39,815 --> 00:27:40,955
which is the beginnings

728
00:27:41,414 --> 00:27:43,974
of the ideas that were shoehorned into four

729
00:27:43,974 --> 00:27:47,480
g to do direct satellite. Four g's satellite

730
00:27:47,619 --> 00:27:50,180
was like low bandwidth. It was putting people

731
00:27:50,180 --> 00:27:51,320
back into SMS.

732
00:27:51,860 --> 00:27:54,180
And the cell line was, be out in

733
00:27:54,180 --> 00:27:55,860
the world, be in the hills. If you

734
00:27:55,860 --> 00:27:58,660
get lost, your satellite commons is gonna work

735
00:27:58,660 --> 00:28:00,514
its EPIRB, the emergency

736
00:28:00,815 --> 00:28:01,315
beacon

737
00:28:01,694 --> 00:28:04,034
of satellites. Right? Right. Yeah.

738
00:28:04,335 --> 00:28:06,014
You've got a Because it's low band it's

739
00:28:06,014 --> 00:28:08,014
low bandwidth. You're not sending a whole lot

740
00:28:08,014 --> 00:28:10,014
of data. You know? It's just it's just

741
00:28:10,014 --> 00:28:12,269
it's like your cell location or whatever.

742
00:28:12,670 --> 00:28:14,690
So five g, the NTN,

743
00:28:15,150 --> 00:28:16,990
people are saying, well, why do we have

744
00:28:16,990 --> 00:28:19,070
to say that it's not actually much data?

745
00:28:19,070 --> 00:28:22,269
I mean, these these dishies, these square antennas,

746
00:28:22,269 --> 00:28:24,509
they're getting, like, two, three hundred megabits out

747
00:28:24,509 --> 00:28:25,330
of this service.

748
00:28:26,054 --> 00:28:28,554
Inside the technology of one of these fractally

749
00:28:28,774 --> 00:28:31,335
designed antennas in a handset, we ought to

750
00:28:31,335 --> 00:28:33,514
be able to get five, ten,

751
00:28:34,134 --> 00:28:34,634
15.

752
00:28:35,174 --> 00:28:37,494
That's usable bandwidth, guys. I mean, there are

753
00:28:37,494 --> 00:28:40,054
places around the world where you don't get

754
00:28:40,054 --> 00:28:42,420
that on the cellular frequencies terrestrially.

755
00:28:42,960 --> 00:28:43,460
So

756
00:28:43,840 --> 00:28:46,080
how about we punch up? So I think

757
00:28:46,080 --> 00:28:48,559
it's reasonable to say that six g is

758
00:28:48,559 --> 00:28:51,440
gonna take the NTN model five g has

759
00:28:51,440 --> 00:28:52,259
as a concept

760
00:28:52,720 --> 00:28:54,900
and run with it a lot harder.

761
00:28:55,275 --> 00:28:57,674
We're not maybe gonna be seeing beams in

762
00:28:57,674 --> 00:29:00,095
the sense of steerable stuff on the ground,

763
00:29:00,234 --> 00:29:02,255
but we may well be seeing beaming

764
00:29:02,634 --> 00:29:05,375
in terms of steerable stuff going into space.

765
00:29:05,755 --> 00:29:09,059
But that also means a technology shift in

766
00:29:09,059 --> 00:29:12,099
terms of implementing this stuff because rolling out

767
00:29:12,099 --> 00:29:13,079
a new g

768
00:29:13,619 --> 00:29:15,640
is just about selling a handset.

769
00:29:16,019 --> 00:29:17,859
I mean, you sure want to sell that

770
00:29:17,859 --> 00:29:18,359
handset.

771
00:29:18,820 --> 00:29:20,660
That is a lot of income here is

772
00:29:20,660 --> 00:29:22,599
selling people upgradable handsets.

773
00:29:23,065 --> 00:29:25,384
But behind the scenes, you have to go

774
00:29:25,384 --> 00:29:26,445
out and reengineer

775
00:29:26,904 --> 00:29:30,605
every single cellular radio mast with new technology.

776
00:29:31,065 --> 00:29:33,005
We might have reached the point where

777
00:29:33,384 --> 00:29:36,265
the bandwidth constraints, the upper and lower gates

778
00:29:36,265 --> 00:29:39,210
on those chipsets doing the radio layer. Maybe

779
00:29:39,210 --> 00:29:41,130
that's got big enough that it's got the

780
00:29:41,130 --> 00:29:41,630
capability

781
00:29:41,930 --> 00:29:43,869
baked in to be reprogrammed

782
00:29:44,809 --> 00:29:47,309
so that you can do narrower signal processing

783
00:29:47,450 --> 00:29:49,769
tricks here. Maybe they're not gonna have to

784
00:29:49,769 --> 00:29:51,690
do a significant amount of work to go

785
00:29:51,690 --> 00:29:53,414
five g to six g, But I wouldn't

786
00:29:53,414 --> 00:29:55,735
hold my breath on it because I'm pretty

787
00:29:55,735 --> 00:29:58,315
damn sure they had to do major uplift

788
00:29:58,455 --> 00:30:00,455
to go three g to four g. And

789
00:30:00,455 --> 00:30:02,795
I've been getting a sense that the technology

790
00:30:02,934 --> 00:30:05,255
shift four g to five g has been

791
00:30:05,255 --> 00:30:06,715
shrinking the radio

792
00:30:07,250 --> 00:30:10,130
distance footprint of a cell tower and putting

793
00:30:10,130 --> 00:30:12,450
a lot more cell towers out there on

794
00:30:12,450 --> 00:30:13,110
the ground.

795
00:30:13,490 --> 00:30:14,390
We see

796
00:30:14,930 --> 00:30:17,809
billboards posted on telegraph poles here when a

797
00:30:17,809 --> 00:30:20,095
new cell is being deployed, And it's the,

798
00:30:20,095 --> 00:30:21,954
you know, you have the right to complain,

799
00:30:22,015 --> 00:30:24,015
you may write to, if you are concerned

800
00:30:24,015 --> 00:30:26,835
about type thing. And you'd see them

801
00:30:27,294 --> 00:30:29,694
you'd probably see three or four on a

802
00:30:29,694 --> 00:30:32,174
walk downtown, and it's like you see one

803
00:30:32,174 --> 00:30:35,259
every other telegraph pole now. Yeah. So those

804
00:30:35,259 --> 00:30:37,519
cell counts are rising massively.

805
00:30:37,980 --> 00:30:40,779
I think it's possible that they're gonna try

806
00:30:40,779 --> 00:30:43,660
and avoid having to do major reengineering at

807
00:30:43,660 --> 00:30:46,460
the physical layer and do software tricks to

808
00:30:46,460 --> 00:30:48,554
do the uplift, but there's a lot of

809
00:30:48,554 --> 00:30:50,554
back end technology that's gonna have to go

810
00:30:50,554 --> 00:30:52,714
on here. So what so what I I

811
00:30:52,714 --> 00:30:54,394
wanna go back to direct to device real

812
00:30:54,394 --> 00:30:54,894
quick.

813
00:30:55,515 --> 00:30:58,234
So do you see a future where that's

814
00:30:58,234 --> 00:30:59,994
a hybrid thing, sort of like Wi Fi

815
00:30:59,994 --> 00:31:01,755
was allowed to oh, yeah. We can use

816
00:31:01,755 --> 00:31:03,099
Wi Fi. We get this other thing. Do

817
00:31:03,099 --> 00:31:04,079
we see a a

818
00:31:04,700 --> 00:31:07,179
terrestrial low earth orbit hybrid type of arrangement

819
00:31:07,179 --> 00:31:08,960
with multiple beams at the same time?

820
00:31:09,419 --> 00:31:09,919
Well,

821
00:31:10,779 --> 00:31:13,099
if you ask on the technology space, I

822
00:31:13,099 --> 00:31:15,659
think it's possible the technology is gonna, in

823
00:31:15,659 --> 00:31:16,125
principle,

824
00:31:21,005 --> 00:31:23,605
pumping this stuff down low is more than

825
00:31:23,605 --> 00:31:25,184
making people go high.

826
00:31:25,565 --> 00:31:27,325
But the problem is I don't think the

827
00:31:27,325 --> 00:31:28,865
constraint here is technical.

828
00:31:29,220 --> 00:31:31,220
The constraint here is people who want to

829
00:31:31,220 --> 00:31:32,759
make that a money stream.

830
00:31:33,220 --> 00:31:33,720
So

831
00:31:34,019 --> 00:31:36,919
you said just earlier, Starlink and Kuiper,

832
00:31:37,299 --> 00:31:39,380
and we've got the European equivalent and the

833
00:31:39,380 --> 00:31:40,919
emerging Chinese equivalent.

834
00:31:41,394 --> 00:31:44,035
I don't understand the money behind these guys,

835
00:31:44,035 --> 00:31:46,535
but I don't believe the AT and Ts

836
00:31:46,914 --> 00:31:49,494
and the T Mo's and the SoftBank's

837
00:31:50,035 --> 00:31:51,795
have got a huge amount of sway in

838
00:31:51,795 --> 00:31:54,115
this. So if this stuff is gonna happen,

839
00:31:54,115 --> 00:31:54,615
Tom,

840
00:31:55,119 --> 00:31:57,380
there's gonna be some money conversations

841
00:31:57,759 --> 00:32:00,579
going on here about cross licensing and charging

842
00:32:00,640 --> 00:32:01,700
to make it happen.

843
00:32:02,240 --> 00:32:04,259
And that is gonna be fascinating.

844
00:32:04,799 --> 00:32:06,319
Yeah. That's gonna be the big thing is

845
00:32:06,319 --> 00:32:08,259
who makes money and how. So

846
00:32:08,825 --> 00:32:10,585
so you would say that six g is

847
00:32:10,585 --> 00:32:13,244
primarily the hype around six g is primarily

848
00:32:13,944 --> 00:32:14,605
the direct

849
00:32:15,384 --> 00:32:16,125
to satellite

850
00:32:16,744 --> 00:32:18,045
play that I haven't

851
00:32:18,424 --> 00:32:18,924
yet

852
00:32:19,384 --> 00:32:22,470
I haven't yet seen anyone describe a device

853
00:32:22,630 --> 00:32:25,429
because the horizon for six g deployment is

854
00:32:25,429 --> 00:32:26,329
twenty thirty.

855
00:32:26,950 --> 00:32:28,169
And in some senses,

856
00:32:28,630 --> 00:32:31,029
it's like they're sweating the asset a lot

857
00:32:31,029 --> 00:32:33,429
harder than they did. When four g came

858
00:32:33,429 --> 00:32:35,690
out, we kind of knew there was technology

859
00:32:35,829 --> 00:32:38,250
that hadn't made the cut. And I think

860
00:32:38,444 --> 00:32:40,284
if they'd been honest, they would have said

861
00:32:40,284 --> 00:32:41,884
it's actually 3.5

862
00:32:41,884 --> 00:32:43,904
g, but nobody liked the branding.

863
00:32:44,204 --> 00:32:46,684
And I think four to five g, they

864
00:32:46,684 --> 00:32:48,764
packed in some of the technology that didn't

865
00:32:48,764 --> 00:32:51,085
make the cut for four, and the stuff

866
00:32:51,085 --> 00:32:52,845
that got left off the edge is what's

867
00:32:52,845 --> 00:32:55,309
gonna be packaged into what's coming in six.

868
00:32:55,690 --> 00:32:58,970
Now there are various aspects to this that

869
00:32:58,970 --> 00:33:01,210
I think are a reflection of the shift

870
00:33:01,210 --> 00:33:03,049
in behavior. I mean, when it was three

871
00:33:03,049 --> 00:33:05,210
g to four g, it really was the

872
00:33:05,210 --> 00:33:07,710
beginnings of my phone is always on.

873
00:33:08,174 --> 00:33:10,174
Before then, people had viewed the phone as

874
00:33:10,174 --> 00:33:12,495
the constrained channel. I get a message. It

875
00:33:12,495 --> 00:33:15,215
stops. I'm billed per message. Four g is

876
00:33:15,215 --> 00:33:17,134
I'm paying for the service of data all

877
00:33:17,134 --> 00:33:17,795
the time.

878
00:33:18,495 --> 00:33:20,674
Five g is now saying,

879
00:33:21,079 --> 00:33:24,059
wow. We have the entire population of developed

880
00:33:24,119 --> 00:33:27,400
economies carrying two devices, an Ooyah Ring and

881
00:33:27,400 --> 00:33:29,660
a phone and an Apple Watch, three devices,

882
00:33:29,720 --> 00:33:32,039
and a pacemaker that can be remotely managed

883
00:33:32,039 --> 00:33:33,799
by the doctor, and their car has a

884
00:33:33,799 --> 00:33:35,019
chip, five devices

885
00:33:35,674 --> 00:33:38,095
permanently on per person walking around.

886
00:33:38,554 --> 00:33:40,234
We need to have a heck of a

887
00:33:40,234 --> 00:33:42,974
lot more device capacity in the network.

888
00:33:43,275 --> 00:33:45,674
And I think that's also partly why people

889
00:33:45,674 --> 00:33:48,279
are saying service no good here, Because they

890
00:33:48,279 --> 00:33:51,180
were used to being bleeding edge, new phone.

891
00:33:51,400 --> 00:33:54,380
I get 300 megabits of download. I'm watching

892
00:33:54,599 --> 00:33:56,859
the NFL finals on my screen.

893
00:33:57,559 --> 00:34:01,115
And then suddenly now there's, you know, 200

894
00:34:01,115 --> 00:34:03,994
more customers sharing the cell. You're not getting

895
00:34:03,994 --> 00:34:06,734
that experience anymore. You're getting a more constrained

896
00:34:06,875 --> 00:34:07,375
bandwidth.

897
00:34:07,835 --> 00:34:10,255
Five g is bringing a lot more bandwidth

898
00:34:10,315 --> 00:34:12,795
to the table because it fringes up into

899
00:34:12,795 --> 00:34:13,295
terahertz.

900
00:34:13,980 --> 00:34:16,219
Six g is gonna be living its best

901
00:34:16,219 --> 00:34:16,719
life

902
00:34:17,019 --> 00:34:17,519
solidly

903
00:34:18,139 --> 00:34:21,659
in terahertz frequency rates when it can. There

904
00:34:21,659 --> 00:34:24,380
is not gonna be ubiquity of bandwidth, but

905
00:34:24,380 --> 00:34:26,460
there's going to have to be a lot

906
00:34:26,460 --> 00:34:26,960
more.

907
00:34:27,454 --> 00:34:30,114
Okay. A lot. And I think that satellite

908
00:34:30,175 --> 00:34:32,894
uplift could well be we can all see

909
00:34:32,894 --> 00:34:34,974
the sky. If we can see the sky,

910
00:34:34,974 --> 00:34:36,594
we can see the sky cells.

911
00:34:37,054 --> 00:34:39,315
Yeah. Which is which is going to improve

912
00:34:39,534 --> 00:34:40,034
bandwidth

913
00:34:40,494 --> 00:34:40,994
dramatically.

914
00:34:42,449 --> 00:34:46,150
Now will it harm delay and jitter? Yes.

915
00:34:46,369 --> 00:34:48,469
And so I always find it funny when

916
00:34:48,610 --> 00:34:51,090
when we get into this thing of, well,

917
00:34:51,090 --> 00:34:53,030
six g is gonna be better for gaming.

918
00:34:53,409 --> 00:34:55,190
Well, it's gonna be better for gaming,

919
00:34:55,505 --> 00:34:57,425
or five g is better for gaming than

920
00:34:57,425 --> 00:35:00,065
four g. And you're like No. It's not.

921
00:35:00,224 --> 00:35:01,684
Not. It's not.

922
00:35:02,065 --> 00:35:04,545
I mean, higher bandwidth does give you lower

923
00:35:04,545 --> 00:35:05,045
delay

924
00:35:05,505 --> 00:35:07,905
and potentially lower jitter depending on how you

925
00:35:07,905 --> 00:35:08,644
play things.

926
00:35:09,050 --> 00:35:11,530
But that doesn't necessarily translate because you're using

927
00:35:11,530 --> 00:35:12,590
so much more bandwidth.

928
00:35:13,050 --> 00:35:16,269
That doesn't necessarily translate into anything useful.

929
00:35:16,889 --> 00:35:17,949
And I mean,

930
00:35:18,489 --> 00:35:19,309
this this

931
00:35:20,090 --> 00:35:20,590
compulsion

932
00:35:21,449 --> 00:35:22,109
to say

933
00:35:22,605 --> 00:35:25,804
because it's higher frequency and smarter encoding, it's

934
00:35:25,804 --> 00:35:26,304
better

935
00:35:26,605 --> 00:35:28,445
in those spaces. I look at it and

936
00:35:28,445 --> 00:35:31,664
think, and in aggregate, across all the customers

937
00:35:31,724 --> 00:35:32,224
available,

938
00:35:32,844 --> 00:35:34,900
there is less loss, there is less drop,

939
00:35:34,900 --> 00:35:37,039
there is less jitter and less delay.

940
00:35:37,340 --> 00:35:38,720
For a single individual,

941
00:35:39,500 --> 00:35:40,559
is that necessarily

942
00:35:40,860 --> 00:35:43,260
gonna be your Lyft experience if you choose

943
00:35:43,260 --> 00:35:44,079
to absolutely

944
00:35:44,539 --> 00:35:45,760
soak this machine

945
00:35:46,059 --> 00:35:47,820
with a stream of bits so that you

946
00:35:47,820 --> 00:35:50,014
can be what we call colloquially

947
00:35:50,315 --> 00:35:51,694
a loping bastard.

948
00:35:52,155 --> 00:35:52,655
Well,

949
00:35:53,114 --> 00:35:55,194
no. I don't think so. Yeah. Because I

950
00:35:55,194 --> 00:35:57,835
think the layering effect of taking care of

951
00:35:57,835 --> 00:36:00,234
all that forward error correcting down in the

952
00:36:00,234 --> 00:36:02,680
link layer and doing the cell to cell

953
00:36:02,680 --> 00:36:04,539
jump and the frequency hopping

954
00:36:04,840 --> 00:36:06,920
in broad brush sense, and then on top

955
00:36:06,920 --> 00:36:09,480
of that, intruding an IP packet, and then

956
00:36:09,480 --> 00:36:12,360
on top of that, intruding the question, where

957
00:36:12,360 --> 00:36:13,579
is the game server?

958
00:36:14,224 --> 00:36:16,784
I I don't think there's some Nirvana of

959
00:36:16,784 --> 00:36:19,585
fantastic gameplay out there. No. I don't I

960
00:36:19,585 --> 00:36:21,505
don't either. I think, you know and and

961
00:36:21,505 --> 00:36:23,344
even so the other thing you talked about

962
00:36:23,344 --> 00:36:24,085
was, like,

963
00:36:24,545 --> 00:36:25,045
cars

964
00:36:25,505 --> 00:36:27,925
in particular are becoming data hogs.

965
00:36:28,670 --> 00:36:31,489
They are becoming holy mackerel. How much data

966
00:36:31,789 --> 00:36:34,769
a car puts off or consumes is insane.

967
00:36:35,469 --> 00:36:37,950
And Yep. You know, so here you are

968
00:36:37,950 --> 00:36:40,930
on your gaming console in your house using

969
00:36:40,989 --> 00:36:42,824
five g or six g.

970
00:36:43,224 --> 00:36:45,704
Bada bing bada bing. And then an all

971
00:36:45,704 --> 00:36:46,204
electric,

972
00:36:46,744 --> 00:36:49,704
all autonomous car drives by and consumes a

973
00:36:49,704 --> 00:36:51,005
100% of the revenue.

974
00:36:53,864 --> 00:36:56,125
And you and your character dies.

975
00:36:59,670 --> 00:37:00,570
That's So

976
00:37:01,269 --> 00:37:02,010
you talked

977
00:37:02,389 --> 00:37:04,890
a little bit earlier when we started this

978
00:37:04,949 --> 00:37:05,449
rap

979
00:37:05,750 --> 00:37:06,230
about

980
00:37:07,909 --> 00:37:08,409
strength

981
00:37:09,829 --> 00:37:12,150
for the engineers was the amount of power

982
00:37:12,150 --> 00:37:14,364
you could physically carry around with you to

983
00:37:14,364 --> 00:37:15,184
drive these

984
00:37:15,484 --> 00:37:18,284
in an almost pre VLSI world. And we

985
00:37:18,284 --> 00:37:21,264
have somehow arrived at this rather odd situation.

986
00:37:21,644 --> 00:37:23,244
We want the phones to be as thin

987
00:37:23,244 --> 00:37:25,340
as a sheet of paper, and we want

988
00:37:25,340 --> 00:37:27,440
them to send the highest frequency

989
00:37:27,739 --> 00:37:28,239
signals

990
00:37:28,539 --> 00:37:29,760
we have ever done.

991
00:37:30,140 --> 00:37:32,160
And we need to have the cells

992
00:37:32,619 --> 00:37:33,119
significantly

993
00:37:33,660 --> 00:37:36,300
closer to me, the phone holder, to do

994
00:37:36,300 --> 00:37:37,519
the join over

995
00:37:37,905 --> 00:37:40,485
tiny battery, thin device, and high frequency.

996
00:37:41,585 --> 00:37:43,265
So all we've done is take the power

997
00:37:43,265 --> 00:37:45,184
budget that was in the handset with the

998
00:37:45,184 --> 00:37:47,505
giant battery, and we've moved it to the

999
00:37:47,505 --> 00:37:50,065
static site where the cell is. But that

1000
00:37:50,065 --> 00:37:50,565
cell

1001
00:37:50,864 --> 00:37:53,445
now needs a lot of volts and current,

1002
00:37:53,719 --> 00:37:56,219
and it's radiating a lot of heat budget.

1003
00:37:56,519 --> 00:37:58,599
And there is a back end burden of

1004
00:37:58,599 --> 00:38:01,480
fiber connecting these things together that has to

1005
00:38:01,480 --> 00:38:03,480
go to the moral equivalent of a main

1006
00:38:03,480 --> 00:38:05,339
data frame somewhere in the streetscape.

1007
00:38:05,719 --> 00:38:08,824
We're building out street furniture to provide this

1008
00:38:08,824 --> 00:38:11,784
communications hub. And, like, kitties sleep on top

1009
00:38:11,784 --> 00:38:14,505
of these things because they're radiating so much

1010
00:38:14,505 --> 00:38:15,005
heat.

1011
00:38:15,625 --> 00:38:17,864
Yeah. You know? Yeah. It's We we are

1012
00:38:17,864 --> 00:38:18,364
consuming

1013
00:38:19,065 --> 00:38:21,005
amazing amounts of power

1014
00:38:21,339 --> 00:38:23,980
to flood the world with radio waves. And

1015
00:38:23,980 --> 00:38:26,719
for what, I ask you, so that Audrey

1016
00:38:27,099 --> 00:38:30,000
can have an ongoing conversation with her fellow

1017
00:38:30,059 --> 00:38:33,280
plants. That's right. That's right. That's exactly right.

1018
00:38:33,585 --> 00:38:36,224
And and, you know, beyond all that, there's

1019
00:38:36,224 --> 00:38:38,804
still the centralizing effect of all of this.

1020
00:38:39,105 --> 00:38:41,525
You can't break into being a cell provider

1021
00:38:41,824 --> 00:38:44,385
any longer. Oh, no. No. That No. No.

1022
00:38:44,385 --> 00:38:45,284
No. This

1023
00:38:46,065 --> 00:38:48,644
this market is no longer about a competitive

1024
00:38:49,059 --> 00:38:51,960
hub of 15 or 20 local providers.

1025
00:38:52,500 --> 00:38:54,279
We are seeing the reconcentration

1026
00:38:54,819 --> 00:38:56,599
down to a much more

1027
00:38:57,460 --> 00:38:59,319
rational number of providers.

1028
00:38:59,699 --> 00:39:02,605
Which, comrade, begs the question, why do we

1029
00:39:02,605 --> 00:39:04,144
even have competition?

1030
00:39:04,684 --> 00:39:07,324
Yeah. Exactly. Do we? I mean, if this

1031
00:39:07,324 --> 00:39:10,125
is a utility function, don't we just want

1032
00:39:10,125 --> 00:39:13,239
one substrate that is available to everyone? Yeah.

1033
00:39:13,239 --> 00:39:15,320
It does Is that where we're going? Is

1034
00:39:15,320 --> 00:39:16,300
physical connectivity

1035
00:39:17,880 --> 00:39:18,380
such

1036
00:39:18,680 --> 00:39:20,460
a such a thing as a commodity

1037
00:39:21,400 --> 00:39:24,280
that we shouldn't even bother? Like, there should

1038
00:39:24,280 --> 00:39:27,420
just be physical connectivity providers in each geographic

1039
00:39:27,480 --> 00:39:27,980
location.

1040
00:39:28,635 --> 00:39:30,815
Now, of course, we all know

1041
00:39:31,195 --> 00:39:31,695
that

1042
00:39:32,155 --> 00:39:33,135
that doesn't necessarily

1043
00:39:33,994 --> 00:39:35,934
work out well for the consumer.

1044
00:39:38,155 --> 00:39:40,335
I'm living in an economy with a national

1045
00:39:40,394 --> 00:39:41,454
broadband network,

1046
00:39:42,489 --> 00:39:45,289
and we're not delighted, but we're not so

1047
00:39:45,289 --> 00:39:48,010
angry we're rioting on the streets. We have

1048
00:39:48,010 --> 00:39:48,510
vibrant

1049
00:39:48,809 --> 00:39:49,309
competition

1050
00:39:50,170 --> 00:39:52,750
above that in the provision of Internet services

1051
00:39:52,890 --> 00:39:54,190
over the common substrate.

1052
00:39:55,125 --> 00:39:57,364
And I would be perfectly happy if the

1053
00:39:57,364 --> 00:39:59,545
radio masts were all owned by

1054
00:40:00,005 --> 00:40:02,724
a single provider and were implemented as a

1055
00:40:02,724 --> 00:40:04,025
common data fabric.

1056
00:40:04,324 --> 00:40:07,125
And my ability to move between masts was

1057
00:40:07,125 --> 00:40:10,199
about their engineering of cell handoff, and the

1058
00:40:10,199 --> 00:40:12,760
build cycle is just like a thin veneer

1059
00:40:12,760 --> 00:40:14,840
on top. I'd be fine in that world,

1060
00:40:14,840 --> 00:40:17,239
Russ. I really don't have a problem. If

1061
00:40:17,239 --> 00:40:19,420
we sweat the acid a bit more rationally

1062
00:40:19,880 --> 00:40:20,699
and compete

1063
00:40:21,000 --> 00:40:23,239
in a more abstract layer above it, I

1064
00:40:23,239 --> 00:40:24,139
could be okay.

1065
00:40:25,074 --> 00:40:27,555
Interesting. Not not that it's an equivalent, but

1066
00:40:27,555 --> 00:40:29,715
in a previous life, I had some exposure

1067
00:40:29,715 --> 00:40:30,614
to metropolitan,

1068
00:40:31,715 --> 00:40:33,574
you know, city run networks.

1069
00:40:34,114 --> 00:40:37,015
And Oh, yeah. It was it was interesting

1070
00:40:37,074 --> 00:40:39,394
the things that were there because it's not

1071
00:40:39,394 --> 00:40:41,150
like they could hire lots of engineers.

1072
00:40:41,449 --> 00:40:43,609
Right? It's not like these and and I'm

1073
00:40:43,609 --> 00:40:45,849
talking city government. I'm not talking national government.

1074
00:40:45,849 --> 00:40:46,730
But but,

1075
00:40:47,449 --> 00:40:49,690
you know, I it it does make me

1076
00:40:49,690 --> 00:40:50,889
wonder, you know, in in the in The

1077
00:40:50,889 --> 00:40:53,130
United States at least, what the the federal

1078
00:40:53,130 --> 00:40:55,049
government is seeing it in exodus of talent

1079
00:40:55,049 --> 00:40:57,364
for various reasons. But, like, if you can't

1080
00:40:57,525 --> 00:40:59,764
like, I I the federal government in The

1081
00:40:59,764 --> 00:41:02,105
US has always had a problem having competitive

1082
00:41:02,244 --> 00:41:02,744
salary.

1083
00:41:03,045 --> 00:41:05,204
Like, if there's an and and, obviously, this

1084
00:41:05,204 --> 00:41:07,684
becomes regional very quickly. But I I don't

1085
00:41:07,684 --> 00:41:08,344
know. Like,

1086
00:41:08,644 --> 00:41:10,644
what happens when you want if you want

1087
00:41:10,644 --> 00:41:13,340
this national substrate, but you can't pay anybody

1088
00:41:13,340 --> 00:41:14,400
to build it for you.

1089
00:41:14,940 --> 00:41:16,619
You know, and and in the meantime, we

1090
00:41:16,619 --> 00:41:18,699
poured all this tax money into building this

1091
00:41:18,699 --> 00:41:20,559
thing and then and we can't pay engineers

1092
00:41:20,619 --> 00:41:21,840
to make it actually work.

1093
00:41:22,539 --> 00:41:23,840
I don't know. I'm I'm skeptical.

1094
00:41:24,140 --> 00:41:26,300
Well, so so so in The US, of

1095
00:41:26,300 --> 00:41:28,925
course, it would never be federal government.

1096
00:41:29,545 --> 00:41:30,765
It would be regional

1097
00:41:31,385 --> 00:41:32,365
like TVA,

1098
00:41:33,224 --> 00:41:34,445
the Texas grid,

1099
00:41:34,825 --> 00:41:36,445
the Florida grid, whatever.

1100
00:41:36,905 --> 00:41:39,005
All the regions have their grids.

1101
00:41:39,730 --> 00:41:41,730
Power grids, it'd be the same thing. It

1102
00:41:41,730 --> 00:41:42,630
would be regional.

1103
00:41:43,010 --> 00:41:44,930
It would not be the federal government doing

1104
00:41:44,930 --> 00:41:46,550
it. That's that's my impression.

1105
00:41:47,489 --> 00:41:50,230
So the funny thing is that before cellular

1106
00:41:51,010 --> 00:41:53,430
phones really took off as a national

1107
00:41:54,125 --> 00:41:57,025
activity with this competitive tension between the providers.

1108
00:41:57,085 --> 00:41:59,264
There were some early attempts to do this

1109
00:41:59,644 --> 00:42:01,885
in the markets that people thought would work,

1110
00:42:01,885 --> 00:42:04,125
and one of them was London. There was

1111
00:42:04,125 --> 00:42:05,105
a pre cellular

1112
00:42:06,219 --> 00:42:09,760
distributed network put up around the Finance District

1113
00:42:10,140 --> 00:42:12,460
to be the way that day traders and

1114
00:42:12,460 --> 00:42:14,539
people like that could be online all the

1115
00:42:14,539 --> 00:42:17,260
time. And it involved the equivalent of what,

1116
00:42:17,260 --> 00:42:18,640
I guess, we call femtocells,

1117
00:42:18,940 --> 00:42:21,744
because the street furniture in London is just

1118
00:42:21,744 --> 00:42:22,244
everywhere.

1119
00:42:22,545 --> 00:42:25,105
Sticking up another white pole with a antenna

1120
00:42:25,105 --> 00:42:26,784
on the top of it was like, well,

1121
00:42:26,784 --> 00:42:29,264
who's gonna notice against the background of all

1122
00:42:29,264 --> 00:42:31,284
the other things we have in the streetscape?

1123
00:42:32,465 --> 00:42:33,284
It tanked.

1124
00:42:33,789 --> 00:42:35,789
I don't think they lasted two years. It

1125
00:42:35,789 --> 00:42:37,569
was just financially unviable.

1126
00:42:38,190 --> 00:42:39,170
But on a technology

1127
00:42:39,469 --> 00:42:39,969
basis,

1128
00:42:40,510 --> 00:42:43,630
everything they were doing then to radiate the

1129
00:42:43,630 --> 00:42:44,609
whole of the cityscape

1130
00:42:45,069 --> 00:42:47,230
with a fabric that people can use to

1131
00:42:47,230 --> 00:42:49,425
do what they're going to do, It's pretty

1132
00:42:49,425 --> 00:42:51,664
much what we are now living in right

1133
00:42:51,664 --> 00:42:54,305
now, but with a different market model, with

1134
00:42:54,305 --> 00:42:56,724
a different charging model. And so

1135
00:42:57,105 --> 00:42:59,344
five g to six g, we're all going

1136
00:42:59,344 --> 00:43:01,505
up into the sky. How many carriers are

1137
00:43:01,505 --> 00:43:03,425
there up in the sky? Well, right now,

1138
00:43:03,425 --> 00:43:04,325
there's one.

1139
00:43:04,929 --> 00:43:07,730
And yet the cellular provider model is that

1140
00:43:07,730 --> 00:43:10,630
there's probably four viable three or four viable

1141
00:43:10,690 --> 00:43:13,190
cellular carriers per nation state.

1142
00:43:13,809 --> 00:43:16,369
All of those guys competing for airtime to

1143
00:43:16,369 --> 00:43:19,110
speak with one company floating up in Liye,

1144
00:43:19,815 --> 00:43:21,755
that is a political nightmare.

1145
00:43:22,295 --> 00:43:24,855
Something's gotta give here. And I think your

1146
00:43:24,855 --> 00:43:27,734
skepticism, Tom, is well grounded. I can dream

1147
00:43:27,734 --> 00:43:29,655
of a nirvana where the city is run

1148
00:43:29,655 --> 00:43:31,734
by people like you, and I would, of

1149
00:43:31,734 --> 00:43:34,474
course, as a consumer, be absolutely delighted

1150
00:43:34,789 --> 00:43:36,170
that you're paid beans.

1151
00:43:36,550 --> 00:43:38,089
But as a comms engineer,

1152
00:43:38,550 --> 00:43:40,710
I can't say the prospect fills me with

1153
00:43:40,710 --> 00:43:41,849
glee because

1154
00:43:42,150 --> 00:43:44,150
it's not a viable living if you can't

1155
00:43:44,150 --> 00:43:45,829
afford to pay them. And if you don't

1156
00:43:45,829 --> 00:43:47,989
have people coming into the sector, you don't

1157
00:43:47,989 --> 00:43:50,684
have the engineers, and the maintenance goes down,

1158
00:43:50,844 --> 00:43:52,864
and the whole thing's looking pretty shaky.

1159
00:43:53,164 --> 00:43:54,844
So there kind of has to be an

1160
00:43:54,844 --> 00:43:57,565
equipoise. Right? We have to have enough people

1161
00:43:57,565 --> 00:44:00,285
behind the scenes running this stuff, pulling good

1162
00:44:00,285 --> 00:44:02,224
wage to be at the level of competency

1163
00:44:02,525 --> 00:44:03,184
and capacity

1164
00:44:03,809 --> 00:44:06,469
to provide what we want. And we demand

1165
00:44:07,090 --> 00:44:09,489
five nines out of this baby. Right? I

1166
00:44:09,489 --> 00:44:10,309
mean, individually,

1167
00:44:10,610 --> 00:44:12,690
all of us experience a lot less than

1168
00:44:12,690 --> 00:44:14,070
that, but this technology

1169
00:44:14,369 --> 00:44:15,110
is now

1170
00:44:15,410 --> 00:44:16,630
lifesaver technology.

1171
00:44:17,335 --> 00:44:19,655
You're gonna die if your cell phone goes

1172
00:44:19,655 --> 00:44:22,954
off because you can't call emergency services.

1173
00:44:23,255 --> 00:44:25,735
You can't call the police. Your mom can't

1174
00:44:25,735 --> 00:44:27,355
call to say I'm having anaphylactic

1175
00:44:27,735 --> 00:44:30,315
shock. This fabric has become

1176
00:44:30,855 --> 00:44:31,675
must work.

1177
00:44:32,949 --> 00:44:35,829
Yeah. Yeah. That's true. And the other thing

1178
00:44:35,829 --> 00:44:37,130
is without the competition,

1179
00:44:37,750 --> 00:44:39,929
it's not only just hard to justify

1180
00:44:41,190 --> 00:44:42,010
the pay,

1181
00:44:42,390 --> 00:44:44,170
it's hard to build a career

1182
00:44:45,074 --> 00:44:45,574
that

1183
00:44:46,034 --> 00:44:48,534
is anything other than stability focused.

1184
00:44:49,554 --> 00:44:51,815
Put it that way. Right? Like,

1185
00:44:52,755 --> 00:44:54,614
road engineers don't become multimillionaires

1186
00:44:55,155 --> 00:44:57,714
most of the time because they work for

1187
00:44:57,714 --> 00:45:00,355
state governments. They're guaranteed never to lose their

1188
00:45:00,355 --> 00:45:02,590
job for the most part, that is true,

1189
00:45:03,210 --> 00:45:04,670
but they're never,

1190
00:45:05,130 --> 00:45:07,550
like, you know, you don't you don't see

1191
00:45:07,930 --> 00:45:09,869
a road engineer on YouTube

1192
00:45:10,730 --> 00:45:12,510
with a following of 10,000,000,

1193
00:45:13,385 --> 00:45:14,984
you know, or what? No. It's more a

1194
00:45:14,984 --> 00:45:16,445
following of 5,000.

1195
00:45:16,664 --> 00:45:19,644
But those 5,000 kids with their Tonka trucks,

1196
00:45:19,945 --> 00:45:21,704
at least a thousand of them are gonna

1197
00:45:21,704 --> 00:45:24,105
grow up to be road engineers. They're feeding

1198
00:45:24,105 --> 00:45:26,789
the beast. But those guys also have a

1199
00:45:26,789 --> 00:45:29,829
significant component of pay that is effectively danger

1200
00:45:29,829 --> 00:45:32,309
money Yeah. Because they signed on the dotted

1201
00:45:32,309 --> 00:45:34,630
line. They will be out at 2AM

1202
00:45:34,630 --> 00:45:37,684
in the snow hooking the wires up. Yeah.

1203
00:45:37,684 --> 00:45:40,565
And this is serious stuff. Yeah. That's true.

1204
00:45:40,565 --> 00:45:41,925
So I don't know. It's it's a different

1205
00:45:41,925 --> 00:45:43,925
kind of career field. And so it's it

1206
00:45:43,925 --> 00:45:45,945
is hard sometimes to figure out

1207
00:45:46,244 --> 00:45:48,025
economically how to make that

1208
00:45:48,565 --> 00:45:49,065
viable,

1209
00:45:50,085 --> 00:45:52,210
in some way. This is part of what

1210
00:45:52,210 --> 00:45:53,890
you get with the electrical grid in The

1211
00:45:53,890 --> 00:45:55,670
United States, is that

1212
00:45:56,210 --> 00:45:57,190
while consumers

1213
00:45:58,289 --> 00:46:01,489
don't move someplace because of consistency of electrical

1214
00:46:01,489 --> 00:46:03,349
grid, it's not something we think about,

1215
00:46:03,809 --> 00:46:04,309
businesses

1216
00:46:04,690 --> 00:46:05,190
do.

1217
00:46:05,684 --> 00:46:06,425
And therefore,

1218
00:46:07,045 --> 00:46:08,105
electrical grids

1219
00:46:08,485 --> 00:46:10,105
do compete with one another

1220
00:46:10,485 --> 00:46:11,704
in the sense of,

1221
00:46:12,085 --> 00:46:14,724
can we draw that new factory? Can we

1222
00:46:14,724 --> 00:46:16,505
draw that new office building

1223
00:46:16,965 --> 00:46:19,949
into our city? And if we don't have

1224
00:46:19,949 --> 00:46:21,409
good electrical grid,

1225
00:46:22,349 --> 00:46:24,130
if we don't have the reputation

1226
00:46:24,510 --> 00:46:25,170
of supporting

1227
00:46:25,949 --> 00:46:26,849
always on,

1228
00:46:27,230 --> 00:46:29,070
then that company is gonna look at us

1229
00:46:29,070 --> 00:46:32,369
less seriously for for putting that new building

1230
00:46:33,054 --> 00:46:35,934
in our in our environs, which is gonna

1231
00:46:35,934 --> 00:46:36,594
hurt us

1232
00:46:36,974 --> 00:46:38,974
from a tax base and other things like

1233
00:46:38,974 --> 00:46:40,994
this. So there's still some

1234
00:46:41,454 --> 00:46:43,875
some kind of competition there, I would think.

1235
00:46:45,030 --> 00:46:48,250
So there is a sell in six g

1236
00:46:49,110 --> 00:46:50,809
that we haven't talked about.

1237
00:46:52,070 --> 00:46:54,869
And I have a feeling it probably is

1238
00:46:54,869 --> 00:46:57,750
going to figure in the marketing materials when

1239
00:46:57,750 --> 00:47:00,864
eventually people decide it's coming around the corner

1240
00:47:00,864 --> 00:47:03,184
and they need to start selling it. And

1241
00:47:03,184 --> 00:47:04,884
it's a two letter acronym,

1242
00:47:06,304 --> 00:47:08,224
and it starts at the beginning of the

1243
00:47:08,224 --> 00:47:08,724
alphabet.

1244
00:47:09,744 --> 00:47:11,844
Six g is going to be

1245
00:47:13,969 --> 00:47:14,469
AI.

1246
00:47:14,849 --> 00:47:17,010
Oh, oh, yes. Yeah. We know that's gonna

1247
00:47:17,010 --> 00:47:17,989
be the big thing.

1248
00:47:18,929 --> 00:47:19,429
So

1249
00:47:19,809 --> 00:47:21,670
let's riddle me this.

1250
00:47:22,289 --> 00:47:23,429
What is the component

1251
00:47:23,730 --> 00:47:26,150
of AI in six g?

1252
00:47:27,035 --> 00:47:29,454
And I have a couple of ideas

1253
00:47:29,755 --> 00:47:32,894
about how this might work. Bandwidth utilization and

1254
00:47:33,835 --> 00:47:34,335
beamforming.

1255
00:47:37,355 --> 00:47:38,335
You're too cynical.

1256
00:47:41,355 --> 00:47:43,950
Yes. But also no, because

1257
00:47:44,650 --> 00:47:47,789
I think there's a thing emerging here

1258
00:47:48,250 --> 00:47:49,230
that is probably,

1259
00:47:50,010 --> 00:47:50,829
in part,

1260
00:47:51,289 --> 00:47:53,470
hopefully, gonna improve the horizon.

1261
00:47:54,090 --> 00:47:56,030
It's improve the local experience.

1262
00:47:56,894 --> 00:47:59,635
So the handset is gonna have NPUs,

1263
00:47:59,934 --> 00:48:02,275
GPUs. The handset's gonna be operating

1264
00:48:02,655 --> 00:48:04,914
a complex local agent model.

1265
00:48:05,295 --> 00:48:07,054
There's going to be a lot more memory

1266
00:48:07,054 --> 00:48:09,474
in the handset once we get over this

1267
00:48:09,840 --> 00:48:12,320
bump in the road of availability of RAM

1268
00:48:12,320 --> 00:48:13,780
chips. These devices

1269
00:48:14,160 --> 00:48:16,820
are gonna be operating some kind of constrained

1270
00:48:17,119 --> 00:48:18,019
machine learning,

1271
00:48:18,960 --> 00:48:19,460
LLM

1272
00:48:20,000 --> 00:48:20,500
environment.

1273
00:48:21,280 --> 00:48:23,780
And so I think the hope is

1274
00:48:24,275 --> 00:48:26,855
you or I will be walking along going,

1275
00:48:27,075 --> 00:48:27,894
hey, Siri.

1276
00:48:28,195 --> 00:48:30,515
Could you construct the purchasing list for the

1277
00:48:30,515 --> 00:48:32,515
food I need by asking the fridge what

1278
00:48:32,515 --> 00:48:34,434
we're gonna have for dinner tomorrow? And then

1279
00:48:34,434 --> 00:48:36,275
when I get to the shop, could you

1280
00:48:36,275 --> 00:48:38,035
guide me round the shop so that I

1281
00:48:38,035 --> 00:48:39,175
can pick the cans

1282
00:48:39,489 --> 00:48:42,610
of premade ultra processed fat to put in

1283
00:48:42,610 --> 00:48:45,110
my shopping pod that is autonomously

1284
00:48:45,570 --> 00:48:48,210
following me along the road? And Siri's gonna

1285
00:48:48,210 --> 00:48:49,989
say, I can do that.

1286
00:48:50,289 --> 00:48:52,950
And what Siri is not gonna say is,

1287
00:48:53,574 --> 00:48:56,695
I'm sorry, George. I can't do that because

1288
00:48:56,695 --> 00:48:59,735
the cell tower doesn't have the bandwidth for

1289
00:48:59,735 --> 00:49:02,135
me to speak to the AI remotely in

1290
00:49:02,135 --> 00:49:02,795
the DC

1291
00:49:03,175 --> 00:49:04,695
for it to do the job and tell

1292
00:49:04,695 --> 00:49:07,574
me what to do. Siri's gonna say, I'm

1293
00:49:07,574 --> 00:49:08,795
running in your phone.

1294
00:49:09,469 --> 00:49:11,469
I can do that. I can handle that

1295
00:49:11,469 --> 00:49:14,349
right now. I'll do low data rate behind

1296
00:49:14,349 --> 00:49:17,010
the scenes. I can handle delay and jitter

1297
00:49:17,309 --> 00:49:20,190
because it's not you experiencing it. I'm gonna

1298
00:49:20,190 --> 00:49:23,054
handle the packet loss and delay. I'll construct

1299
00:49:23,114 --> 00:49:26,235
the world model of George's ideal meal of

1300
00:49:26,235 --> 00:49:26,735
pizza,

1301
00:49:27,114 --> 00:49:29,355
and I will help you buy that pizza

1302
00:49:29,355 --> 00:49:30,894
for one low, low price

1303
00:49:31,195 --> 00:49:34,094
and take it home. I think the belief

1304
00:49:34,155 --> 00:49:35,295
is the AI

1305
00:49:35,889 --> 00:49:38,469
fronted services running in your device

1306
00:49:38,849 --> 00:49:39,829
are gonna homogenize

1307
00:49:40,210 --> 00:49:41,989
the data plane out the back,

1308
00:49:42,289 --> 00:49:44,710
and the phone engineers are all gonna go,

1309
00:49:45,969 --> 00:49:47,109
we dodged a bullet

1310
00:49:47,409 --> 00:49:51,065
because now it's constant plane. George is consuming

1311
00:49:51,284 --> 00:49:53,385
50 megabits all the time,

1312
00:49:53,684 --> 00:49:55,224
but he doesn't peak

1313
00:49:55,605 --> 00:49:57,385
up into the hundreds of megabits.

1314
00:49:57,765 --> 00:49:59,925
He's just on a layer, and I can

1315
00:49:59,925 --> 00:50:03,349
engineer a layer cake of bandwidth behavior to

1316
00:50:03,349 --> 00:50:06,090
handle George's AI constrained agent

1317
00:50:06,469 --> 00:50:07,449
in his handset,

1318
00:50:07,989 --> 00:50:09,849
I think that's gonna be the sell.

1319
00:50:10,949 --> 00:50:13,449
That's interesting. That's a very interesting concept.

1320
00:50:15,284 --> 00:50:17,704
You have been speaking to a simulacrum

1321
00:50:18,244 --> 00:50:19,625
of George Michaelson.

1322
00:50:20,885 --> 00:50:21,444
Yeah. That's

1323
00:50:22,164 --> 00:50:24,244
yeah. That's an interesting that's an interesting sell

1324
00:50:24,244 --> 00:50:24,744
point.

1325
00:50:25,924 --> 00:50:27,784
Now will people want that? Yes.

1326
00:50:28,085 --> 00:50:29,605
Will it be good for them? Well, that's

1327
00:50:29,605 --> 00:50:30,824
another entire show.

1328
00:50:31,569 --> 00:50:34,369
Yeah. Yeah. That's that's another entire set of

1329
00:50:34,369 --> 00:50:34,869
problems.

1330
00:50:35,329 --> 00:50:37,409
I ain't doing that. No. I'm not doing

1331
00:50:37,409 --> 00:50:39,589
it either, but, you know, that's me.

1332
00:50:41,250 --> 00:50:42,849
And I know a lot of people, they

1333
00:50:42,849 --> 00:50:43,909
choose convenience

1334
00:50:44,255 --> 00:50:47,775
over everything, and I'd I worry anyway, that's

1335
00:50:47,775 --> 00:50:48,594
another entire

1336
00:50:49,454 --> 00:50:51,635
that's another, like I said, another entire show.

1337
00:50:51,695 --> 00:50:52,835
You young whippersnapper.

1338
00:50:53,295 --> 00:50:53,795
Yeah.

1339
00:50:54,335 --> 00:50:55,775
I don't know. You know, I go back

1340
00:50:55,775 --> 00:50:57,535
to music, and I think everybody on the

1341
00:50:57,535 --> 00:50:59,519
last podcast recording when I was talking to

1342
00:50:59,519 --> 00:51:01,760
Mike about this, about when we were talking

1343
00:51:01,760 --> 00:51:03,519
with Mike about this, and I was saying

1344
00:51:03,519 --> 00:51:05,440
how much better the music was back in

1345
00:51:05,440 --> 00:51:06,099
the nineties

1346
00:51:06,400 --> 00:51:07,300
and the eighties.

1347
00:51:07,680 --> 00:51:09,760
And even going back a bit into the

1348
00:51:09,760 --> 00:51:10,260
seventies

1349
00:51:10,559 --> 00:51:12,340
when there was that bend over

1350
00:51:12,684 --> 00:51:14,204
where the people who knew how to play

1351
00:51:14,204 --> 00:51:16,864
musical instruments were giving act given access

1352
00:51:17,244 --> 00:51:18,065
to digitalized

1353
00:51:18,764 --> 00:51:20,224
versions of those instruments.

1354
00:51:20,525 --> 00:51:22,204
And they knew what they were trying to

1355
00:51:22,204 --> 00:51:24,784
do, but they could never express it before.

1356
00:51:25,085 --> 00:51:27,069
And all of a sudden, you have people

1357
00:51:27,069 --> 00:51:28,829
like the guy who did Boston, which was,

1358
00:51:28,829 --> 00:51:30,369
by the way, one guy

1359
00:51:30,670 --> 00:51:31,409
sitting in

1360
00:51:31,789 --> 00:51:32,690
in his garage,

1361
00:51:33,230 --> 00:51:35,969
making up things that you attach to electrical

1362
00:51:36,030 --> 00:51:38,130
guitars to make them sound different.

1363
00:51:38,429 --> 00:51:39,949
A lot of people don't realize that's how

1364
00:51:39,949 --> 00:51:42,605
Boston ever really was. And, you know, Kansas

1365
00:51:42,605 --> 00:51:43,984
and all these other big groups.

1366
00:51:44,364 --> 00:51:46,605
But anyway, I was also noticing that I

1367
00:51:46,605 --> 00:51:48,764
was watching, like, the year by year number

1368
00:51:48,764 --> 00:51:51,744
one hits because there's a video out there.

1369
00:51:52,525 --> 00:51:54,385
All the different record labels

1370
00:51:55,500 --> 00:51:56,000
and

1371
00:51:56,460 --> 00:51:59,260
how they competed against each other in different

1372
00:51:59,260 --> 00:52:00,320
styles of music

1373
00:52:00,699 --> 00:52:03,260
and with different artists, and they were looking

1374
00:52:03,260 --> 00:52:04,000
for something.

1375
00:52:04,460 --> 00:52:05,199
And nowadays,

1376
00:52:05,820 --> 00:52:07,920
all music is owned by, like, three companies.

1377
00:52:08,300 --> 00:52:09,045
All of it.

1378
00:52:09,525 --> 00:52:11,545
And and you get to this homogeneity

1379
00:52:12,085 --> 00:52:12,744
that's like,

1380
00:52:13,045 --> 00:52:15,224
no. It's just it's just,

1381
00:52:15,684 --> 00:52:16,184
stop.

1382
00:52:16,724 --> 00:52:19,045
So I don't know. Maybe that's that's just,

1383
00:52:19,605 --> 00:52:21,525
but that that's actually kind of what worries

1384
00:52:21,525 --> 00:52:23,610
me about the scenario you're talking about.

1385
00:52:23,930 --> 00:52:25,630
George's ideal pizza

1386
00:52:25,930 --> 00:52:27,150
is gonna be identical

1387
00:52:27,769 --> 00:52:29,070
to 500,000

1388
00:52:29,130 --> 00:52:31,070
other people's ideal pizzas,

1389
00:52:31,450 --> 00:52:33,390
and George is never gonna realize

1390
00:52:33,769 --> 00:52:36,250
that he's actually not getting his ideal pizza

1391
00:52:36,250 --> 00:52:36,750
anymore.

1392
00:52:37,215 --> 00:52:39,954
He's getting the average of all ideal pizzas.

1393
00:52:40,974 --> 00:52:41,934
And that's there's, like

1394
00:52:43,055 --> 00:52:44,894
But the but the thing here is we

1395
00:52:44,894 --> 00:52:47,695
didn't enter this talking about what you or

1396
00:52:47,695 --> 00:52:49,474
I or Tom as consumers

1397
00:52:50,175 --> 00:52:51,715
would want from a ubiquitous

1398
00:52:52,440 --> 00:52:52,940
data

1399
00:52:53,320 --> 00:52:54,059
lake of

1400
00:52:54,519 --> 00:52:55,659
cellular capability.

1401
00:52:55,960 --> 00:52:58,519
We talked about what are the companies going

1402
00:52:58,519 --> 00:53:00,679
to want to do. Yep. What are they

1403
00:53:00,679 --> 00:53:01,900
trying to sell us?

1404
00:53:02,280 --> 00:53:04,699
And I gotta tell you, it's beams.

1405
00:53:05,159 --> 00:53:07,954
Yeah. It's beams. It's beam forming.

1406
00:53:09,454 --> 00:53:12,494
Yeah. Instead of instead of ball bearings now,

1407
00:53:12,494 --> 00:53:13,235
it's beams

1408
00:53:14,494 --> 00:53:15,315
or plastics.

1409
00:53:16,575 --> 00:53:18,575
The future is plastic. Wait a minute. Isn't

1410
00:53:18,575 --> 00:53:21,775
that, is that, Dustin Hoffman? Is that the

1411
00:53:21,775 --> 00:53:22,275
graduate?

1412
00:53:22,599 --> 00:53:23,339
It is.

1413
00:53:24,119 --> 00:53:26,679
The plastics. All ball bearings or the yeah.

1414
00:53:26,679 --> 00:53:28,059
Or plastics? Plastics.

1415
00:53:28,519 --> 00:53:30,519
Yeah. Yeah. It actually goes all the way

1416
00:53:30,519 --> 00:53:32,139
back to It's a Wonderful Life.

1417
00:53:32,679 --> 00:53:35,079
Plastics, I tell you. I'm offering you a

1418
00:53:35,079 --> 00:53:37,980
a place on the Ground Floor of plastics.

1419
00:53:38,414 --> 00:53:38,914
Like

1420
00:53:41,775 --> 00:53:43,855
Yeah. George got the better deal. He got

1421
00:53:43,855 --> 00:53:45,394
the girl. Just say it.

1422
00:53:46,574 --> 00:53:48,655
You have your plastics. I'll take the girl.

1423
00:53:48,655 --> 00:53:49,155
Thanks.

1424
00:53:49,614 --> 00:53:52,559
You made the right choice. Yep. So anyway,

1425
00:53:52,940 --> 00:53:54,859
alright. Well, I think we've kinda beat this

1426
00:53:54,859 --> 00:53:55,519
to death,

1427
00:53:55,980 --> 00:53:58,000
and maybe we ought to give up

1428
00:53:58,780 --> 00:54:01,839
on any further six g for the moment.

1429
00:54:01,900 --> 00:54:03,980
We'll eventually have Doug Dawson on to talk

1430
00:54:03,980 --> 00:54:04,480
about

1431
00:54:04,994 --> 00:54:07,174
hype, how the hype cycle works on this.

1432
00:54:07,315 --> 00:54:09,494
And, so anyway alright. Great.

1433
00:54:09,875 --> 00:54:12,914
George, anything else before we No. I had

1434
00:54:12,914 --> 00:54:14,914
a blast talking here. I love you guys,

1435
00:54:14,914 --> 00:54:17,014
and Audrey's looking better and better.

1436
00:54:17,980 --> 00:54:20,059
He's got he's gotta do some Audrey needs

1437
00:54:20,059 --> 00:54:20,640
a haircut.

1438
00:54:21,660 --> 00:54:23,760
Oh, Audrey. He didn't mean it.

1439
00:54:26,300 --> 00:54:28,000
Okay. A leaf cut. Whatever.

1440
00:54:29,420 --> 00:54:31,039
So, Tom, anything else?

1441
00:54:31,934 --> 00:54:34,094
Nope. Alright. Nope. Cool. It's been good. Alright,

1442
00:54:34,094 --> 00:54:35,315
Tom. Say your word.

1443
00:54:38,175 --> 00:54:41,454
LinkedIn. Alright. Cool. George, where can people find

1444
00:54:41,454 --> 00:54:43,775
you if they want to follow you? Or

1445
00:54:44,015 --> 00:54:46,755
I'm on the IP Nick blog. Okay. Cool.

1446
00:54:47,420 --> 00:54:48,619
I'm Ross White. You can find me here

1447
00:54:48,619 --> 00:54:50,480
at the hedge, rule eleven dot tech.

1448
00:54:50,780 --> 00:54:51,599
On LinkedIn,

1449
00:54:52,460 --> 00:54:55,260
occasionally on x, more or less depends on

1450
00:54:55,260 --> 00:54:56,539
the time of day and what I feel

1451
00:54:56,539 --> 00:54:58,460
like. And most of the time, it's I

1452
00:54:58,460 --> 00:55:00,780
don't, but anyway. And you can also find

1453
00:55:00,780 --> 00:55:03,039
me connected to a five g network someplace.

1454
00:55:06,514 --> 00:55:09,335
Likely. Likely. Very, very likely possibility.

1455
00:55:09,755 --> 00:55:10,255
And,

1456
00:55:10,675 --> 00:55:11,175
anyway,

1457
00:55:11,954 --> 00:55:12,454
alright.

1458
00:55:12,994 --> 00:55:14,835
We know we live in an attention driven

1459
00:55:14,835 --> 00:55:16,434
economy, and we thank you for listening all

1460
00:55:16,434 --> 00:55:17,954
the way to the bitter end of this

1461
00:55:17,954 --> 00:55:19,769
episode of The Hedge, and and we will

1462
00:55:19,769 --> 00:55:20,269
catch

1463
00:55:22,090 --> 00:55:23,150
you next