1
00:00:00,040 --> 00:00:03,359
Speaker 1: Twenty twenty six, twenty twenty seven, and twenty twenty eight.

2
00:00:03,480 --> 00:00:06,719
Speaker 2: Three years. It sounds like a standard lease agreement, doesn't it.

3
00:00:06,759 --> 00:00:07,200
But it's not.

4
00:00:07,440 --> 00:00:10,720
Speaker 1: No, it certainly doesn't feel like just another three years.

5
00:00:11,000 --> 00:00:13,800
You know, Usually when we sit down to unravel these threads,

6
00:00:14,480 --> 00:00:17,199
looking at the tech landscape, the future of humanity, the

7
00:00:17,280 --> 00:00:21,800
usual light dinner table conversation, we are casting our gays.

8
00:00:21,559 --> 00:00:23,199
Speaker 2: Pretty far out decades out.

9
00:00:23,320 --> 00:00:26,320
Speaker 1: Usually, yeah, exactly, we talked to twenty forty. We're talking

10
00:00:26,359 --> 00:00:29,160
about fusion energy coming in twenty years like it always has,

11
00:00:29,320 --> 00:00:32,399
or you know, colonies on Mars. But the intelligence deck

12
00:00:32,439 --> 00:00:35,920
we have on the desk today, this isn't about someday.

13
00:00:36,000 --> 00:00:38,960
Speaker 2: It is entirely different. We aren't talking about a distant horizon.

14
00:00:39,359 --> 00:00:41,880
We were talking about a window of time that is

15
00:00:41,920 --> 00:00:43,439
staring us right in the face.

16
00:00:43,560 --> 00:00:44,960
Speaker 1: The next thirty six months.

17
00:00:44,679 --> 00:00:46,719
Speaker 2: The next thirty six months, and according to the analysis

18
00:00:46,759 --> 00:00:50,560
we've pulled, specifically looking at the first mover's intelligence reports

19
00:00:50,880 --> 00:00:54,679
and the breakdown from doctor Julia McCoy, these aren't just

20
00:00:54,719 --> 00:00:58,280
sequential years. They are designated as a phase shift.

21
00:00:58,399 --> 00:01:01,000
Speaker 1: A phase shift I want to on that term because

22
00:01:01,039 --> 00:01:07,000
it sounds cool, but it also implies something volatile, unpredictable.

23
00:01:07,319 --> 00:01:10,719
Speaker 2: It is volatile. In physics, a phase shift isn't a

24
00:01:10,760 --> 00:01:12,319
gradual slope. It's a cliff.

25
00:01:12,640 --> 00:01:12,879
Speaker 1: Right.

26
00:01:13,000 --> 00:01:16,239
Speaker 2: Think about water. You heat it up, it gets hotter hotter.

27
00:01:16,439 --> 00:01:18,760
That's linear, you can predict it. But at one hundred

28
00:01:18,799 --> 00:01:20,560
degrees celsius, it doesn't just get.

29
00:01:20,359 --> 00:01:22,599
Speaker 1: Hotter, It becomes something else entirely.

30
00:01:22,719 --> 00:01:26,959
Speaker 2: It turns to steam. It expands sixteen hundred times in volume,

31
00:01:27,359 --> 00:01:30,640
It changes its fundamental state of matter. That is what

32
00:01:30,760 --> 00:01:32,719
human capability is about to undergo.

33
00:01:33,200 --> 00:01:36,519
Speaker 1: So we aren't looking at incremental improvements. This isn't a

34
00:01:36,599 --> 00:01:39,159
slightly faster phone, a slightly brighter screen.

35
00:01:39,040 --> 00:01:42,359
Speaker 2: No, not at all. We are looking at mastering complexity

36
00:01:42,599 --> 00:01:46,079
at a level our species has literally never achieved before.

37
00:01:46,159 --> 00:01:49,159
Speaker 1: And the catalyst for this steam explosion, it's the holy crap,

38
00:01:49,200 --> 00:01:52,040
it's here moment that I think a lot of people

39
00:01:52,200 --> 00:01:54,359
missed while they were scrolling through social.

40
00:01:54,040 --> 00:01:58,040
Speaker 2: Media, exactly the moment where quantum computing went from a

41
00:01:58,079 --> 00:02:00,799
physics experiment and maybe someday thing in a lab to

42
00:02:00,879 --> 00:02:02,920
a reality. It's happening right now.

43
00:02:03,159 --> 00:02:06,079
Speaker 1: Welcome to thrilling Threads. I'm your host, here to pull

44
00:02:06,079 --> 00:02:08,400
on the loose ends of the future until the whole picture.

45
00:02:08,120 --> 00:02:10,159
Speaker 2: Unravels, and I'm here to help make sense of the

46
00:02:10,199 --> 00:02:12,000
tangled mess we find underneath.

47
00:02:12,319 --> 00:02:15,240
Speaker 1: Today, we have a massive thread to pull. It's a collision,

48
00:02:15,280 --> 00:02:19,599
really a head on crash between two titans that are

49
00:02:19,639 --> 00:02:22,960
currently dominating the headlines, but rarely in the same sentence.

50
00:02:23,560 --> 00:02:26,280
In one corner, we have the energy hungry beast that

51
00:02:26,319 --> 00:02:28,639
has taken over our lives, our work, and our stock

52
00:02:28,680 --> 00:02:30,639
markets artificial.

53
00:02:30,159 --> 00:02:33,319
Speaker 2: Intelligence, and in the other corner, the physics bending solution

54
00:02:33,560 --> 00:02:35,319
that might be the only thing that can save in

55
00:02:35,800 --> 00:02:36,879
quantum computing.

56
00:02:37,199 --> 00:02:39,719
Speaker 1: And just to set the table, we are working off

57
00:02:39,759 --> 00:02:43,680
some incredible intelligence reports. Today. We've got analysis on breakthroughs

58
00:02:43,719 --> 00:02:47,000
from Google and Microsoft that are fresh, I mean early

59
00:02:47,039 --> 00:02:50,159
twenty twenty five, twenty twenty six fresh. This isn't old news.

60
00:02:50,479 --> 00:02:53,360
Speaker 2: No, this is happening as we speak. And the mission

61
00:02:53,439 --> 00:02:56,599
for this thread is critical. It's not just academic. We

62
00:02:56,639 --> 00:02:59,680
need to figure out if Google's recent quantum breakthrough has

63
00:02:59,680 --> 00:03:03,639
actually solve the AI energy crisis, or if Microsoft has

64
00:03:03,680 --> 00:03:05,800
a secret weapon hidden up its sleeve that's about to

65
00:03:05,840 --> 00:03:06,960
steal the entire show.

66
00:03:07,120 --> 00:03:09,240
Speaker 1: Okay, let's unpack this because I think most people hear

67
00:03:09,280 --> 00:03:13,599
AI and quantum and the default too cool faster computers.

68
00:03:13,680 --> 00:03:17,879
Speaker 2: Right, my games will look better exactly, Maybe Syria will

69
00:03:17,879 --> 00:03:20,360
finally understand what I'm asking. But you're telling me this

70
00:03:20,439 --> 00:03:22,759
is actually an energy story at its core.

71
00:03:23,199 --> 00:03:25,879
Speaker 1: It is fundamentally an energy story. In fact, if you

72
00:03:25,960 --> 00:03:28,879
don't look at it through the lens of thermodynamics, you

73
00:03:29,000 --> 00:03:32,199
miss the entire economic reality of what's happening. You missed

74
00:03:32,199 --> 00:03:32,800
the whole point.

75
00:03:32,960 --> 00:03:34,599
Speaker 2: So lay up the scene for us. Let's start with

76
00:03:34,639 --> 00:03:39,840
the problem. We know. AI is blowing up chat GPT, Gemini, Claude,

77
00:03:40,560 --> 00:03:44,199
the image generators. It feels like magic, but magic usually

78
00:03:44,280 --> 00:03:46,879
has a cost. What is the cost of that magic?

79
00:03:46,960 --> 00:03:52,840
Speaker 1: The cost is electricity, massive, gluttonous, almost unbelievable amounts of it.

80
00:03:52,919 --> 00:03:53,680
Speaker 2: Put in perspective.

81
00:03:53,719 --> 00:03:56,039
Speaker 1: Okay, so you look at the trajectory. The big picture

82
00:03:56,159 --> 00:03:58,960
number is that data centers are set to double their

83
00:03:59,000 --> 00:04:02,599
electricity consumption by twenty thirty double. That's already huge, it is,

84
00:04:02,639 --> 00:04:05,280
But that's just the aggregate number. It's a bit misleading

85
00:04:05,319 --> 00:04:08,400
because it smooths things out. The scary part is the density.

86
00:04:08,680 --> 00:04:10,840
Speaker 2: The density, what do you mean? I mean the power

87
00:04:10,919 --> 00:04:13,800
draw of a single piece of equipment a single high

88
00:04:13,919 --> 00:04:17,600
end AI chip like the Nvidia H one hundreds or

89
00:04:17,639 --> 00:04:21,040
the new Blackwall architecture we're seeing roll out, running at

90
00:04:21,079 --> 00:04:25,120
full training capacity, draws an incredible amount of power.

91
00:04:25,720 --> 00:04:26,199
Speaker 1: How much?

92
00:04:26,759 --> 00:04:29,759
Speaker 2: We are talking about a single chip using as much

93
00:04:29,800 --> 00:04:31,839
power as an entire American household?

94
00:04:31,839 --> 00:04:34,399
Speaker 1: Wait, hold on, let me just process that one chip

95
00:04:34,839 --> 00:04:37,120
a piece of silicon the size of a dinner plate.

96
00:04:37,319 --> 00:04:40,439
Speaker 2: Roughly. Yes, When you factor in the chip itself, the

97
00:04:40,480 --> 00:04:43,519
memory transfer, and crucially the cooling required to keep it

98
00:04:43,519 --> 00:04:46,680
from melting into slag, you are looking at kilowatts of

99
00:04:46,800 --> 00:04:50,040
draw for one unit. Now imagine a data center.

100
00:04:50,199 --> 00:04:51,160
Speaker 1: Do not just using one.

101
00:04:51,199 --> 00:04:53,199
Speaker 2: We aren't building them with a few hundred chips anymore.

102
00:04:53,240 --> 00:04:55,920
We are building what they call superclusters with tens of

103
00:04:55,959 --> 00:04:58,680
thousands of them. Some plans call for hundreds of thousands.

104
00:04:58,800 --> 00:05:01,800
Speaker 1: So we are essentially building small cities that do nothing

105
00:05:01,839 --> 00:05:04,720
but sit there, hum and eat electricity to run a

106
00:05:04,800 --> 00:05:05,240
chat butt.

107
00:05:05,319 --> 00:05:08,120
Speaker 2: That is exactly what we're doing, and we are hitting

108
00:05:08,199 --> 00:05:11,360
a wall, a physical energy based wall. We call it

109
00:05:11,360 --> 00:05:12,319
the consumption crisis.

110
00:05:12,360 --> 00:05:13,839
Speaker 1: The consumption crisis.

111
00:05:14,160 --> 00:05:17,279
Speaker 2: Yeah, classical computing, the silicon transistors we've used for fifty

112
00:05:17,360 --> 00:05:22,720
years cannot keep growing exponentially without literally destroying the energy grid.

113
00:05:22,839 --> 00:05:24,680
There just isn't enough power to go around.

114
00:05:24,920 --> 00:05:27,240
Speaker 1: So we're driving a Ferrari, but we're running out of

115
00:05:27,279 --> 00:05:28,000
gas stations.

116
00:05:28,160 --> 00:05:30,480
Speaker 2: It's worse. It's much worse than that. We're driving a

117
00:05:30,519 --> 00:05:33,439
Ferrari and the engine is getting so hot it's about

118
00:05:33,480 --> 00:05:35,800
to melt the chassis, and the only way to cool

119
00:05:35,800 --> 00:05:37,720
it down is to pour more gas on it.

120
00:05:37,920 --> 00:05:40,360
Speaker 1: That sounds problematic. It sounds like a death spiral.

121
00:05:40,439 --> 00:05:44,240
Speaker 2: It's a thermodynamic death spiral. And that's not hyperbole.

122
00:05:44,360 --> 00:05:47,360
Speaker 1: Okay, this brings us to the physics of why. This

123
00:05:47,480 --> 00:05:49,879
is the part that I think is so fascinating. I

124
00:05:49,879 --> 00:05:53,759
saw Landauer's principle in our source notes. This sounds like

125
00:05:53,800 --> 00:05:56,480
the kind of thing that gave me nightmares in college physics.

126
00:05:57,079 --> 00:05:59,319
But it seems to be the villain of this story.

127
00:05:59,399 --> 00:06:00,959
Speaker 2: It is the fundamental villain.

128
00:06:01,040 --> 00:06:04,040
Speaker 1: Yes, break it down for us. Why are our computers

129
00:06:04,079 --> 00:06:05,920
so ridiculously inefficient?

130
00:06:06,040 --> 00:06:09,079
Speaker 2: It's fascinating, Actually, Rolf Landauer back in nineteen sixty one

131
00:06:09,079 --> 00:06:11,639
at IBM, realized something profound.

132
00:06:11,240 --> 00:06:13,199
Speaker 1: About information sixty years ago.

133
00:06:13,560 --> 00:06:17,759
Speaker 2: Over sixty years ago, he realized that information is physical.

134
00:06:18,319 --> 00:06:21,199
It's not just an abstract concept floating in the ether.

135
00:06:21,399 --> 00:06:24,399
It has to be stored represented by something real.

136
00:06:24,600 --> 00:06:27,519
Speaker 1: Okay, that makes sense, a magnetic bit, a voltage in

137
00:06:27,560 --> 00:06:28,720
a transistor exactly.

138
00:06:28,879 --> 00:06:31,920
Speaker 2: And Landauer's principle states that there is a minimum energy

139
00:06:31,920 --> 00:06:35,600
cost to erasing information. This is tied to entropy, the

140
00:06:35,639 --> 00:06:38,120
second law of thermodynamics.

141
00:06:37,439 --> 00:06:40,279
Speaker 1: Entropy, the tendency towards chaos and disorder.

142
00:06:40,360 --> 00:06:43,600
Speaker 2: Precisely, every time a classical computer deletes a bit, turns

143
00:06:43,600 --> 00:06:45,839
a one to a zero, or wipes a memory block,

144
00:06:46,199 --> 00:06:48,120
it has to dissipate energy as heat.

145
00:06:48,560 --> 00:06:51,720
Speaker 1: It must, so the act of forgetting creates heat.

146
00:06:52,079 --> 00:06:53,959
Speaker 2: That is a perfect way to put it. You cannot

147
00:06:54,000 --> 00:06:57,360
destroy information without increasing the entropy of the universe.

148
00:06:57,399 --> 00:06:59,079
Speaker 1: That feels so profound it is.

149
00:06:59,399 --> 00:07:02,040
Speaker 2: Think about us standard logic gait in your computer like

150
00:07:02,079 --> 00:07:04,319
an A and D gate. It takes two inputs, say

151
00:07:04,319 --> 00:07:07,199
a one in a zero, and gives one output a zero. Okay,

152
00:07:07,519 --> 00:07:09,800
but what happened to the information about the two separate inputs?

153
00:07:09,800 --> 00:07:11,519
You had two bits of information, Now you have one.

154
00:07:11,560 --> 00:07:12,399
Where did the other one go?

155
00:07:12,560 --> 00:07:14,680
Speaker 1: It was erased, It's gone, it's gone.

156
00:07:14,720 --> 00:07:19,079
Speaker 2: And that erasure manifests physically as heat every single.

157
00:07:18,800 --> 00:07:23,560
Speaker 1: Time, and classical computers do this how often, billions trillions

158
00:07:23,560 --> 00:07:25,680
of time to second in every single chip. It's a

159
00:07:25,680 --> 00:07:29,439
relentless cycle of right erase right erase. As the problems

160
00:07:29,480 --> 00:07:31,720
we try to solve, like training a massive AI model

161
00:07:31,759 --> 00:07:36,160
with trillions of parameters, get more complex that cycle accelerates.

162
00:07:36,360 --> 00:07:39,360
Speaker 2: So more complexity means more erasing.

163
00:07:39,000 --> 00:07:42,680
Speaker 1: Which means more energy use, which means more heat. The

164
00:07:42,720 --> 00:07:46,079
heat output becomes unmanageable. That's why you see these crazy

165
00:07:46,079 --> 00:07:49,079
stories about data centers being built under water or in

166
00:07:49,120 --> 00:07:51,720
the Arctic circle. They're not doing it for fun. They're

167
00:07:51,759 --> 00:07:53,439
doing it because of Landauer's principle.

168
00:07:53,519 --> 00:07:55,399
Speaker 2: It's the entropy tax. We have to pay it every

169
00:07:55,399 --> 00:07:56,360
time we compute the.

170
00:07:56,399 --> 00:07:59,399
Speaker 1: Entropy tax exactly, and with classical chips we are paying

171
00:07:59,399 --> 00:08:01,839
a higher and higher tax rate. The faster we try

172
00:08:01,839 --> 00:08:04,000
to go, we're getting taxed into oblivion.

173
00:08:04,120 --> 00:08:07,000
Speaker 2: So enter the hero of our story, the tax evator,

174
00:08:07,639 --> 00:08:09,040
quantum computing.

175
00:08:08,920 --> 00:08:10,680
Speaker 1: The ultimate tax shelter.

176
00:08:10,839 --> 00:08:14,079
Speaker 2: Now, I've heard about superposition and entanglement, the spooky stuff

177
00:08:14,079 --> 00:08:16,639
where things are in two places at once. But you're saying,

178
00:08:16,720 --> 00:08:19,360
the real killer feature for the energy grid isn't the

179
00:08:19,399 --> 00:08:23,519
spooky stuff. It's reversibility. This is the mind explosion moment

180
00:08:23,560 --> 00:08:27,040
for most people. Yes, the spooky stuff is cool, it's

181
00:08:27,079 --> 00:08:29,800
what gives it the power, But the efficiency comes from

182
00:08:29,800 --> 00:08:30,360
something else.

183
00:08:30,639 --> 00:08:31,000
Speaker 1: Okay.

184
00:08:31,279 --> 00:08:35,559
Speaker 2: Quantum computers handle information in a fundamentally different way. In

185
00:08:35,639 --> 00:08:40,039
quantum mechanics, operations are modeled by what we call unitary transformations.

186
00:08:40,399 --> 00:08:44,960
Speaker 1: Whoa hold on unitary transformations? You lost me unpack that, okay?

187
00:08:45,039 --> 00:08:49,279
Speaker 2: Okay. To put simply, it just means the operations are reversible.

188
00:08:49,480 --> 00:08:52,039
You can run the computation forward and then you can

189
00:08:52,120 --> 00:08:54,480
run it backward to get to the exact starting state.

190
00:08:54,600 --> 00:08:56,879
Speaker 1: Reversible meaning they don't erase.

191
00:08:56,720 --> 00:08:58,919
Speaker 2: Meaning they can uncompute.

192
00:08:59,279 --> 00:09:03,320
Speaker 1: Okay. An analogy. Uncomputing sounds like time travel. Doesn't sound real.

193
00:09:03,440 --> 00:09:06,200
Speaker 2: Think of it like walking through a pristine field of snow.

194
00:09:06,480 --> 00:09:09,879
Classical computing is like stomping through the snow, digging holes,

195
00:09:09,960 --> 00:09:13,320
kicking it around. It's messy. To reset the path or

196
00:09:13,360 --> 00:09:15,960
hide your tracks, you have to shovel fresh snow over it.

197
00:09:16,320 --> 00:09:17,399
That's a lot of work.

198
00:09:17,240 --> 00:09:20,720
Speaker 1: Right, lots of sweaty work shoveling snow. And that shoveling

199
00:09:20,759 --> 00:09:20,960
is the.

200
00:09:20,960 --> 00:09:27,120
Speaker 2: Heat generation exactly. That's Landauer's principle and action reversible computing.

201
00:09:27,240 --> 00:09:32,159
Quantum computing is like carefully walking backward, stepping exactly into

202
00:09:32,200 --> 00:09:34,240
your own footprints until you're back at the start.

203
00:09:34,360 --> 00:09:36,480
Speaker 1: You haven't disturbed the snowfield.

204
00:09:35,919 --> 00:09:38,559
Speaker 2: You haven't increased the chaos or entropy of the system.

205
00:09:38,600 --> 00:09:40,519
You've left it exactly as you found it.

206
00:09:40,960 --> 00:09:43,480
Speaker 1: So because you aren't destroying the state of the pit,

207
00:09:44,360 --> 00:09:47,759
you aren't triggering Landauer's penalty. You aren't paying the tax.

208
00:09:47,600 --> 00:09:51,279
Speaker 2: Exactly, you evade the tax for the core computation. At least,

209
00:09:51,320 --> 00:09:54,039
there's still energy needed to run the machine. But the

210
00:09:54,080 --> 00:09:58,120
computation itself is in principle nearly energy free.

211
00:09:58,240 --> 00:10:00,519
Speaker 1: That is wild. So it's not it's just that they

212
00:10:00,519 --> 00:10:04,039
are faster. Everyone says quantum is faster. Yeah, but for

213
00:10:04,080 --> 00:10:07,679
the energy grid, the headline is that they are sustainable.

214
00:10:07,159 --> 00:10:10,320
Speaker 2: For complex problems. Yes, we aren't talking about ten percent

215
00:10:10,399 --> 00:10:13,120
better battery life on your laptop. We are talking about

216
00:10:13,120 --> 00:10:16,559
doing calculations that would require a nuclear power plant's worth

217
00:10:16,559 --> 00:10:19,879
of energy on a classical supercomputer, okay, and doing that

218
00:10:19,919 --> 00:10:22,639
same calculation using the energy of a light bulb on

219
00:10:22,679 --> 00:10:25,759
a quantum machine. It's a different universe of efficiency.

220
00:10:26,000 --> 00:10:28,480
Speaker 1: That is the miracle we need. I mean, if we

221
00:10:28,480 --> 00:10:31,759
want AI to keep getting smarter, to reach that personal

222
00:10:31,799 --> 00:10:34,799
intelligence level. The reports talk about where it's actually useful

223
00:10:35,080 --> 00:10:36,480
without burning down the planet.

224
00:10:36,679 --> 00:10:39,440
Speaker 2: We need this switch, we do, and that's why the

225
00:10:39,519 --> 00:10:42,240
race between the tech giants has suddenly shifted from R

226
00:10:42,279 --> 00:10:45,960
and D science projects to survival mode. This is no

227
00:10:46,039 --> 00:10:48,639
longer a choice, It's a necessity.

228
00:10:48,559 --> 00:10:51,360
Speaker 1: Which brings us to the first giant Google.

229
00:10:51,159 --> 00:10:54,720
Speaker 2: Google Quantum AI, the behemoth, the one that's been making

230
00:10:54,759 --> 00:10:56,279
the most noise for the longest time.

231
00:10:56,440 --> 00:10:58,039
Speaker 1: They've been at this for a while. I remember the

232
00:10:58,120 --> 00:11:00,639
quantum supremacy announcement back in twenty tine nineteen with the

233
00:11:00,639 --> 00:11:01,639
Sycamore processor.

234
00:11:01,799 --> 00:11:02,480
Speaker 2: I remember that.

235
00:11:02,679 --> 00:11:05,919
Speaker 1: They claimed they did a calculation in two hundred seconds

236
00:11:06,200 --> 00:11:11,240
that would take a supercomputer ten thousand years. IBM disputed it,

237
00:11:11,279 --> 00:11:11,559
of course.

238
00:11:11,600 --> 00:11:14,000
Speaker 2: It was a whole thing, a lot of academic squabbling, yes,

239
00:11:14,440 --> 00:11:16,279
but it put the world on notice.

240
00:11:16,320 --> 00:11:18,320
Speaker 1: But the first mover's report says they just hit a

241
00:11:18,360 --> 00:11:21,519
much more significant milestone in this twenty twenty five twenty

242
00:11:21,559 --> 00:11:24,480
twenty six window, something that makes a premacy look like

243
00:11:24,480 --> 00:11:25,240
a warm up act.

244
00:11:25,240 --> 00:11:28,720
Speaker 2: They demonstrated the first logical quibit prototype.

245
00:11:28,879 --> 00:11:32,759
Speaker 1: Okay, jargon alert, we need to beat precise here logical quibbit?

246
00:11:34,039 --> 00:11:35,919
How is that different from the quibits we've heard about

247
00:11:35,919 --> 00:11:38,559
for the last decade. This sounds like a marketing term.

248
00:11:38,759 --> 00:11:42,000
Speaker 2: It's not marketing, it's well, it's everything. This is the

249
00:11:42,080 --> 00:11:45,759
single biggest hurdle in the field. Think of a regular

250
00:11:45,879 --> 00:11:51,000
physical quibbot as a very fidgety, anxious child.

251
00:11:51,120 --> 00:11:52,080
Speaker 1: Okay, I can picture that.

252
00:11:52,159 --> 00:11:56,519
Speaker 2: It's easily distracted. A tiny fluctuation in temperature, a stray

253
00:11:56,600 --> 00:12:00,240
electromagnetic wave from your phone, a cosmic ray from a

254
00:12:00,240 --> 00:12:04,000
distant supernova. Anything makes it lose its memory. This is

255
00:12:04,039 --> 00:12:07,159
called decoherence. It's incredibly error prone.

256
00:12:07,240 --> 00:12:10,559
Speaker 1: So up until now quantum computers were like really fast

257
00:12:10,600 --> 00:12:13,360
calculators that got the answer wrong fifty percent of the time.

258
00:12:13,279 --> 00:12:15,519
Speaker 2: Or the signal would just dissolve into noise before you

259
00:12:15,519 --> 00:12:17,480
could even finish the math. You'd ask it what's two

260
00:12:17,519 --> 00:12:19,559
plus two? And before it could answer, it would forget

261
00:12:19,559 --> 00:12:19,960
the question.

262
00:12:20,320 --> 00:12:22,840
Speaker 1: Not very useful. So what's the logical quibit? How does

263
00:12:22,879 --> 00:12:24,519
it fix a fidgety child problem?

264
00:12:24,559 --> 00:12:27,960
Speaker 2: A logical quibit is error correction. Imagine you take that

265
00:12:28,039 --> 00:12:30,799
one fidgety child and you surround him with one hundred

266
00:12:30,879 --> 00:12:33,840
other children. A whole classroom of them a whole classroom,

267
00:12:34,000 --> 00:12:36,679
and they all hold hands. They form a collective. If

268
00:12:36,720 --> 00:12:38,639
the one in the middle trips or gets distracted, the

269
00:12:38,639 --> 00:12:41,240
others hold them up. They keep them in line. The

270
00:12:41,279 --> 00:12:44,200
information isn't stored in one kid. It's stored in the

271
00:12:44,240 --> 00:12:45,080
pattern of the group.

272
00:12:45,159 --> 00:12:49,080
Speaker 1: It's redundancy like a rate array for hard drives, but

273
00:12:49,120 --> 00:12:50,039
for quantum bits.

274
00:12:50,120 --> 00:12:54,200
Speaker 2: It's an active error correction system, specifically something called surface code.

275
00:12:54,639 --> 00:12:58,480
By weaving many physical quibets together, Google created a single

276
00:12:58,559 --> 00:13:02,320
logical unit that stays stable. They prove that making the

277
00:13:02,320 --> 00:13:04,639
group bigger actually reduced the error rate.

278
00:13:04,799 --> 00:13:07,000
Speaker 1: So the more kids holding hands, the more stable the

279
00:13:07,039 --> 00:13:07,799
group becomes.

280
00:13:07,840 --> 00:13:10,840
Speaker 2: Exactly and that was the Holy Grail moment. It proved

281
00:13:10,840 --> 00:13:13,320
that this wasn't an insurmountable physics problem.

282
00:13:13,440 --> 00:13:16,960
Speaker 1: So demonstrating this means they crack the code on reliability.

283
00:13:17,799 --> 00:13:20,799
It's no longer just theoretical physics. It works. It's an

284
00:13:20,840 --> 00:13:23,120
engineering problem. Now, it's a scaling problem.

285
00:13:23,159 --> 00:13:26,200
Speaker 2: It is. It's the transistor moment. It's the jump from

286
00:13:26,279 --> 00:13:29,879
vacuum tubes to solid state and Google isn't stopping there.

287
00:13:30,080 --> 00:13:33,960
Their public roadmap is aggressive. They're planning to achieve one

288
00:13:34,039 --> 00:13:35,919
million quibuts in the coming years.

289
00:13:36,000 --> 00:13:39,320
Speaker 1: One million physical quibbits just for context. Right now we're

290
00:13:39,320 --> 00:13:42,759
dealing with what a few hundred maybe a thousand in

291
00:13:42,799 --> 00:13:43,879
the best lab systems.

292
00:13:44,080 --> 00:13:47,279
Speaker 2: In terms of high quality physical equibitts, yes, but in

293
00:13:47,360 --> 00:13:50,399
terms of logical quibits, the currency that actually matters for

294
00:13:50,559 --> 00:13:54,039
running useful programs, we are barely at step one. We

295
00:13:54,120 --> 00:13:56,159
have maybe one two a handful.

296
00:13:56,440 --> 00:13:59,480
Speaker 1: So jumping to a million physical equibits to support thousands

297
00:13:59,519 --> 00:14:02,120
of logic ones is I mean, that's not a step,

298
00:14:02,279 --> 00:14:03,799
that's a leap to another planet.

299
00:14:03,919 --> 00:14:07,080
Speaker 2: It's a total shift in processing capability. It's the difference

300
00:14:07,080 --> 00:14:09,799
between counting on your fingers and simulating a galaxy.

301
00:14:10,159 --> 00:14:13,879
Speaker 1: But there's always a butt. There has to be a catch.

302
00:14:14,000 --> 00:14:15,000
Speaker 2: There is always a butt.

303
00:14:15,080 --> 00:14:17,600
Speaker 1: And here's where it gets really interesting for our energy story.

304
00:14:18,039 --> 00:14:21,480
The report points out a massive achilles heel for Google.

305
00:14:22,279 --> 00:14:23,600
The big chill yay.

306
00:14:23,639 --> 00:14:27,919
Speaker 2: The reality check. Google uses what are called superconducting quantum computers.

307
00:14:28,120 --> 00:14:30,759
That's their chosen architecture. It means it means they rely

308
00:14:30,840 --> 00:14:34,879
on materials that have zero electrical resistance, an amazing property.

309
00:14:35,120 --> 00:14:37,960
But to get that property super conductivity, you need to

310
00:14:37,960 --> 00:14:38,840
cool the chips down.

311
00:14:39,000 --> 00:14:40,440
Speaker 1: How cool all we talk and put it into the

312
00:14:40,440 --> 00:14:41,000
fridge cool?

313
00:14:41,279 --> 00:14:42,960
Speaker 2: I wish, no, you need to get them to about

314
00:14:43,000 --> 00:14:44,080
fifteen millikelvin.

315
00:14:44,120 --> 00:14:45,639
Speaker 1: Put that in perspective for me, I don't know what

316
00:14:45,679 --> 00:14:46,440
a mil kelvin is.

317
00:14:46,480 --> 00:14:50,639
Speaker 2: Okay, so absolute zero. The coldest possible temperature in the

318
00:14:50,720 --> 00:14:55,679
universe is zero kelvin deep space. The vacuum between galaxies

319
00:14:55,799 --> 00:14:57,600
is about two point seven.

320
00:14:57,399 --> 00:14:59,720
Speaker 1: Kelvin, colder than deep space orders.

321
00:14:59,480 --> 00:15:01,960
Speaker 2: Of magnet colder than deep space. It is one of

322
00:15:01,960 --> 00:15:04,240
the coldest places in the known universe.

323
00:15:04,679 --> 00:15:07,919
Speaker 1: Inside that machine, we're talking about those giant gold and

324
00:15:08,039 --> 00:15:11,360
chandelier looking things we see in the press photos right exactly.

325
00:15:11,840 --> 00:15:15,120
Speaker 2: Those are called dilution refrigerators. They use a mix of

326
00:15:15,320 --> 00:15:19,759
helium three and helium four isotopes to literally pump heat

327
00:15:19,879 --> 00:15:24,159
out of the system stage by stage. They are incredibly complex,

328
00:15:24,320 --> 00:15:29,080
incredibly expensive, and crucially for our conversation, they are power hungry.

329
00:15:29,279 --> 00:15:32,039
Speaker 1: How power hungry? The report says running one of these

330
00:15:32,039 --> 00:15:34,679
systems consumes around twenty five kilowatts of power.

331
00:15:34,799 --> 00:15:37,679
Speaker 2: Twenty five kilowatts, and crucially, most of that electricity isn't

332
00:15:37,679 --> 00:15:39,919
doing any math, is not computing anything.

333
00:15:40,039 --> 00:15:41,200
Speaker 1: It's just running the freezer.

334
00:15:41,320 --> 00:15:43,919
Speaker 2: It's just keeping the system frozen. Its overhead, it's the

335
00:15:43,960 --> 00:15:46,759
cost of entry before you even do a single calculation.

336
00:15:47,159 --> 00:15:49,919
Speaker 1: It's massive overhead. If you want to scale this up

337
00:15:49,919 --> 00:15:53,080
to millions of quivts, you need acres of these fridges.

338
00:15:53,440 --> 00:15:55,960
You're building a warehouse of quigentic freezers.

339
00:15:56,039 --> 00:15:59,080
Speaker 2: That's the problem. You need a power substation just for

340
00:15:59,120 --> 00:16:01,039
the refrigeration plant for your computer.

341
00:16:01,200 --> 00:16:05,120
Speaker 1: So we're solving the AI energy crisis by building giant

342
00:16:05,159 --> 00:16:08,759
freezers that use a ton of energy. There's an irony,

343
00:16:08,759 --> 00:16:09,840
there's a huge irony.

344
00:16:09,879 --> 00:16:12,639
Speaker 2: It's a trade off. It's still better than classical for

345
00:16:12,679 --> 00:16:15,759
certain specific problems, but it's not the perfect free lunch

346
00:16:15,799 --> 00:16:18,399
we might hope for. And this is where the source

347
00:16:18,440 --> 00:16:19,440
draws are really sharp.

348
00:16:19,480 --> 00:16:24,600
Speaker 1: Contrast, right, it compares this to neutral atom quantum computers.

349
00:16:24,679 --> 00:16:26,919
Speaker 2: Neutral atom a completely different approach.

350
00:16:27,080 --> 00:16:27,799
Speaker 1: How do those work?

351
00:16:28,000 --> 00:16:31,200
Speaker 2: They work at room temperature. Well, the atoms themselves are

352
00:16:31,279 --> 00:16:34,440
laser cool to be almost motionless. But the machine itself

353
00:16:34,480 --> 00:16:37,039
sits in a standard server rack in a normal room,

354
00:16:37,399 --> 00:16:38,879
no giant golden.

355
00:16:38,600 --> 00:16:40,840
Speaker 1: Chandeliers, and the power draw the.

356
00:16:40,840 --> 00:16:44,559
Speaker 2: Report pegs them at roughly ten kilowatts.

357
00:16:43,960 --> 00:16:46,440
Speaker 1: So less than half that's already a huge difference in

358
00:16:46,480 --> 00:16:49,559
efficiency right out of the gate, a huge difference.

359
00:16:49,759 --> 00:16:53,279
Speaker 2: So you have this dynamic where Google is the clear

360
00:16:53,360 --> 00:16:56,879
leader in crevit quality and scale. They have the logical quibit,

361
00:16:57,600 --> 00:17:02,519
but they are hauling around this massive energy guzzling refrigeration baggage.

362
00:17:02,559 --> 00:17:05,400
Speaker 1: They are they are betting that their speed of development

363
00:17:05,440 --> 00:17:09,599
and their software ecosystem will outweigh the infrastructure cost. But

364
00:17:09,759 --> 00:17:14,440
that vulnerability, that massive energy bill, that's exactly where the

365
00:17:14,480 --> 00:17:18,119
wild cart steps in. Microsoft is sleeping giant. I feel

366
00:17:18,160 --> 00:17:19,880
like Microsoft has been the quiet kid in the back

367
00:17:19,880 --> 00:17:22,079
of the class on this one for a while. They've

368
00:17:22,079 --> 00:17:25,680
been talking about this topological stuff for a decade, and honestly,

369
00:17:25,720 --> 00:17:27,720
a lot of people in the field thought it was vaporware,

370
00:17:28,559 --> 00:17:29,160
a dead end.

371
00:17:29,240 --> 00:17:31,000
Speaker 2: A lot of people wrote them off completely, and.

372
00:17:30,920 --> 00:17:34,519
Speaker 1: Then suddenly, boom, February twenty twenty five, they dropped the mic.

373
00:17:34,599 --> 00:17:38,640
Speaker 2: They really did. They announced the Majorina one quantum chip.

374
00:17:38,799 --> 00:17:41,920
Speaker 1: Majorana sounds like a spell from Harry Potter, yeah, or

375
00:17:41,960 --> 00:17:42,759
apasta sauce.

376
00:17:42,960 --> 00:17:46,640
Speaker 2: It's named after a Tori Majorana, a brilliant, mysterious Italian

377
00:17:46,640 --> 00:17:50,440
physicist who theorized a particle that is its own antiparticle

378
00:17:50,440 --> 00:17:53,759
back in the nineteen thirties and then vanag vanished. Yeah,

379
00:17:53,799 --> 00:17:55,960
just disappeared off a boat. It's a whole other story.

380
00:17:56,680 --> 00:18:00,160
But the tech is what matters, they claim, and we

381
00:18:00,279 --> 00:18:03,279
have to emphasize claim because this is cutting edge to

382
00:18:03,359 --> 00:18:06,960
be able to fit over one million quivots on a

383
00:18:07,079 --> 00:18:08,559
chip the size of your palm.

384
00:18:08,880 --> 00:18:13,319
Speaker 1: Whoa stop rewind that Google needs a warehouse of fridges

385
00:18:13,359 --> 00:18:17,079
for a million kubits. Microsoft says the size of my palm.

386
00:18:17,240 --> 00:18:18,359
Speaker 2: The size of your palm?

387
00:18:18,519 --> 00:18:21,359
Speaker 1: Oh, I mean that defies the sense of scale we

388
00:18:21,440 --> 00:18:23,319
just established. That sounds like science fiction.

389
00:18:23,519 --> 00:18:26,200
Speaker 2: It comes down to the physics. It's a completely different paradigm.

390
00:18:26,279 --> 00:18:29,720
They are using what are called topological superconductor material.

391
00:18:29,880 --> 00:18:32,680
Speaker 1: Okay, expert mode, help me out. Topological. I know what

392
00:18:32,720 --> 00:18:35,440
a topographical map is with the mountains and valleys, but

393
00:18:35,480 --> 00:18:36,880
what is a topological computer.

394
00:18:37,000 --> 00:18:38,880
Speaker 2: Okay, Think of a knot in a piece of string.

395
00:18:39,240 --> 00:18:41,599
If you have a simple loop of string lying on

396
00:18:41,640 --> 00:18:43,839
the table, it's just a circle. If you tie a

397
00:18:43,920 --> 00:18:47,440
knot in it, that knot is topologically protected me meaning

398
00:18:47,440 --> 00:18:50,119
you can twist the string, stretch it, wiggle it around,

399
00:18:50,319 --> 00:18:54,480
vibrate the table. But the knot stays a knot. The

400
00:18:54,519 --> 00:18:57,920
fundamental property of knottedness doesn't change, right.

401
00:18:58,039 --> 00:18:59,680
Speaker 1: The only way to remove the knot is to.

402
00:18:59,599 --> 00:19:04,319
Speaker 2: Cut the exactly. Microsoft is using that principle to store information.

403
00:19:04,799 --> 00:19:08,759
In Google's machine, the information is fragile. It's like balancing

404
00:19:08,839 --> 00:19:11,720
a pencil on its tip. One tiny bump and it

405
00:19:11,759 --> 00:19:16,640
falls over. It decoheres, okay. In Microsoft's theoretical machine, the

406
00:19:16,759 --> 00:19:19,799
data is stored in the knot, specifically in the way

407
00:19:19,880 --> 00:19:24,240
these exotic quasi particles, these majorana particles, are braided around

408
00:19:24,240 --> 00:19:27,599
each other. The information is woven into the very physics

409
00:19:27,680 --> 00:19:28,400
of the material.

410
00:19:28,599 --> 00:19:31,160
Speaker 1: So the information is naturally robust. It's protected by.

411
00:19:31,079 --> 00:19:33,880
Speaker 2: The knot exactly. Local noise, a little bit of heat,

412
00:19:33,920 --> 00:19:37,200
a vibration, It doesn't untie the knot. So you don't

413
00:19:37,240 --> 00:19:40,400
need that massive army of one hundred physical quibits holding

414
00:19:40,440 --> 00:19:44,240
hands to make one logical quibbet. Their physical cuibbits are

415
00:19:44,440 --> 00:19:46,319
in theory already logical quibits.

416
00:19:46,359 --> 00:19:48,599
Speaker 1: That's the game changer. They don't need the error correction overhead.

417
00:19:48,599 --> 00:19:50,480
Speaker 2: They don't need the massive overhead. And if and it

418
00:19:50,559 --> 00:19:54,680
is a colossal texas sized I, this works at scale.

419
00:19:55,000 --> 00:19:56,559
Microsoft leapfrogs everyone.

420
00:19:56,640 --> 00:19:58,480
Speaker 1: It wouldn't just be a player. They would dominate the

421
00:19:58,519 --> 00:19:59,240
quantum space.

422
00:19:59,319 --> 00:20:01,440
Speaker 2: They'd own it. It would be a monopoly because of

423
00:20:01,480 --> 00:20:02,559
the form factor.

424
00:20:02,279 --> 00:20:04,759
Speaker 1: A palm sized chip versus a warehouse of freezers.

425
00:20:05,039 --> 00:20:07,440
Speaker 2: If you can put a million kubits on a palm

426
00:20:07,440 --> 00:20:11,400
sized chip that were except you know, slightly less extreme temperatures,

427
00:20:11,599 --> 00:20:13,359
you can put it in a server rack, You can

428
00:20:13,359 --> 00:20:14,880
put it in a satellite, you can put on a

429
00:20:14,880 --> 00:20:17,880
fighter jet. You don't need to build a specialized frozen

430
00:20:17,880 --> 00:20:19,000
cathedral to house it.

431
00:20:19,480 --> 00:20:24,000
Speaker 1: This reshapes the next decade entirely. If Microsoft pulls this off.

432
00:20:24,599 --> 00:20:28,400
Google's massive fridge farms look like dinosaurs.

433
00:20:28,599 --> 00:20:32,640
Speaker 2: They look like the eni At computer from the nineteen forties. Big, impressive,

434
00:20:33,000 --> 00:20:36,160
room sized, took a whole team to run, but completely

435
00:20:36,200 --> 00:20:38,279
and utterly obsolete the moment.

436
00:20:38,000 --> 00:20:40,519
Speaker 1: The microchip arrived. But I have to play Dell's advocate

437
00:20:40,559 --> 00:20:44,279
here sounds too good to be true. Microsoft has claimed

438
00:20:44,319 --> 00:20:47,000
to have found these midrona particles before, and then had

439
00:20:47,000 --> 00:20:51,359
to retract papers. The scientific community is, let's say.

440
00:20:51,240 --> 00:20:55,119
Speaker 2: Skeptical, extremely skeptical, and for good reason. In twenty eighteen

441
00:20:55,240 --> 00:20:58,079
and again in twenty twenty one, there were major retractions

442
00:20:58,119 --> 00:21:00,799
from research groups they were working with. It was embarrassing

443
00:21:00,839 --> 00:21:03,519
for the field and for Microsoft. But this time they've

444
00:21:03,519 --> 00:21:06,359
published data on something called the fractional quantum Hall effect,

445
00:21:06,440 --> 00:21:08,799
which is supposed to be the definitive fingerprint of this

446
00:21:08,920 --> 00:21:09,920
topological state.

447
00:21:10,160 --> 00:21:11,920
Speaker 1: So they think they really have it. This time.

448
00:21:12,200 --> 00:21:14,519
Speaker 2: They seem confident enough to put a product name on

449
00:21:14,599 --> 00:21:17,559
it and announce a roadmap. That's a big step, but

450
00:21:17,599 --> 00:21:18,839
it's high risk, high.

451
00:21:18,759 --> 00:21:21,400
Speaker 1: Reward, the highest reward in computing history. Probably.

452
00:21:21,400 --> 00:21:23,920
Speaker 2: I don't think that's an exaggeration, but we.

453
00:21:23,880 --> 00:21:25,880
Speaker 1: Can't count out the others. There are more than two

454
00:21:25,920 --> 00:21:29,359
horses in this race. The report mentions a third player,

455
00:21:30,039 --> 00:21:31,079
a dark horse.

456
00:21:31,880 --> 00:21:35,400
Speaker 2: Ion Q ion Q, Yes, the pure play.

457
00:21:35,640 --> 00:21:38,759
Speaker 1: They aren't Google or Microsoft. They're not a trillion dollar

458
00:21:38,839 --> 00:21:40,359
behemoth with a quantum division.

459
00:21:40,880 --> 00:21:43,400
Speaker 2: This is all they do right, market cap around what

460
00:21:43,519 --> 00:21:46,960
eighteen billion dollars, which is huge for a startup, but

461
00:21:47,000 --> 00:21:49,559
a drop in the bucket for their competitors. They use

462
00:21:49,640 --> 00:21:51,880
trapped ion architecture trapped ions.

463
00:21:51,920 --> 00:21:54,000
Speaker 1: Okay, so this is a third way. This isn't super

464
00:21:54,000 --> 00:21:57,200
conducting loops or topological knots. This is where they take

465
00:21:57,279 --> 00:22:00,119
single atoms, charged atoms and hold them in place with

466
00:22:00,200 --> 00:22:01,519
electromagnetic fields.

467
00:22:01,519 --> 00:22:05,160
Speaker 2: Correct, they use uurbium atoms. Usually they levitate a little

468
00:22:05,200 --> 00:22:07,480
line of them in a perfect vacuum chamber, and then

469
00:22:07,480 --> 00:22:10,839
they zap them with precisely tuned lasers to do the computing.

470
00:22:11,079 --> 00:22:13,039
Speaker 1: What's the advantage there? Why do it that way?

471
00:22:13,400 --> 00:22:17,640
Speaker 2: The advantage is that nature makes the quibbets. Every single

472
00:22:17,759 --> 00:22:21,559
uturbium atom is perfectly identical to every other uturbium atom

473
00:22:21,559 --> 00:22:25,680
in the universe. Ah, no manufacturing defense exactly In Google's approach,

474
00:22:25,720 --> 00:22:28,599
you have to build the quibbit on a chip using lithography.

475
00:22:29,160 --> 00:22:32,839
No two are ever perfectly alike. You get variations. Ion

476
00:22:32,960 --> 00:22:37,480
q uses nature's perfection. Their equipots are incredibly high quality

477
00:22:37,480 --> 00:22:38,000
and stable.

478
00:22:38,440 --> 00:22:40,319
Speaker 1: And what's their status? Where are they in the race?

479
00:22:40,839 --> 00:22:44,240
Speaker 2: They are currently unprofitable, which is expected, but they are

480
00:22:44,240 --> 00:22:47,519
shipping hardware. You can access their machines on the cloud

481
00:22:47,599 --> 00:22:51,440
right now. The key for them is connectivity and activity.

482
00:22:51,519 --> 00:22:54,799
In a chip like Google's equibit can usually only talk

483
00:22:54,839 --> 00:22:57,200
to its immediate neighbors. It's like a game of telephone.

484
00:22:57,559 --> 00:22:59,200
To get a message from one side to the other

485
00:22:59,240 --> 00:23:01,680
takes a lot of stuf. In an ion trap, you

486
00:23:01,720 --> 00:23:03,799
can wiggle the whole line of atoms so that any

487
00:23:03,880 --> 00:23:06,920
quibit can talk directly to any other quibit. It's like

488
00:23:06,960 --> 00:23:09,160
a conference call versus a game of telephone.

489
00:23:09,200 --> 00:23:12,319
Speaker 1: That sounds much more efficient for running software the actual

490
00:23:12,319 --> 00:23:13,519
algorithms it is.

491
00:23:13,839 --> 00:23:18,119
Speaker 2: It dramatically simplifies the algorithms, but they struggle with scaling up.

492
00:23:18,440 --> 00:23:22,119
It's hard to trap and control millions of individual atoms.

493
00:23:22,400 --> 00:23:25,240
So they're great on quality but behind on quantity.

494
00:23:25,440 --> 00:23:29,200
Speaker 1: So you have three competing philosophies. Google's scale at all

495
00:23:29,279 --> 00:23:34,000
costs freezer farm, Microsoft's high risk, high reward magic chip,

496
00:23:34,559 --> 00:23:38,240
and ion Que's slow and steady quality first approach.

497
00:23:38,599 --> 00:23:41,319
Speaker 2: That's a great summary. And if Google's cold approach hits

498
00:23:41,319 --> 00:23:45,000
an energy wall and Microsoft's magic not physics doesn't pan out,

499
00:23:45,039 --> 00:23:47,480
ion q is sitting there with a system that just works.

500
00:23:47,559 --> 00:23:50,039
Speaker 1: It's the hedging strategy for the market.

501
00:23:49,799 --> 00:23:52,200
Speaker 2: It is, and for investors that's the key. This isn't

502
00:23:52,200 --> 00:23:53,039
a one horse race.

503
00:23:53,200 --> 00:23:56,359
Speaker 1: Let's talk about that. The investment angle. The source. Doctor

504
00:23:56,440 --> 00:24:00,079
McCoy calls the twenty twenties the investment opportunity of a

505
00:23:59,880 --> 00:24:03,480
lifetime for this convergence. That's a bold claim.

506
00:24:03,640 --> 00:24:05,440
Speaker 2: It's hard to argue with that, we are looking at

507
00:24:05,480 --> 00:24:07,759
a fundamental platform shift. This is like the invention of

508
00:24:07,759 --> 00:24:11,000
the Internet or the microchip. It changes everything. You have AI,

509
00:24:11,279 --> 00:24:14,240
which is a world changing application that needs quantum, which

510
00:24:14,279 --> 00:24:16,880
is a world changing platform. The valuation of the winner

511
00:24:16,880 --> 00:24:18,400
here isn't going to be billions. It's going to be

512
00:24:18,400 --> 00:24:19,160
in the trillions.

513
00:24:19,519 --> 00:24:22,880
Speaker 1: It's a winner take all or winner take most dynamic.

514
00:24:22,519 --> 00:24:25,400
Speaker 2: Isn't it. It usually is in tech infrastructure. Look at

515
00:24:25,400 --> 00:24:30,200
cloud computing. You've got Aws, Azure, Google Cloud. That's basically it.

516
00:24:30,880 --> 00:24:35,200
The company is investing early, Google, Microsoft, IONQ. They will

517
00:24:35,200 --> 00:24:37,279
either be worth ten x their current.

518
00:24:37,079 --> 00:24:39,359
Speaker 1: Value or they get acquired by the one who wins. Yeah,

519
00:24:39,359 --> 00:24:41,359
there's no prize for second place precisely.

520
00:24:41,759 --> 00:24:44,400
Speaker 2: But here is my advice, and it echoes the report.

521
00:24:44,960 --> 00:24:48,480
Don't just look for the fastest computer. Everyone chases speed.

522
00:24:48,599 --> 00:24:49,759
That's the sexy.

523
00:24:49,480 --> 00:24:51,440
Speaker 1: Metric, right, how many operations per second?

524
00:24:51,559 --> 00:24:53,759
Speaker 2: Yeah, but that's a trap. Look for the most efficient one.

525
00:24:53,920 --> 00:24:56,960
Speaker 1: Because of the energy wall. We're coming full circle.

526
00:24:56,759 --> 00:24:59,279
Speaker 2: Because of the energy wall. The winner is the one

527
00:24:59,319 --> 00:25:02,240
who solves the energy problem. If your computer is a

528
00:25:02,279 --> 00:25:05,400
million times faster but requires the power output of Chicago

529
00:25:05,440 --> 00:25:08,279
to run, nobody can afford to turn it on, the

530
00:25:08,319 --> 00:25:11,559
economics of energy will dictate the winner of the quantum race,

531
00:25:12,079 --> 00:25:13,200
not raw speed.

532
00:25:13,440 --> 00:25:15,880
Speaker 1: That is a crucial distinction. It's not a drag race.

533
00:25:16,200 --> 00:25:19,519
It's an efficiency race. It's la masse, not a quarter mile.

534
00:25:19,880 --> 00:25:22,799
Speaker 2: Correct, and the company that builds the most fuel efficient

535
00:25:22,839 --> 00:25:24,799
engine is the one that will win the twenty four

536
00:25:24,799 --> 00:25:25,359
hour race.

537
00:25:25,720 --> 00:25:28,359
Speaker 1: So what does this all mean for us, the listeners,

538
00:25:29,000 --> 00:25:32,880
the people not buying billion dollars server farms or trading

539
00:25:32,920 --> 00:25:36,640
stock options. We've got this hardware war happening, But what

540
00:25:36,720 --> 00:25:38,960
happens when the software starts running on these things?

541
00:25:39,200 --> 00:25:41,440
Speaker 2: This is where we move beyond speed and talk about

542
00:25:41,440 --> 00:25:42,599
solving the impossible.

543
00:25:42,720 --> 00:25:44,319
Speaker 1: The impossible. I like this ount of that.

544
00:25:44,480 --> 00:25:48,319
Speaker 2: I mean it literally. Quantum isn't for better graphics on

545
00:25:48,359 --> 00:25:51,160
your video games or faster emails. Our phones are already

546
00:25:51,200 --> 00:25:54,759
great at that. It's for problems that are literally impossible

547
00:25:54,839 --> 00:25:57,799
for classical computers to solve because the number of variables

548
00:25:57,799 --> 00:26:02,319
explodes exponentially, the combinatorial explosion exactly. Give me an example

549
00:26:02,359 --> 00:26:02,680
from the.

550
00:26:02,640 --> 00:26:06,480
Speaker 1: Source, let's see it mentions energy systems optimizing wind farms.

551
00:26:07,079 --> 00:26:11,599
Speaker 2: Perfect example, think about a huge offshore wind farm hundreds

552
00:26:11,599 --> 00:26:12,319
of turbines.

553
00:26:12,359 --> 00:26:14,720
Speaker 1: Okay, I see wind turbines all the time. They spin,

554
00:26:14,960 --> 00:26:17,240
they make power. What's the problem. It seems pretty simple.

555
00:26:17,279 --> 00:26:20,839
Speaker 2: The problem is fluid dynamics. The wind is chaotic. If

556
00:26:20,920 --> 00:26:23,880
the front row of turbines faces the wind directly, they

557
00:26:23,880 --> 00:26:26,880
extract energy, but they also create a wake, like the

558
00:26:26,920 --> 00:26:30,039
wake behind a boat. It's turbulent, messy.

559
00:26:29,720 --> 00:26:33,319
Speaker 1: Air, right, and that messy air hits the turbines.

560
00:26:32,960 --> 00:26:35,599
Speaker 2: In the back row, and it kills their efficiency. The

561
00:26:35,640 --> 00:26:38,319
back row is drafting like a Nascar But in this case,

562
00:26:38,359 --> 00:26:39,960
that's bad for generating power.

563
00:26:40,039 --> 00:26:42,559
Speaker 1: So what's the solution. Just space them farther apart.

564
00:26:43,119 --> 00:26:45,559
Speaker 2: You could, but that costs a fortune in real estate

565
00:26:45,599 --> 00:26:49,519
and undersea cables. The optimal solution is to angle every

566
00:26:49,559 --> 00:26:52,240
blade on every turbine slightly differently in.

567
00:26:52,200 --> 00:26:53,799
Speaker 1: Real time to steer the wake.

568
00:26:53,759 --> 00:26:56,599
Speaker 2: To steer the wake around the other turbines, But the

569
00:26:56,640 --> 00:27:01,480
wind changes every second, the direction, the speed. Calculating the

570
00:27:01,519 --> 00:27:05,720
optimal angle for hundreds of turbines thousands of blades in

571
00:27:05,839 --> 00:27:09,880
real time that involves so many variables that a classical

572
00:27:09,920 --> 00:27:12,880
supercomputer just chokes on it. It can't keep up.

573
00:27:12,839 --> 00:27:15,279
Speaker 1: Too many possibilities to check an impossible number.

574
00:27:15,799 --> 00:27:18,440
Speaker 2: But a quantum computer can model that chaos. It can

575
00:27:18,480 --> 00:27:20,680
look at all the possibilities at once and find the

576
00:27:20,720 --> 00:27:25,519
global optimum, the perfect setting, instantly. The source estimates this

577
00:27:25,559 --> 00:27:28,240
could boost the efficiency of a wind farm by twenty

578
00:27:28,279 --> 00:27:28,960
percent or more.

579
00:27:29,039 --> 00:27:31,960
Speaker 1: Twenty percent. That's massive. That's like building a whole new

580
00:27:31,960 --> 00:27:33,160
power plant for free.

581
00:27:33,319 --> 00:27:37,359
Speaker 2: It is. And here's the beautiful irony. The energy needed

582
00:27:37,400 --> 00:27:41,400
to run that quantum optimization algorithm is tiny compared to

583
00:27:41,440 --> 00:27:44,680
the gigawatts of extra renewable energy you gain from the

584
00:27:44,720 --> 00:27:45,160
wind farm.

585
00:27:45,240 --> 00:27:47,240
Speaker 1: So you spanned a little bit of quantum energy to

586
00:27:47,279 --> 00:27:49,480
save massive amounts of real world energy.

587
00:27:49,599 --> 00:27:53,319
Speaker 2: Exactly. It's leverage, a tiny push on a very big level.

588
00:27:53,359 --> 00:27:55,119
Speaker 1: What else where else can we apply that lever?

589
00:27:55,440 --> 00:27:58,799
Speaker 2: Material science? This is the big one for me. This

590
00:27:58,839 --> 00:28:01,720
is where we stop discussing materials by accident like we

591
00:28:01,759 --> 00:28:04,400
have for most of human history, and start designing them

592
00:28:04,480 --> 00:28:05,359
atom by atom.

593
00:28:05,480 --> 00:28:07,279
Speaker 1: Better batteries, that's the one everyone talks about.

594
00:28:07,359 --> 00:28:11,000
Speaker 2: Better batteries, absolutely lighter, more powerful, no rare earth metals.

595
00:28:11,319 --> 00:28:12,839
But let's go deeper. Fertilizer.

596
00:28:12,880 --> 00:28:16,920
Speaker 1: Fertilizer, You're going from quantum computing to farming. That feels

597
00:28:16,920 --> 00:28:17,640
like a leap.

598
00:28:17,559 --> 00:28:20,359
Speaker 2: Stick with me. The way we make ammonia for fertilizer

599
00:28:20,440 --> 00:28:24,279
right now is a process called haberbosh. It involves incredibly

600
00:28:24,359 --> 00:28:28,160
high heat and high pressure to rip stable nitrogen molecules

601
00:28:28,200 --> 00:28:30,359
out of the air. It's brute force.

602
00:28:30,200 --> 00:28:32,359
Speaker 1: Chemistry, and it must use a lot of energy.

603
00:28:32,440 --> 00:28:35,759
Speaker 2: It consumes about one percent two percent of the world's

604
00:28:35,880 --> 00:28:41,079
entire energy supply one process, one single industrial process. It's

605
00:28:41,119 --> 00:28:43,920
one of the largest single sources of carbon emissions on

606
00:28:43,960 --> 00:28:44,599
the planet.

607
00:28:44,799 --> 00:28:47,000
Speaker 1: That is staggering. I had no idea.

608
00:28:47,079 --> 00:28:50,720
Speaker 2: Now consider this. Bacteria in the soil do this every

609
00:28:50,720 --> 00:28:53,599
single day at room temperature and normal pressure, using an

610
00:28:53,759 --> 00:28:55,319
enzyme called nitrogenase.

611
00:28:55,559 --> 00:28:57,559
Speaker 1: So nature has better tech than we do.

612
00:28:57,759 --> 00:29:01,759
Speaker 2: Much better, infinitely more elegant. But we can't replicate it

613
00:29:01,799 --> 00:29:06,200
because we can't fully simulate exactly how that complex enzyme works.

614
00:29:06,440 --> 00:29:10,799
The electron interactions are fundamentally quantum mechanical.

615
00:29:10,440 --> 00:29:13,640
Speaker 1: And classical computers can't do quantum mechanics. It seems like

616
00:29:13,680 --> 00:29:14,440
they should be able to.

617
00:29:14,680 --> 00:29:17,880
Speaker 2: The physicist Richard Feinman said it best decades ago. He said,

618
00:29:18,160 --> 00:29:20,519
nature isn't classical, damn it. And if you want to

619
00:29:20,519 --> 00:29:22,680
make a simulation of nature, you'd better make it.

620
00:29:22,759 --> 00:29:27,559
Speaker 1: Quantum mechanical nature runs on quantum physics, electrons, chemical bonds,

621
00:29:27,839 --> 00:29:28,880
molecular orbitals.

622
00:29:29,039 --> 00:29:32,640
Speaker 2: Right, Trying to simulate a complex molecule on a silicon

623
00:29:32,720 --> 00:29:36,079
chip is like trying to describe a Rembrandt painting using

624
00:29:36,119 --> 00:29:38,599
only the words yes and no. You just can't capture

625
00:29:38,599 --> 00:29:41,559
the nuance. A quantum computer, on the other hand, is

626
00:29:41,599 --> 00:29:44,519
the quantum system. It models the problem faithfully.

627
00:29:44,799 --> 00:29:48,680
Speaker 1: So if we unlock that, we could produce fertilizer with

628
00:29:48,759 --> 00:29:51,440
a tiny fraction of the energy cost. We could design

629
00:29:51,559 --> 00:29:53,920
room temperature superconductors.

630
00:29:53,200 --> 00:29:56,119
Speaker 2: All on the table. If we solve the catalyst for

631
00:29:56,240 --> 00:30:00,000
nitrogen fixation, we could cut global carbon emissions by a huge,

632
00:30:00,039 --> 00:30:03,440
huge amount, almost overnight. That is the scale of application

633
00:30:03,559 --> 00:30:06,799
we are talking about. It's civilization altering.

634
00:30:06,480 --> 00:30:09,599
Speaker 1: And this can expect to AI. The report uses the

635
00:30:09,640 --> 00:30:12,200
phrase quantum inspired optimization.

636
00:30:12,640 --> 00:30:15,240
Speaker 2: That sounds like buzzword salad, doesn't it like something from

637
00:30:15,279 --> 00:30:16,119
a marketing burcher.

638
00:30:16,279 --> 00:30:17,960
Speaker 1: It really does. What does it actually mean?

639
00:30:18,119 --> 00:30:21,440
Speaker 2: But it's real. It means using these quantum principles to

640
00:30:21,559 --> 00:30:25,480
train our AI models more efficiently. Right now, training AI

641
00:30:25,640 --> 00:30:28,880
is a brute force search through a massive possibility space

642
00:30:28,920 --> 00:30:30,319
to find the right setting.

643
00:30:30,119 --> 00:30:32,440
Speaker 1: Right, turning all the little knobs until it works.

644
00:30:32,279 --> 00:30:34,759
Speaker 2: And it costs hundreds of millions of dollars in electricity.

645
00:30:35,200 --> 00:30:38,279
Quantum optimization algorithms could find the best path through that

646
00:30:38,359 --> 00:30:41,599
space much more efficiently. The goal is to make AI

647
00:30:41,680 --> 00:30:43,640
deliver better performance with fewer.

648
00:30:43,359 --> 00:30:45,440
Speaker 1: Operations, less energy per task.

649
00:30:45,519 --> 00:30:49,039
Speaker 2: Making AI sustainable. That is the holy grail. Instead of

650
00:30:49,039 --> 00:30:52,480
brute forcing intelligence with more and more power. We find

651
00:30:52,480 --> 00:30:55,720
the elegant path. We make it smarter, not just bigger.

652
00:30:56,240 --> 00:30:58,680
Speaker 1: So we're at the convergence point. We have the hardware

653
00:30:58,799 --> 00:31:02,400
race Google versus Microsoft versus ion Q. We have the

654
00:31:02,440 --> 00:31:07,880
application explosion batteries, energy grids, fertilizer AI itself. Now, let's

655
00:31:07,880 --> 00:31:10,440
look at the calendar again. That three year window twenty

656
00:31:10,480 --> 00:31:11,960
twenty six to twenty twenty.

657
00:31:11,680 --> 00:31:12,880
Speaker 2: Eight, the phase shift window.

658
00:31:13,000 --> 00:31:15,920
Speaker 1: The source highlighted a really critical statistic, almost a warning

659
00:31:16,279 --> 00:31:18,079
about the power demand discrepancy.

660
00:31:18,160 --> 00:31:21,119
Speaker 2: Yes, this is why governments and national labs are sweating

661
00:31:21,160 --> 00:31:23,880
right now. The electrical power demand for a full scale,

662
00:31:24,039 --> 00:31:27,079
fault tolerant quantum computer could differ by up to two

663
00:31:27,319 --> 00:31:28,359
orders of magnitude.

664
00:31:28,359 --> 00:31:30,680
Speaker 1: Two orders of magnitude, so one hundred times.

665
00:31:30,480 --> 00:31:33,559
Speaker 2: One hundred times different depending on which architecture wins.

666
00:31:33,599 --> 00:31:36,079
Speaker 1: So if we bet on the wrong horse, if a

667
00:31:36,119 --> 00:31:40,720
country builds its national quantum strategy around one technology, if we.

668
00:31:40,680 --> 00:31:45,039
Speaker 2: Build our national infrastructure around a technology that requires one

669
00:31:45,079 --> 00:31:47,960
hundred x more power, say we go all in on

670
00:31:48,000 --> 00:31:52,039
the giant fridges, and then Microsoft's low power topological chip

671
00:31:52,400 --> 00:31:56,640
actually works. We've wasted billions and locked ourselves into an

672
00:31:56,720 --> 00:31:58,759
unsustainable grid for decades.

673
00:31:58,880 --> 00:32:01,160
Speaker 1: You'd have the quantum equivalent of a country full of

674
00:32:01,200 --> 00:32:03,480
Betamax players in a VHS.

675
00:32:02,960 --> 00:32:06,640
Speaker 2: World exactly, but a Betamax player that requires its own

676
00:32:06,759 --> 00:32:10,240
power plant to run. Governments and industries have to decide

677
00:32:10,240 --> 00:32:12,680
now which path to back, and it's a huge gamble.

678
00:32:12,720 --> 00:32:15,079
Speaker 1: It's high stakes poker with the national grid. Is the

679
00:32:15,079 --> 00:32:15,440
buy in?

680
00:32:15,720 --> 00:32:18,240
Speaker 2: It is And by twenty twenty eight, the prediction in

681
00:32:18,279 --> 00:32:20,799
the report is that we will see the first commercially

682
00:32:20,880 --> 00:32:24,880
viable large scale quantum computers solving real world problems, not

683
00:32:24,960 --> 00:32:28,039
just lab demas real valuable commercial problem.

684
00:32:28,160 --> 00:32:30,720
Speaker 1: That's just around the corner. That's not science fiction anymore.

685
00:32:30,759 --> 00:32:31,559
Speaker 2: It's a business plan.

686
00:32:31,799 --> 00:32:34,839
Speaker 1: And the source mentions Google's vision of personal intelligence. What

687
00:32:35,000 --> 00:32:35,319
is that.

688
00:32:35,480 --> 00:32:38,599
Speaker 2: It's the next step beyond large language models. It's an

689
00:32:38,640 --> 00:32:40,799
AI that knows you better than you know yourself.

690
00:32:40,920 --> 00:32:43,160
Speaker 1: That's both terrifying and incredibly exciting.

691
00:32:43,240 --> 00:32:46,160
Speaker 2: Yes, it is. Imagine an AI that has a perfect

692
00:32:46,200 --> 00:32:49,559
memory of every conversation you've ever had, every document you've

693
00:32:49,559 --> 00:32:53,720
ever read. It understands your context, anticipates your health needs

694
00:32:53,720 --> 00:32:57,480
based on your biometrics, manages your finances with perfect foresight,

695
00:32:57,680 --> 00:32:57,960
like a.

696
00:32:57,920 --> 00:33:00,559
Speaker 1: True digital assistant. Not the dumbsby we have.

697
00:33:00,559 --> 00:33:04,400
Speaker 2: Now exactly, but that level of complexity, that constant personalized

698
00:33:04,400 --> 00:33:08,119
computation is impossible to run cheaply on today's hardware. It

699
00:33:08,119 --> 00:33:10,000
would be too slow and too expensive.

700
00:33:10,319 --> 00:33:13,480
Speaker 1: So the quantum breakthrough is the prerequisite for that AI breakthrough.

701
00:33:13,559 --> 00:33:17,799
Speaker 2: Precisely, the convergence of AI and quantum is the phase shift.

702
00:33:17,960 --> 00:33:20,920
It's the moment we stop struggling against complexity and start

703
00:33:20,960 --> 00:33:21,720
mastering it.

704
00:33:21,720 --> 00:33:23,480
Speaker 1: It's a lot to take in. The scale of this

705
00:33:23,799 --> 00:33:25,160
is immense.

706
00:33:24,920 --> 00:33:26,319
Speaker 2: It is. It touches everything.

707
00:33:26,519 --> 00:33:29,599
Speaker 1: So let's try to synthesize this. We've covered the energy crisis,

708
00:33:29,759 --> 00:33:33,160
the physics of why the corporate players, the global stakes.

709
00:33:33,640 --> 00:33:36,119
When you step back from all the details, what's the

710
00:33:36,160 --> 00:33:38,079
one thing that keeps you up at night about this?

711
00:33:39,119 --> 00:33:42,119
Speaker 2: You know, usually with these topics. It's the risks, the

712
00:33:42,240 --> 00:33:46,519
terminator scenarios, the AI safety questions. But here what keeps

713
00:33:46,559 --> 00:33:49,799
me up is the sheer magnitude of the shift in logic.

714
00:33:50,319 --> 00:33:53,720
How so, we are building computers that work on fundamentally

715
00:33:53,720 --> 00:33:57,039
different principles of physics than anything we have ever built.

716
00:33:57,519 --> 00:34:00,880
We are stepping away from the deterministic, binary world of

717
00:34:01,000 --> 00:34:03,759
zeros in one black and white yes and no, and

718
00:34:03,799 --> 00:34:08,679
we are entering the probabilistic, complex, entangled world of nature itself.

719
00:34:08,840 --> 00:34:13,000
Speaker 1: It feels profound, it's philosophical almost. We're building a computer

720
00:34:13,079 --> 00:34:14,679
that thinks more like the universe does.

721
00:34:14,840 --> 00:34:17,559
Speaker 2: It is, but on a very practical level, it's the

722
00:34:17,599 --> 00:34:20,440
talent gap. The shortage of people who understand this stuff

723
00:34:20,480 --> 00:34:24,519
is severe. We have maybe a few thousand real experts

724
00:34:24,519 --> 00:34:26,719
in the world. We are trying to build the engine

725
00:34:26,719 --> 00:34:28,639
of the twenty first century, and we don't have enough

726
00:34:28,679 --> 00:34:30,079
mechanics who speak the language.

727
00:34:30,159 --> 00:34:32,760
Speaker 1: That's a real bottleneck. The tech could be ready, but

728
00:34:32,800 --> 00:34:34,239
we won't have the people to use it.

729
00:34:33,920 --> 00:34:36,840
Speaker 2: It is, But my optimism comes from that idea of

730
00:34:36,880 --> 00:34:39,400
reversibility we talked about the beginning, the idea that we

731
00:34:39,440 --> 00:34:43,079
can compute without destruction, that we can be efficient. That

732
00:34:43,079 --> 00:34:45,159
feels like a mature step for civilization.

733
00:34:45,639 --> 00:34:47,960
Speaker 1: And my takeaway is that the race is on between

734
00:34:48,000 --> 00:34:51,519
the tech giants, but the real winner, as you said,

735
00:34:52,239 --> 00:34:55,920
is whoever solves the energy efficiency puzzle. It's not about

736
00:34:55,960 --> 00:34:58,840
who gets to the finish line first. It's about who

737
00:34:58,880 --> 00:35:01,880
has enough gas left in the tank to keep going beautifully.

738
00:35:01,880 --> 00:35:04,039
Speaker 2: Put that's the whole story in a nutshell.

739
00:35:04,679 --> 00:35:07,719
Speaker 1: So let's leave our listeners with some actionable advice. We

740
00:35:07,760 --> 00:35:09,320
don't just want to blow their minds. We want to

741
00:35:09,320 --> 00:35:11,159
give them a toolkit for this new world.

742
00:35:11,440 --> 00:35:13,280
Speaker 2: Right, what do you do on Monday morning?

743
00:35:13,559 --> 00:35:15,960
Speaker 1: For the investors listening, what's the play?

744
00:35:16,239 --> 00:35:19,719
Speaker 2: Watch who is solving the energy problem? Don't get distracted

745
00:35:19,719 --> 00:35:22,719
by press releases about quibet counts alone. If a company

746
00:35:22,719 --> 00:35:25,559
says we have ten thousand kibbuts ask how much power

747
00:35:25,559 --> 00:35:28,199
does it take to run them? Look at the cooling requirements,

748
00:35:28,239 --> 00:35:31,360
look at the total system cost of ownership. The efficiency

749
00:35:31,440 --> 00:35:32,360
is the key metric.

750
00:35:32,480 --> 00:35:35,760
Speaker 1: Okay. For the tech workers, the coders, the engineers.

751
00:35:35,679 --> 00:35:40,280
Speaker 2: Start learning quantum principles immediately. You don't need a PhD

752
00:35:40,400 --> 00:35:45,079
in physics, but you need to understand the logic, understand

753
00:35:45,119 --> 00:35:49,039
the new types of algorithms. Learn the basic languages like

754
00:35:49,119 --> 00:35:52,920
qushtab from Microsoft or a circ from Google. The talent

755
00:35:53,000 --> 00:35:56,119
gap is a massive opportunity. If you can bridge the

756
00:35:56,119 --> 00:35:59,440
world to classical coding with quantum concepts, you are writing

757
00:35:59,480 --> 00:36:01,440
your own tip get for the next twenty years.

758
00:36:01,639 --> 00:36:03,599
Speaker 1: Be the translator between the two worlds.

759
00:36:03,679 --> 00:36:05,679
Speaker 2: Be the translator. You'll be invaluable.

760
00:36:05,880 --> 00:36:10,679
Speaker 1: And finally, for the business leaders, the CEOs, the strategists.

761
00:36:10,440 --> 00:36:12,800
Speaker 2: Look at your own industry and identify the problems that

762
00:36:12,840 --> 00:36:16,559
are currently too complex. The ones your supercomputers choke on,

763
00:36:17,000 --> 00:36:19,760
the supply chain logistics that are always messy, the drug

764
00:36:19,800 --> 00:36:23,400
discovery simulations that take years, the material designs that are

765
00:36:23,440 --> 00:36:24,880
too expensive to test in the.

766
00:36:24,840 --> 00:36:26,599
Speaker 1: Real world, the impossible problems.

767
00:36:26,679 --> 00:36:30,320
Speaker 2: Those are your quantum opportunities. Start cataloging them now. Be

768
00:36:30,400 --> 00:36:32,800
ready to deploy when the tech lands in twenty twenty eight,

769
00:36:32,880 --> 00:36:34,320
because your competitors will be.

770
00:36:34,519 --> 00:36:37,239
Speaker 1: Twenty twenty eight. Mark your calendars. Everyone, The age of

771
00:36:37,320 --> 00:36:40,320
quantum abundance is coming. It's coming fast, and that brings

772
00:36:40,400 --> 00:36:42,000
us to the end of this thread. But we really

773
00:36:42,079 --> 00:36:45,199
want to hear from you, the listener. We've painted a

774
00:36:45,239 --> 00:36:48,360
picture of a future where AI could consume the power

775
00:36:48,360 --> 00:36:51,599
of nations, but this new form of computing might just

776
00:36:51,639 --> 00:36:52,280
save it.

777
00:36:52,280 --> 00:36:56,199
Speaker 2: It's the ultimate ds X machina, a solution appearing from

778
00:36:56,239 --> 00:36:57,599
the very fabric.

779
00:36:57,159 --> 00:37:00,639
Speaker 1: Of physics exactly. So here is the direct question for you.

780
00:37:01,199 --> 00:37:04,559
We're facing a future where AI consumes the power of nations,

781
00:37:04,840 --> 00:37:09,000
but quantum might save it. Do you trust big tech

782
00:37:09,519 --> 00:37:13,360
to prioritize the energy efficient solution or will the race

783
00:37:13,440 --> 00:37:15,000
for raw speed win out.

784
00:37:15,400 --> 00:37:18,159
Speaker 2: Will they choose the Ferrari engine that melts the chassis

785
00:37:18,199 --> 00:37:20,440
because it looks good in a demo or the efficient

786
00:37:20,480 --> 00:37:22,719
engine that saves the planet but is maybe a little

787
00:37:22,840 --> 00:37:23,679
less flashy.

788
00:37:23,840 --> 00:37:26,440
Speaker 1: It's a question of profit versus prudence. Leave a comment

789
00:37:26,480 --> 00:37:28,519
and let us know your stand We really want to

790
00:37:28,559 --> 00:37:29,400
know what you think about this.

791
00:37:29,599 --> 00:37:32,360
Speaker 2: We do this conversation needs to happen out in the open.

792
00:37:32,599 --> 00:37:34,239
Speaker 1: Thanks for joining us on thrilling Threads.

793
00:37:34,360 --> 00:37:34,719
Speaker 2: Thank you.

794
00:37:34,960 --> 00:37:35,960
Speaker 1: Welcome to the future.

795
00:37:36,079 --> 00:37:37,920
Speaker 2: It's already here, you just have to know where to look.