1
00:00:00,000 --> 00:00:05,792
Linus Tech Tips asked who on
earth buys a $6,000 AI laptop?

2
00:00:06,078 --> 00:00:09,260
And meanwhile, in my house, I
have a machine with two graphics

3
00:00:09,261 --> 00:00:10,847
cards that never turns off.

4
00:00:11,522 --> 00:00:16,688
AI cloud keeps getting cheaper, and AI
hardware keeps getting more expensive,

5
00:00:16,807 --> 00:00:19,181
and somehow the workstation is back.

6
00:00:19,524 --> 00:00:22,558
Chauncey's off this week Scott
Slater is joining us from New

7
00:00:22,558 --> 00:00:24,829
York, and this is the Local Host.

8
00:00:42,576 --> 00:00:44,397
Frankcx: Welcome everyone.

9
00:00:44,586 --> 00:00:45,503
We are back.

10
00:00:45,525 --> 00:00:49,291
We took last week off, and if you
notice from our five pictures,

11
00:00:49,292 --> 00:00:50,523
we've got a little bit different.

12
00:00:50,523 --> 00:00:51,654
It's kinda like Sesame Street.

13
00:00:51,654 --> 00:00:55,579
One of these faces is not, the
same, but we've, replaced Scott or,

14
00:00:55,579 --> 00:00:57,481
Chauncey Larson with Scott Slater.

15
00:00:57,781 --> 00:00:58,429
Robert: Permanently?

16
00:00:58,722 --> 00:00:59,647
Frankcx: welcome back.

17
00:01:00,004 --> 00:01:00,756
Robert: permanently

18
00:01:00,837 --> 00:01:03,306
Frankcx: hey, let's first do
a little bit around the horn.

19
00:01:03,323 --> 00:01:06,506
I think, Maybe almost in the round
the horn it would be like, where

20
00:01:06,514 --> 00:01:08,358
have you been in the last week?

21
00:01:08,673 --> 00:01:09,944
I have not been anywhere.

22
00:01:09,945 --> 00:01:13,547
I've been, sitting here watching
sadly, the Seattle Seahawks lose.

23
00:01:13,557 --> 00:01:16,696
Sam Darnold better figure out how
to stop throwing interceptions,

24
00:01:17,153 --> 00:01:18,910
but I think he feels good.

25
00:01:18,927 --> 00:01:22,050
Like, it feels like an easy game
this week going up against the 6

26
00:01:22,050 --> 00:01:24,110
and 0 Kansas City Chiefs in Seattle.

27
00:01:24,457 --> 00:01:28,300
but Canine is back into Seattle,
so we'll see how that game goes.

28
00:01:28,313 --> 00:01:29,831
but hey, let's go around the horn.

29
00:01:29,840 --> 00:01:33,080
Neil, I think you've been traveling,
so maybe, give an introduction

30
00:01:33,095 --> 00:01:34,383
and tell us where you've been.

31
00:01:35,322 --> 00:01:35,595
Neil: Ray.

32
00:01:35,613 --> 00:01:37,011
Great to see everyone again.

33
00:01:37,017 --> 00:01:41,841
Neil Mysak, disclosing that I'm a
member of Microsoft Surface Engineering.

34
00:01:42,180 --> 00:01:47,393
The last week, I went to Chicago for
a Surface Laptop Ultra launch event.

35
00:01:47,425 --> 00:01:48,469
great success.

36
00:01:48,532 --> 00:01:51,561
Showed local coding, live on the device.

37
00:01:51,809 --> 00:01:57,125
on my way from Chicago to Arizona,
stress tested the power of Laptop

38
00:01:57,133 --> 00:02:00,836
Ultra, actually built an offline
PowerPoint generator, and then shared

39
00:02:00,836 --> 00:02:04,068
that at the ASU Agentic AI Summit.

40
00:02:04,307 --> 00:02:07,563
will.i.am made a guest appearance, which
I'll share more on throughout the show.

41
00:02:07,831 --> 00:02:08,106
but

42
00:02:08,401 --> 00:02:09,782
Robert: Wow

43
00:02:09,923 --> 00:02:12,803
Neil: Chicago, and
Phoenix in the last week.

44
00:02:12,911 --> 00:02:13,635
Frankcx: Welcome back

45
00:02:13,735 --> 00:02:14,627
Jacob: How do we keep up?

46
00:02:14,769 --> 00:02:15,952
Frankcx: Yeah, it's good to have you.

47
00:02:16,677 --> 00:02:17,168
Hey, Robert,

48
00:02:17,168 --> 00:02:18,703
Robert: I'm in the middle
of… You can see though now

49
00:02:18,704 --> 00:02:22,079
my office is put back together,
different pink color on the wall.

50
00:02:22,081 --> 00:02:23,137
We're in the middle of construction.

51
00:02:23,138 --> 00:02:26,243
You might hear, a patio stone
cutter in the background, but,

52
00:02:26,279 --> 00:02:27,802
hopefully AI Will keep that out.

53
00:02:28,157 --> 00:02:30,960
And as soon as we're done here, I
won't dox myself 'cause this will be--

54
00:02:30,970 --> 00:02:34,361
I'll be home by the time this airs,
but I'm on my way to Columbus, to go

55
00:02:34,365 --> 00:02:36,228
see our son play soccer this weekend.

56
00:02:36,229 --> 00:02:39,099
Maybe take in a Columbus Blue
Jackets game while we're there,

57
00:02:39,720 --> 00:02:41,110
and, looking forward to it.

58
00:02:41,152 --> 00:02:43,096
Frankcx: the Ohio State University?

59
00:02:43,211 --> 00:02:44,927
Robert: it, happens to be there, yes.

60
00:02:45,257 --> 00:02:46,703
I won't be there, but yes

61
00:02:48,236 --> 00:02:49,116
Frankcx: He's… Yeah, my

62
00:02:49,376 --> 00:02:51,673
Jacob: Will you tell us the exact
address you'll be at, Robert?

63
00:02:51,757 --> 00:02:56,355
Frankcx: he always wanted me to dot
the I because he was a Ohio native,

64
00:02:56,355 --> 00:02:57,759
but I never got the chance to do

65
00:02:57,760 --> 00:02:57,979
that.

66
00:02:57,996 --> 00:02:58,939
Let my dad down,

67
00:02:59,098 --> 00:02:59,834
Robert: sorry

68
00:03:00,497 --> 00:03:03,720
Frankcx: Speaking of
euphoniums and tubas, Jacob

69
00:03:05,750 --> 00:03:06,978
Jacob: How's it going, everyone?

70
00:03:07,245 --> 00:03:11,662
last week we didn't have the pod, but, I
wasn't on because I was in San Francisco

71
00:03:11,695 --> 00:03:17,718
at a Best Buy in downtown San Francisco
with, folks like co-founder of NVIDIA,

72
00:03:17,737 --> 00:03:23,185
folks like the founder of OpenClaw,
Peter, folks like the founder of Comfy

73
00:03:23,185 --> 00:03:26,841
UI, which we showed on stage, Microsoft
showed on stage a couple weeks ago.

74
00:03:27,358 --> 00:03:31,412
and about 400 people buying
their first RTX Spark devices.

75
00:03:32,914 --> 00:03:34,723
Yes, which was awesome

76
00:03:34,863 --> 00:03:36,206
Frankcx: other OEMs too?

77
00:03:37,463 --> 00:03:41,350
Jacob: this is a chance on this podcast
to say that I don't believe any of the

78
00:03:41,363 --> 00:03:47,690
purchases were Surface Laptop Ultras
because, that specific location did not

79
00:03:47,716 --> 00:03:50,537
assort Surface Laptop Ultras in person.

80
00:03:50,751 --> 00:03:51,747
Frankcx: transparency

81
00:03:51,780 --> 00:03:55,773
Jacob: sales online, tons of excitement
about the platform, but, I was the

82
00:03:55,784 --> 00:03:59,270
lone representative for Microsoft
there to, to help kinda connect

83
00:03:59,270 --> 00:04:00,923
the dots and, and be a part of it,

84
00:04:01,145 --> 00:04:03,565
Robert: keeping track, that
was Jacob's third low-key

85
00:04:03,565 --> 00:04:05,333
brag already on this podcast.

86
00:04:05,334 --> 00:04:05,791
Just…

87
00:04:06,884 --> 00:04:11,052
Frankcx: Well, Neil worked in Neil
I Am, or Neil I Am, Will.i.am.

88
00:04:11,752 --> 00:04:12,168
Neil: like that.

89
00:04:12,169 --> 00:04:12,454
My new

90
00:04:12,574 --> 00:04:14,056
Jacob: I feel I am

91
00:04:15,254 --> 00:04:16,260
Frankcx: You are a DJ.

92
00:04:16,346 --> 00:04:17,450
Is that your DJ name?

93
00:04:17,703 --> 00:04:19,234
Neil: no, I go with Neil Nebula.

94
00:04:19,288 --> 00:04:23,414
and I can share my, recent Redmond
debut in the Microsoft tree houses,

95
00:04:23,527 --> 00:04:24,279
Frankcx: Ooh,

96
00:04:24,484 --> 00:04:26,209
Neil: recently connected as well, so

97
00:04:26,330 --> 00:04:28,614
Frankcx: I will cut in the
video if you have some to share.

98
00:04:28,614 --> 00:04:31,853
I'll put it in the pod
right here, so awesome.

99
00:04:32,267 --> 00:04:34,275
All right, well, Chauncey
is out this week.

100
00:04:34,284 --> 00:04:37,799
I think he's down at a wedding or
some other event down in Mexico.

101
00:04:37,870 --> 00:04:41,002
But we are honored to
have Scott Slater here.

102
00:04:41,030 --> 00:04:44,194
Scott joins us from
New York, in the house.

103
00:04:44,216 --> 00:04:48,423
Scott's old school, Microsoft
Surface and Microsoft's, we call

104
00:04:48,423 --> 00:04:50,801
it an M2, like he's a sales leader.

105
00:04:51,095 --> 00:04:53,399
so Scott, I'll turn it over
to you to kinda introduce

106
00:04:53,400 --> 00:04:55,641
yourself and how are you doing?

107
00:04:55,979 --> 00:04:58,874
Scott Slater: Yeah, and I'll say
the honor's all mine to be with you

108
00:04:58,882 --> 00:05:01,337
fine gentlemen, and I wasn't sure
where you were starting when you

109
00:05:01,338 --> 00:05:03,092
started sharing I thought it was old.

110
00:05:03,198 --> 00:05:03,410
Frankcx: old

111
00:05:03,667 --> 00:05:05,372
Scott Slater: yeah,
pleasure to join you all.

112
00:05:05,372 --> 00:05:07,773
Looking forward to having
a fun conversation.

113
00:05:08,112 --> 00:05:11,950
I'm one of the leaders in Microsoft's
Windows specialist organization.

114
00:05:12,348 --> 00:05:17,203
no great stories this week, but I was
in and out of New York City meeting

115
00:05:17,206 --> 00:05:20,897
with different large customers,
which is generally how I spend my

116
00:05:20,897 --> 00:05:24,280
weeks, meeting from customer to
customer across different industries

117
00:05:25,974 --> 00:05:27,356
Frankcx: Scott, I'm gonna open it up.

118
00:05:27,358 --> 00:05:28,876
We're changing the format a little bit.

119
00:05:28,949 --> 00:05:33,116
I've got these top questions of the
week that I'd love to be able to ask,

120
00:05:33,116 --> 00:05:37,052
and as kind of the host week, my first
question is actually to Scott, because

121
00:05:37,052 --> 00:05:40,448
it sounds like making that trip into
New York City, meeting with those

122
00:05:40,457 --> 00:05:45,790
valued customers at Microsoft, what
are customers actually asking about

123
00:05:45,823 --> 00:05:49,278
in AI and specifically AI on device?

124
00:05:49,326 --> 00:05:52,972
I mean, sounds like you're on the
Windows side now, so Windows is obviously

125
00:05:52,975 --> 00:05:56,481
more than just AI on device, but,
coming from the Surface side, it'd be

126
00:05:56,482 --> 00:05:59,854
interesting to get your perspective as
to how customers are thinking about it.

127
00:06:00,352 --> 00:06:02,005
Scott Slater: Yeah, it's
a good question, Frank.

128
00:06:02,060 --> 00:06:06,397
I think, the conversation flips
at some point with customers.

129
00:06:06,398 --> 00:06:09,186
I think some customers start
talking about the hardware.

130
00:06:09,586 --> 00:06:12,925
They may ask about the chip, the silicone.

131
00:06:12,936 --> 00:06:15,064
They may ask what an NPU is.

132
00:06:15,076 --> 00:06:21,980
They may ask about memory and tops
and GPUs and, and after getting a

133
00:06:21,981 --> 00:06:26,269
little bit into that conversation,
it kinda changes to, you know, can

134
00:06:26,270 --> 00:06:31,387
I, can I use the devices I have today
or do I need to get new devices?

135
00:06:31,606 --> 00:06:33,771
And that's kinda where
the conversation flips.

136
00:06:34,086 --> 00:06:38,823
And then at some point the conversation
goes from there to, where is my data

137
00:06:38,861 --> 00:06:44,797
and is it secure as I'm running these
AI workloads on these new AI PCs?

138
00:06:45,165 --> 00:06:49,724
And then at some point it shifts from
there to like, does everybody need

139
00:06:49,725 --> 00:06:53,588
this in my environment or do only
some people need it in my environment?

140
00:06:54,044 --> 00:06:54,416
And so

141
00:06:54,417 --> 00:06:57,104
it's interesting how these
conversations are going with customers.

142
00:06:57,110 --> 00:07:01,980
But I usually I see it starting with
like understanding about today's new

143
00:07:01,989 --> 00:07:07,019
hardware and then it quickly shifts
to who needs this and can I do this on

144
00:07:07,019 --> 00:07:08,914
what I have or do I need something new?

145
00:07:09,140 --> 00:07:12,947
Frankcx: I find that an interesting
question because, as I was watching a lot

146
00:07:12,947 --> 00:07:18,479
of the, reviews of the RTX Spark, device,
and whether that's a Surface Laptop Ultra

147
00:07:18,487 --> 00:07:22,504
or the other OEMs that are selling those
devices, one of them that came to mind

148
00:07:22,504 --> 00:07:26,424
was Linus Tech Tips, a guy that's been
in the internet for quite some time.

149
00:07:26,424 --> 00:07:29,617
He's got quite a bit of experience
out there, sits up in Vancouver,

150
00:07:30,000 --> 00:07:31,456
watching over all of this.

151
00:07:31,461 --> 00:07:39,341
But his actual take on it was like, who on
earth would spend $6,000 for an AI laptop

152
00:07:39,406 --> 00:07:44,060
when you can through a tool like Tailscale
or another tool that allows you to

153
00:07:44,060 --> 00:07:47,542
tunnel into an AI device, into your home?

154
00:07:47,630 --> 00:07:51,289
And I'll, I'll say that firstfor- and
foremost, that's honestly probably

155
00:07:51,290 --> 00:07:53,652
the way that I use AI the most today.

156
00:07:54,112 --> 00:07:58,063
I have a device that's literally
right behind me, AI models.

157
00:07:58,091 --> 00:08:01,812
I'll talk a little bit about this,
Ride Edge app that I've developed,

158
00:08:02,115 --> 00:08:03,713
but, that's the way I use it.

159
00:08:03,715 --> 00:08:09,371
I use my phone, in some ways to Tailscale,
, a nice portal and a nice kind of tunnel

160
00:08:09,395 --> 00:08:13,382
application that cr- creates a secure
loca- secure tunnel into the device.

161
00:08:13,803 --> 00:08:17,214
So I would wonder, like, why
would people even buy a laptop?

162
00:08:17,220 --> 00:08:20,940
But Jacob, I know that you were
at the Best Buy, and you were

163
00:08:20,943 --> 00:08:24,597
watching people walk out the
door with a bunch of AI laptops.

164
00:08:24,614 --> 00:08:27,071
You know, what was the reason
that they were doing that?

165
00:08:28,437 --> 00:08:31,688
Jacob: Yeah, I mean, first of all,
I, I think, your perspective, Frank,

166
00:08:31,700 --> 00:08:34,336
is mostly because you're using
the hardware you have available.

167
00:08:34,599 --> 00:08:38,271
Almost to, to, to Scott's point
earlier, like customers are asking

168
00:08:38,271 --> 00:08:41,316
like, "Do I need to buy something new
or can I use what I already have?"

169
00:08:41,570 --> 00:08:46,418
Like, if you've got dual 3090s, like
you're already set up really well.

170
00:08:46,427 --> 00:08:49,792
Like that's gonna cost you… Like
that whole setup is probably 10

171
00:08:49,792 --> 00:08:52,827
grand on the open market these days
with the price of graphics cards.

172
00:08:53,308 --> 00:08:55,832
so I would say that yes, you're
kind of an edge case in that you're

173
00:08:55,832 --> 00:08:59,907
already set up well, and you don't
have one of these, from what I know,

174
00:08:59,998 --> 00:09:02,826
Laptop Ultras or RTX Spark devices yet

175
00:09:02,911 --> 00:09:03,589
Frankcx: one, but nobody

176
00:09:03,799 --> 00:09:05,023
Robert: Get in line

177
00:09:05,208 --> 00:09:05,382
Frankcx: as

178
00:09:05,508 --> 00:09:06,086
Jacob: Yeah

179
00:09:06,177 --> 00:09:08,988
Frankcx: influencer, now
I'm not as big as, Neil I am

180
00:09:13,017 --> 00:09:17,034
Jacob: but what I will say is that
customers are just looking for devices

181
00:09:17,034 --> 00:09:19,483
that match their needs in my perspective.

182
00:09:19,520 --> 00:09:22,430
And so the folks that walked out of
that Best Buy, they were excited.

183
00:09:22,658 --> 00:09:25,969
it was a little bit more on the
consumer side where folks are looking

184
00:09:25,969 --> 00:09:28,997
for their AI work, they're looking
for their gaming work, they're

185
00:09:28,999 --> 00:09:30,680
looking for their development work.

186
00:09:30,969 --> 00:09:33,925
They're-- Like some of them
were AI founders themselves.

187
00:09:33,925 --> 00:09:37,869
I got pitched on a number of different,
startups that week on like, "Hey, do you

188
00:09:37,869 --> 00:09:41,400
have access to the Microsoft socials?
Like you work in marketing. can you get

189
00:09:41,427 --> 00:09:44,498
me connected with the right person here
or there?" I'm like: Okay, I'd love to

190
00:09:44,499 --> 00:09:46,143
hear your idea, but let's chat about it.

191
00:09:46,197 --> 00:09:50,834
But I think that users just want a
device that they can turn on, that

192
00:09:50,835 --> 00:09:54,882
they can use, and by the way, a ton
of great announcements this week and

193
00:09:54,882 --> 00:09:59,661
the last couple weeks from Microsoft
and other partners on how it's easier

194
00:09:59,661 --> 00:10:03,889
than ever to just have local AI built
into the tools that you're using.

195
00:10:03,965 --> 00:10:04,694
Robert: an RTX

196
00:10:04,797 --> 00:10:07,118
Jacob: like GitHub Copilot
CLI, stuff like that.

197
00:10:07,167 --> 00:10:11,621
Robert: device, laptop as a server,
or you can take it with you.

198
00:10:11,654 --> 00:10:15,323
Now, look it, full disclosure, I
think we talked last podcast, do

199
00:10:15,324 --> 00:10:19,036
I spend six grand on a laptop or
do I spend six grand on a patio?

200
00:10:19,832 --> 00:10:22,826
I think by, if you hear the noise
You know, what my decision was.

201
00:10:23,540 --> 00:10:25,137
And no, it's not very mobile.

202
00:10:25,875 --> 00:10:29,414
but I… Look it, I love the idea
of going to Neal's Point, going on a

203
00:10:29,415 --> 00:10:34,507
plane, going offline, and just grooving
for four or five hours on a flight.

204
00:10:34,957 --> 00:10:37,341
That, dude, that to me is totally worth it

205
00:10:37,956 --> 00:10:39,151
Scott Slater: know, I, I'll

206
00:10:39,246 --> 00:10:41,510
Frankcx: Neil, I guess maybe
the… Oh, go ahead, Scott

207
00:10:41,587 --> 00:10:44,499
Scott Slater: Yeah, I was just gonna
say, I, I think I, I really agree

208
00:10:44,500 --> 00:10:46,270
with, with Jacob and Robert here.

209
00:10:46,339 --> 00:10:49,909
I don't think… If you're trying
to save a few bucks on tokens,

210
00:10:49,965 --> 00:10:51,349
I don- I think Linus is right.

211
00:10:51,349 --> 00:10:53,191
You're, you're not
buying the device, right?

212
00:10:53,557 --> 00:10:58,704
But, for, for builders, for creators,
they're, they're buying freedom, right?

213
00:10:58,738 --> 00:11:04,268
Like, look, when you're running on a
meter, your behavior changes, right?

214
00:11:04,381 --> 00:11:09,695
Like, if every try costs money,
you're gonna play it safe a little

215
00:11:09,696 --> 00:11:14,411
more, but if you take that meter
away, you're gonna try weird ideas.

216
00:11:14,411 --> 00:11:17,040
You're gonna run crazy tests overnight.

217
00:11:17,040 --> 00:11:20,483
You're gonna try 10 things
versus one thing, right?

218
00:11:20,496 --> 00:11:24,621
Running on a meter, are
more careful, right?

219
00:11:24,652 --> 00:11:30,588
And what you want is, is to take that away
so you can drive the curiosity and, like,

220
00:11:31,000 --> 00:11:31,164
Robert: More

221
00:11:31,323 --> 00:11:32,229
Scott Slater: That, that's right.

222
00:11:32,260 --> 00:11:32,724
You know?

223
00:11:32,762 --> 00:11:36,850
And, and look, the people that are
the most curious, that don't have

224
00:11:36,850 --> 00:11:40,819
that restriction and are the most
creative, those are the ones who

225
00:11:40,821 --> 00:11:45,787
think of the next big things, I think
that's the big thing here, to enable

226
00:11:46,001 --> 00:11:47,843
the freedom to get that outcome

227
00:11:48,790 --> 00:11:50,427
Neil: Totally agree with that, Scott.

228
00:11:50,874 --> 00:11:53,779
I was gonna say, there's been a
big push from enterprises around

229
00:11:54,024 --> 00:11:57,805
token economics, and we're running
this workload in the cloud.

230
00:11:57,814 --> 00:12:01,338
What facet of that workload can come
to the edge, and what does that,

231
00:12:01,367 --> 00:12:04,485
correspond to in terms of an ROI
model over the cost of this device?

232
00:12:04,863 --> 00:12:08,181
And additional value add is, what
is the cost of lost experimentation?

233
00:12:08,210 --> 00:12:12,861
If I have a finite token budget and I
have my role specifically in scope, I'm

234
00:12:12,862 --> 00:12:17,699
preserving those tokens to accelerate
what I need to accomplish, and I'm not

235
00:12:17,708 --> 00:12:19,406
thinking about doing and achieving more.

236
00:12:19,407 --> 00:12:24,019
So this notion of unmetered
intelligence, right, taking away that.

237
00:12:24,028 --> 00:12:29,123
finite token limit not only helps
accelerate productivity, whether you're

238
00:12:29,125 --> 00:12:33,766
35,000 feet in the air, on-site with
a client in an air-gapped environment,

239
00:12:34,029 --> 00:12:39,239
but it also unlocks, the opportunity
to create tooling that wouldn't

240
00:12:39,239 --> 00:12:43,693
be possible when you're operating
under just a fixed cap mindset

241
00:12:44,334 --> 00:12:45,382
Scott Slater: know, Neil, I,

242
00:12:46,412 --> 00:12:50,830
I had a developer say that exact
thing last week, and they wanted

243
00:12:50,841 --> 00:12:52,893
the right to fail cheaply,

244
00:12:53,193 --> 00:12:54,845
Neil: Mm.

245
00:12:54,993 --> 00:12:59,469
Scott Slater: wanted the right to not
ask permission, and I think that's part

246
00:12:59,469 --> 00:13:03,157
of why people want these devices where
they don't have to ask permission,

247
00:13:03,159 --> 00:13:04,869
and the cloud doesn't give them that.

248
00:13:05,183 --> 00:13:07,909
You know, in these big enterprise
companies, they need to ask

249
00:13:07,924 --> 00:13:10,165
permission to have more tokens

250
00:13:12,325 --> 00:13:15,644
Robert: if we wanna tie this back to
the original question to Scott about

251
00:13:15,645 --> 00:13:18,749
what customers are saying and then the
token economics piece, I learned this

252
00:13:18,749 --> 00:13:23,821
week through talking with customers the
amount of overspend that is happening in

253
00:13:23,823 --> 00:13:32,825
the cloud, either due to token usage or
token cost, raw token cost, dwarfs the

254
00:13:32,834 --> 00:13:38,524
increased price is for RAM or storage
or GPUs or anything in the hardware.

255
00:13:38,590 --> 00:13:40,598
It absolutely dwarfs it.

256
00:13:40,618 --> 00:13:45,917
So for some of us that are in the
Windows and PC game, we're like, "Oh

257
00:13:45,917 --> 00:13:48,898
my God, the world's on fire." To these
guys, it's like, "Eh, it's a drop in

258
00:13:48,899 --> 00:13:51,213
the bucket" for, for the potential

259
00:13:51,266 --> 00:13:51,506
Jacob: Угу

260
00:13:51,582 --> 00:13:51,972
Robert: could save.

261
00:13:52,630 --> 00:13:53,839
It was a huge eye-opener

262
00:13:54,975 --> 00:13:58,323
Frankcx: this is a question back to Scott
too, and I'd love other people's input

263
00:13:58,326 --> 00:14:01,334
because it feels like everybody's doing
a little bit of vibe coding out there.

264
00:14:01,799 --> 00:14:06,805
But Scott, as you look at, new sales
reps that you're hiring on and, bringing

265
00:14:06,805 --> 00:14:10,901
into the fold, obviously different reps
have different skills of relationship

266
00:14:10,919 --> 00:14:12,908
compared to technology background.

267
00:14:13,265 --> 00:14:17,454
But with the vibe coding story that's out
there these days, what happens when every

268
00:14:17,455 --> 00:14:22,679
seller can build software of their own and
be able to tell a story to their customers

269
00:14:22,688 --> 00:14:24,840
to showcase the value of these devices?

270
00:14:24,842 --> 00:14:26,431
I don't know if you've
started to see that.

271
00:14:26,697 --> 00:14:30,112
Obviously, Robert works on your
team, so that might be one of your

272
00:14:30,344 --> 00:14:34,238
example reps that, you have out
there that can showcase, like, I can

273
00:14:34,240 --> 00:14:39,347
customize something to really sell the
customer on the story of what AI has.

274
00:14:39,681 --> 00:14:41,429
But I'm wondering, do
you have a take on that?

275
00:14:41,867 --> 00:14:44,561
Scott Slater: Yeah, I've thought
about that, and I think what worries

276
00:14:44,576 --> 00:14:50,045
me a little bit is that, look, every
person, every seller can build today.

277
00:14:50,382 --> 00:14:55,525
They can build apps, dashboards,
trackers, automation, right?

278
00:14:55,861 --> 00:14:58,108
Building isn't the hard part anymore.

279
00:14:58,166 --> 00:14:59,726
That's the easy part now.

280
00:15:00,025 --> 00:15:06,191
But keeping something running and,
that's a hard thing to do, or knowing

281
00:15:06,379 --> 00:15:11,271
which problem you should actually
build something for, I think that's the

282
00:15:11,272 --> 00:15:13,312
harder part today, in my opinion, right?

283
00:15:13,533 --> 00:15:17,526
The building part is easier and, and
I think what we're gonna see soon

284
00:15:17,526 --> 00:15:21,923
is, like, thousands of these little
apps and dashboards and trackers

285
00:15:22,179 --> 00:15:25,144
that one person owns or nobody owns.

286
00:15:25,178 --> 00:15:30,301
And it means is IT's not owning this,
and I think that's a problem that's gonna

287
00:15:30,301 --> 00:15:33,198
really raise up soon in all organizations

288
00:15:33,198 --> 00:15:38,101
As all these things get built, they're not
necessarily solving for the right things

289
00:15:39,348 --> 00:15:39,994
Neil: I have an ig-

290
00:15:40,217 --> 00:15:43,332
Frankcx: for the rest of the team here,
like being able to customize something

291
00:15:43,333 --> 00:15:48,015
that kind of specifically aims at
the customer's voice, if you've seen

292
00:15:48,023 --> 00:15:50,061
value in that type of storytelling?

293
00:15:51,400 --> 00:15:52,287
Neil: 100%.

294
00:15:52,305 --> 00:15:59,361
I walk into every customer partner meeting
with a custom demo pack, and to Scott's

295
00:15:59,365 --> 00:16:04,602
point, art of possible, doing my homework,
meeting with the reps that already know

296
00:16:04,602 --> 00:16:09,960
that customer, doing my research online,
so that I can come prepared with a custom

297
00:16:09,979 --> 00:16:14,305
artifact, functional local AI solution
in every customer meeting has been my

298
00:16:14,306 --> 00:16:16,278
mindset for the last eight, nine months.

299
00:16:16,611 --> 00:16:19,634
Keep in mind, I have an electrical
computer engineering background.

300
00:16:19,839 --> 00:16:22,282
I hadn't touched code in 15 years, right?

301
00:16:22,282 --> 00:16:26,956
And so the ability to now, again, create
these prototypes in my favorite language,

302
00:16:26,970 --> 00:16:30,623
English, meet customers where they are
on their journey, extremely impactful.

303
00:16:30,916 --> 00:16:32,881
I do wanna highlight the
inverse of that, though.

304
00:16:32,940 --> 00:16:37,102
In a world where you can build
anything and creation is no longer

305
00:16:37,103 --> 00:16:40,588
the bottleneck, choosing which
problems to solve, super important.

306
00:16:40,890 --> 00:16:45,332
I met with a customer, and they had,
essentially a rogue experiment gone wrong.

307
00:16:45,608 --> 00:16:48,830
Someone came up with the idea
of a brand-new ERP platform.

308
00:16:49,118 --> 00:16:54,977
They consumed six figures in tokens, only
to find out that they had already licensed

309
00:16:54,978 --> 00:16:57,327
a perfectly functional ERP platform.

310
00:16:57,630 --> 00:17:02,091
So that's over 100 grand down the
drain on top of that time investment.

311
00:17:02,489 --> 00:17:05,560
So again, when in a world where
you can build anything, make sure

312
00:17:05,561 --> 00:17:06,662
you're building the right thing.

313
00:17:06,934 --> 00:17:10,478
Avoid that, disaster scenario
of 100 grand in tokens down the

314
00:17:10,478 --> 00:17:12,526
drain to build a duplicate system

315
00:17:13,329 --> 00:17:19,720
Robert: And the, the other side to that
is say you have 1,000 things people

316
00:17:19,720 --> 00:17:22,896
built, but they're built by 1,000
different people, and then what becomes

317
00:17:22,920 --> 00:17:27,101
the standard for… I mean, everybody's
building something that they think

318
00:17:27,101 --> 00:17:31,740
is awesome, but now we have like 100
different versions of something, and

319
00:17:31,740 --> 00:17:35,406
nobody knows which is the, you know,
which is the blessed one and, and whatnot.

320
00:17:35,407 --> 00:17:38,777
I had to build a customer
one-pager today, for one of our

321
00:17:38,777 --> 00:17:40,767
OEMs and couldn't find anything.

322
00:17:40,860 --> 00:17:43,274
Took a long while to do it, and the
first thought that came to my mind is,

323
00:17:43,277 --> 00:17:47,485
"Oh, I should share this." And I'm like,
"No." I know what that's gonna create

324
00:17:47,485 --> 00:17:51,794
is the problem of this proliferation,
and then the governance and ownership of

325
00:17:52,255 --> 00:17:53,989
what you should or shouldn't be sharing

326
00:17:55,191 --> 00:17:58,432
Frankcx: I'd love to turn this to
Jacob, because Jacob, you saying that

327
00:17:58,432 --> 00:18:01,703
you stood in a Best Buy with a bunch
of influential people, but also watched

328
00:18:01,704 --> 00:18:03,420
a lot of devices go out the door.

329
00:18:03,870 --> 00:18:07,102
I'm reminded of, when I was at Build
one time and somebody came up and

330
00:18:07,103 --> 00:18:12,880
literally bought a $4,000 Surface
laptop, and now I see that there

331
00:18:12,880 --> 00:18:14,491
is a return of the workstation.

332
00:18:14,845 --> 00:18:17,039
the workstation, I would say, once died.

333
00:18:17,093 --> 00:18:20,482
we didn't see a value because people
were pushing so much up to Azure.

334
00:18:20,781 --> 00:18:25,978
Laura, coworker of, Robert's and Neil's
in the past, or still Neil, she pointed

335
00:18:25,978 --> 00:18:30,481
out, like, I spent a decade pushing
people to start using Azure in the

336
00:18:30,481 --> 00:18:34,997
cloud, and not that Azure-- I think
the Azure numbers are still secure, but

337
00:18:34,997 --> 00:18:38,237
we're starting to see people start to
think about a workstation again, like

338
00:18:38,237 --> 00:18:43,699
the power of a super powerful device,
a device that can make money, for you.

339
00:18:43,707 --> 00:18:47,322
So I'm wondering, Jacob, when people
were dishing out that much money…

340
00:18:47,629 --> 00:18:50,730
Well, of course, everybody in San
Francisco's got unlimited cash flow.

341
00:18:52,408 --> 00:18:53,328
Jacob: Not necessarily

342
00:18:53,428 --> 00:18:55,138
Frankcx: I'm like, what
do you think about that?

343
00:18:55,221 --> 00:18:57,012
Like, the return of the workstation?

344
00:18:57,988 --> 00:19:03,928
Jacob: the first person that bought an
RTX Spark device in San Francisco, he was

345
00:19:03,929 --> 00:19:08,835
on camera and shouted out to everyone,
like, essentially, so excited to use this

346
00:19:08,852 --> 00:19:16,656
to build without limits and to be able to
not have to worry about my token costs and

347
00:19:16,664 --> 00:19:19,782
also do some awesome gaming and stuff."

348
00:19:19,848 --> 00:19:22,440
So he was going to build
his business around it.

349
00:19:22,448 --> 00:19:24,173
What his business was, I have no clue,

350
00:19:24,269 --> 00:19:24,361
Robert: to

351
00:19:24,589 --> 00:19:26,609
Jacob: but he was gonna build his
business around this investment.

352
00:19:26,821 --> 00:19:27,372
Robert: How long did it

353
00:19:27,387 --> 00:19:27,865
Jacob: What's that?

354
00:19:28,118 --> 00:19:28,512
Robert: that?

355
00:19:30,529 --> 00:19:30,791
thought he

356
00:19:30,894 --> 00:19:32,858
Jacob: you can look at the PCMag video.

357
00:19:32,876 --> 00:19:34,726
Like, they've got an Instagram reel up.

358
00:19:34,738 --> 00:19:37,741
They've probably got TikToks up
where it highlights what he said

359
00:19:37,771 --> 00:19:38,926
when he was walking out of that.

360
00:19:39,248 --> 00:19:42,129
But it was… Maybe I butchered it,
but it was something to that effect

361
00:19:42,608 --> 00:19:45,074
They're like, "Oh, we wanna clap,"
but, "Oh, he keeps on going.

362
00:19:45,085 --> 00:19:48,805
Oh, he keeps on going." It was like a
quasi interview question, and everyone was

363
00:19:48,805 --> 00:19:50,296
cheering and clapping and it was exciting.

364
00:19:50,460 --> 00:19:54,178
That was the type of sentiment that it
was kind of a hybrid of what Scott and

365
00:19:54,178 --> 00:19:57,745
Robert and y- like what we were already
talking about that like kind of opens up

366
00:19:57,745 --> 00:20:01,578
your ability to create, your ability to,
to not feel like you're compromised by the

367
00:20:01,579 --> 00:20:07,104
amount of, tokens you have or even by like
the sovereignty of your data, like where

368
00:20:07,104 --> 00:20:12,766
it's going, like what, what, you know,
what, who has access to your sensitive.

369
00:20:12,774 --> 00:20:15,756
For a lot of these founders that
I talk to, like who has access

370
00:20:15,759 --> 00:20:17,523
to their sensitive information?

371
00:20:18,152 --> 00:20:22,567
and they're not just gonna use a
standard, you know, consumer, AI model

372
00:20:22,976 --> 00:20:26,345
in the cloud for building their own
business that may be competing with

373
00:20:26,381 --> 00:20:29,116
that cloud solution that already exists.

374
00:20:29,271 --> 00:20:29,647
Frankcx: Jacob,

375
00:20:29,793 --> 00:20:31,090
Jacob: And so that's
something interesting.

376
00:20:31,097 --> 00:20:36,089
There are obviously enterprise versions
of these that have license agreements,

377
00:20:36,117 --> 00:20:41,119
but it's still an interesting thing
like to think about that privacy

378
00:20:41,189 --> 00:20:44,111
Frankcx: Well, I know that Jacob, you like
to make fun of me because I make reference

379
00:20:44,111 --> 00:20:48,244
to some of the older technologies
that I've worked with, but it would be

380
00:20:48,246 --> 00:20:50,520
that, you know, we watch that pendulum

381
00:20:50,660 --> 00:20:51,239
Jacob: Got it, thanks

382
00:20:51,540 --> 00:20:55,293
Frankcx: where we talk about going
from like a mainframe to a PC, and

383
00:20:55,293 --> 00:20:59,006
I would argue like Lotus 1-2-3 was
probably like what caused a lot of

384
00:20:59,015 --> 00:21:03,768
people to start buying a PC what is that
application that allowed them to do that?

385
00:21:03,786 --> 00:21:06,907
I always name drop that I worked for
this company, Silicon Graphics, that

386
00:21:06,908 --> 00:21:11,251
got put literally out of business by
Nvidia at the time by selling a $300

387
00:21:11,264 --> 00:21:15,937
gaming card, that business caused
people to go from like a large device

388
00:21:15,938 --> 00:21:20,774
to a small device again because of
this $300 gaming card and be able to do

389
00:21:20,785 --> 00:21:22,774
things like Photoshop on your device.

390
00:21:23,236 --> 00:21:27,172
But I would say like, what's that thing
that's really moving the needle or moving

391
00:21:27,173 --> 00:21:32,432
the dial, you know, that causes people
to buy a workstation, like at $6,000?

392
00:21:32,448 --> 00:21:33,966
Is it open weight models?

393
00:21:33,969 --> 00:21:35,756
Is it agentic coding?

394
00:21:36,131 --> 00:21:37,912
Like, you know, what are
you seeing out there?

395
00:21:37,912 --> 00:21:41,064
And I guess for Neil and Robert
and Scott, as well, like what are

396
00:21:41,066 --> 00:21:44,039
you guys seeing that is kind of
pushing people to have interest in

397
00:21:44,040 --> 00:21:45,724
these workstation class devices?

398
00:21:46,418 --> 00:21:47,130
Robert: you go first

399
00:21:47,964 --> 00:21:49,208
Scott Slater: yeah, you have the honor.

400
00:21:49,299 --> 00:21:53,219
So you go, then I'll go, 'cause I
think I'll have a different perspective

401
00:21:53,646 --> 00:21:58,605
Robert: I think the open weight
models is the one that's getting

402
00:21:58,605 --> 00:22:04,723
people hooked and interested, as
the first kind of step to this.

403
00:22:05,056 --> 00:22:08,041
I think it goes back to, I don't
know who said it, but about

404
00:22:08,052 --> 00:22:12,061
the, like, the unlimited, like,
limitless creativity, right?

405
00:22:12,085 --> 00:22:17,805
I have the ability to do this without a
connection, without oversight, without

406
00:22:18,076 --> 00:22:20,282
token spend, all of these kind of things.

407
00:22:20,582 --> 00:22:26,100
I would love nothing more than
to have a AI workstation, just to

408
00:22:26,231 --> 00:22:28,311
run everything I wanna do off of.

409
00:22:28,311 --> 00:22:35,172
There's probably 80% of the things I don't
do right now is because I don't have AI

410
00:22:35,186 --> 00:22:37,199
workstation sitting right here beside me.

411
00:22:37,486 --> 00:22:41,038
there's a whole bunch of stuff I wanna
try out and do, but I'm not gonna burn

412
00:22:41,075 --> 00:22:42,896
tokens to go, to go figure that out.

413
00:22:42,901 --> 00:22:45,458
I'll wait till, the chains
come off, so to speak

414
00:22:47,224 --> 00:22:48,179
Scott Slater: I'll add to that.

415
00:22:48,192 --> 00:22:53,186
what I'm excited about is
what's going to happen, right?

416
00:22:53,240 --> 00:22:57,063
Like, so I know we're talking
about on-device AI and, you

417
00:22:57,064 --> 00:22:58,238
know, what's exciting there.

418
00:22:58,238 --> 00:23:04,243
I think what the exciting part is the
stuff that users don't even notice, right?

419
00:23:04,555 --> 00:23:08,054
And I think that's just gonna
continue to evolve, right?

420
00:23:08,055 --> 00:23:14,011
So we know, like, now there's AI models
that make video calls better, right?

421
00:23:14,037 --> 00:23:19,969
Or that can do live captions, or
you can search files faster, right?

422
00:23:20,110 --> 00:23:23,873
Or stuff that can work when you're
30,000 feet on a plane with no

423
00:23:23,881 --> 00:23:28,141
connectivity, and like, that's the
stuff that gets me real excited.

424
00:23:28,142 --> 00:23:31,802
Like, don't get me wrong, I'm still
excited about local models and, you

425
00:23:31,803 --> 00:23:35,781
know, I'm prepared to talk about local
models today, and I know we announced

426
00:23:35,783 --> 00:23:37,786
a whole bunch, about a week ago, right?

427
00:23:37,786 --> 00:23:40,845
Like MAI, which stands for
Microsoft AI, by the way.

428
00:23:40,896 --> 00:23:41,379
Robert: Yeah

429
00:23:41,632 --> 00:23:43,840
Scott Slater: A super creative
name that we came up with, right?

430
00:23:44,086 --> 00:23:45,550
I'm really excited about that.

431
00:23:45,565 --> 00:23:51,646
Like, the stuff that nobody knows and it's
just there that the new hardware enables.

432
00:23:51,997 --> 00:23:56,573
And, new hardware plus, the models and the
apps that are tapping into those models.

433
00:23:56,613 --> 00:24:01,967
I'll try not to talk a lot about Microsoft
products, but, the hybrid intelligence

434
00:24:01,980 --> 00:24:04,768
stuff really I'm excited about, right?

435
00:24:04,775 --> 00:24:10,295
'Cause the idea of that, that was
announced, is that a user wouldn't

436
00:24:10,304 --> 00:24:16,420
know, Like, the hybrid intelligence
would decide if it's leveraging a

437
00:24:16,420 --> 00:24:21,691
cloud model or a local model based
on some efficiency metric, right?

438
00:24:21,709 --> 00:24:24,044
Like performance, latency, cost.

439
00:24:24,417 --> 00:24:25,721
Like, that's really…

440
00:24:26,050 --> 00:24:26,354
Robert: Yeah.

441
00:24:26,407 --> 00:24:26,884
Scott Slater: Privacy.

442
00:24:26,894 --> 00:24:28,313
Like, that's really interesting.

443
00:24:28,328 --> 00:24:32,241
Like, if you don't need to know as
a user and you never notice it, but

444
00:24:32,242 --> 00:24:34,390
you have the right hardware, right?

445
00:24:34,634 --> 00:24:38,583
Frankcx: Yeah, Neil, I wonder, like,
when you're talking to customers, that

446
00:24:38,584 --> 00:24:42,128
story that, Scott is, talking about
it is really starting to resonate

447
00:24:42,167 --> 00:24:45,680
with a lot of our commercial or
enterprise-grade customers out there?

448
00:24:46,636 --> 00:24:47,588
Neil: 100%.

449
00:24:47,613 --> 00:24:53,464
we've been promoting local hybrid
intelligence, for years, and now that it's

450
00:24:53,484 --> 00:24:59,506
coming essentially baked into the tools
customers already own, that barrier to

451
00:24:59,507 --> 00:25:02,286
adopt hybrid intelligence has disappeared.

452
00:25:02,562 --> 00:25:06,042
Before, you had to go out and
either buy or build a routing

453
00:25:06,045 --> 00:25:09,571
mechanism, which frontier customers
are doing, but now we're meeting

454
00:25:09,572 --> 00:25:11,005
them right where they are, right?

455
00:25:11,020 --> 00:25:15,103
that auto-routing capability in
GitHub Copilot with HydroFusion,

456
00:25:15,384 --> 00:25:16,493
absolute game changer.

457
00:25:16,785 --> 00:25:20,401
I'll highlight on the topic of
unlimited creativity, I learned

458
00:25:20,401 --> 00:25:23,750
during my time at ASU, will.i.am
is actually a professor there.

459
00:25:24,053 --> 00:25:28,926
He has 200 students, in the creative
department, and NVIDIA gave him a grant.

460
00:25:28,937 --> 00:25:33,762
All 200 of his students have a
Spark device, to tinker with, build

461
00:25:33,763 --> 00:25:34,781
with throughout their semester.

462
00:25:34,960 --> 00:25:37,347
Frankcx: Can I sit in on that course?

463
00:25:37,708 --> 00:25:39,876
Neil: going back to school
myself to attend this one.

464
00:25:40,136 --> 00:25:41,498
and so think about that, right?

465
00:25:41,538 --> 00:25:45,176
Unlimited creativity for will.i.am
students throughout the semester,

466
00:25:45,392 --> 00:25:49,455
essentially creating a digital twin
of their creative, persona, whether

467
00:25:49,455 --> 00:25:51,490
they're, a musician, artist or more.

468
00:25:51,716 --> 00:25:54,701
we've really just hit the tip of the
iceberg in terms of what's possible

469
00:25:54,707 --> 00:25:59,534
when we bring unmetered intelligence,
and hybrid intelligence together,

470
00:26:00,215 --> 00:26:01,916
unlocking a new era of computing

471
00:26:02,539 --> 00:26:03,233
Scott Slater: So Neil

472
00:26:03,734 --> 00:26:05,990
Jacob: I mean, you guys have
been doing this for a while.

473
00:26:05,992 --> 00:26:10,868
Like, I remember Robert, Neil, you showed
me a demo two years ago of, essentially

474
00:26:10,868 --> 00:26:14,908
deciding where that workload should go,
you know, whether it's cloud or it's not.

475
00:26:14,998 --> 00:26:17,446
and I remember being blown away
by that, going like, "Okay."

476
00:26:17,446 --> 00:26:20,470
And that took some effort to do.

477
00:26:20,501 --> 00:26:22,973
Like, we would talk about, oh,
it's just some lines of code,

478
00:26:23,013 --> 00:26:24,209
but, it took some effort to do.

479
00:26:24,209 --> 00:26:26,647
You'd have to have the right
accounts, the right things, and,

480
00:26:26,647 --> 00:26:28,066
having that built in is awesome.

481
00:26:28,113 --> 00:26:32,488
And I know Robert, shameless plug for
the Surface IT Pro blog, but we've got,

482
00:26:32,548 --> 00:26:34,984
a blog post that just landed this week.

483
00:26:35,258 --> 00:26:37,579
it was actually authored by one
of our very own, so we've got a

484
00:26:37,579 --> 00:26:38,864
thought leader here on this call.

485
00:26:38,876 --> 00:26:40,402
Robert, you wanna talk at all about that?

486
00:26:41,448 --> 00:26:42,989
Robert: how awkward.

487
00:26:43,251 --> 00:26:44,145
Frankcx: How awkward.

488
00:26:45,714 --> 00:26:48,925
Robert: listen, to your point, Jacob,
full disclosure, I was in Surface.

489
00:26:48,951 --> 00:26:49,793
Now I'm in Windows.

490
00:26:49,798 --> 00:26:53,515
Before I left, we wrote a hybrid
intelligence blog, waited till the

491
00:26:53,515 --> 00:26:56,474
announcement came, so the people
knew what we were talking about

492
00:26:56,474 --> 00:27:00,166
with hybrid intelligence, which
essentially now replaces unmetered

493
00:27:00,169 --> 00:27:02,355
intelligence, in the nomenclature.

494
00:27:02,358 --> 00:27:04,446
But it's all about, what
we've been talking about.

495
00:27:04,476 --> 00:27:08,549
Lookit, it's all about what our blog
or what our podcast is about, and

496
00:27:08,550 --> 00:27:10,137
just talks about it in those terms.

497
00:27:10,370 --> 00:27:13,055
Frankcx: kinda hoping you'd
title it Super Intelligence,

498
00:27:13,275 --> 00:27:13,730
Robert: Oh.

499
00:27:14,534 --> 00:27:16,751
I did, but it did not make it past legal

500
00:27:19,320 --> 00:27:22,274
There is another one coming
out, by the way, on, on device

501
00:27:22,278 --> 00:27:24,029
economics as well that, I think

502
00:27:24,456 --> 00:27:25,107
Frankcx: Definitely check

503
00:27:25,152 --> 00:27:26,480
Robert: our audience will enjoy

504
00:27:26,594 --> 00:27:28,021
Frankcx: blog that Robert put together.

505
00:27:28,042 --> 00:27:29,379
Super insightful.

506
00:27:30,412 --> 00:27:32,399
I want to turn the wheels a little bit.

507
00:27:32,449 --> 00:27:36,637
we talk about this idea of the return
of the workstation, but I'm also

508
00:27:36,639 --> 00:27:40,300
reminded of, what Satya has put out
there in the past in some of his blog

509
00:27:40,301 --> 00:27:43,177
posts around keeping the intelligence.

510
00:27:43,191 --> 00:27:48,450
You know, you shouldn't have to share
your own intelligence with a cloud model.

511
00:27:48,474 --> 00:27:51,777
And I think, Scott, you made mention
of these new models from Microsoft,

512
00:27:51,777 --> 00:27:58,249
like Microsoft AI, and what Mustafa
out there is doing with his big giant

513
00:27:58,259 --> 00:28:00,246
team down in the Silicon Valley area.

514
00:28:00,532 --> 00:28:01,882
beautiful office, by the way.

515
00:28:01,883 --> 00:28:04,401
I got to visit once or
twice, and, pretty amazing.

516
00:28:04,824 --> 00:28:08,471
But, by the way, they offer free food
down there, which they don't give

517
00:28:08,510 --> 00:28:12,301
on the Microsoft campus, so I guess
that's a great in- incentive as to

518
00:28:12,451 --> 00:28:15,304
Scott Slater: I think they
grow grapes on the rooftop also

519
00:28:16,884 --> 00:28:17,850
Frankcx: Of course they do.

520
00:28:18,506 --> 00:28:21,806
but one of the things that really
stands out to me is that new model

521
00:28:21,806 --> 00:28:25,525
is now available in Azure as a
model that you could actually load

522
00:28:25,525 --> 00:28:30,599
yourself on your expensive Azure
infrastructure, but it's all yours,

523
00:28:30,663 --> 00:28:32,506
meaning, like, it's not trained.

524
00:28:32,518 --> 00:28:35,185
It doesn't learn from, the
things that you share from it.

525
00:28:35,198 --> 00:28:37,344
It's private into that space.

526
00:28:37,344 --> 00:28:41,057
And I wonder, like, that seems
to be another reason to me,

527
00:28:41,069 --> 00:28:44,642
like keeping the workstation and
keeping the data, into that space.

528
00:28:44,659 --> 00:28:48,025
And I'm wondering even Jacob or
Neil or Robert or even Scott, like,

529
00:28:48,025 --> 00:28:52,271
as you're talking to customers,
this idea of IP protection.

530
00:28:52,281 --> 00:28:57,134
Like, I can't share that information,
with anybody outside of my own space.

531
00:28:57,152 --> 00:29:01,350
And I also talk about this idea of
this second wave of AI, like these

532
00:29:01,360 --> 00:29:04,894
customers that just couldn't take
part in the first wave because they

533
00:29:04,894 --> 00:29:08,916
didn't want to share with Sam Altman
or Dario Amodei, you know, all of their

534
00:29:08,916 --> 00:29:10,876
internal, you know, trade secrets.

535
00:29:11,152 --> 00:29:15,645
But now with these new devices and, you
know, Microsoft artificial intelligence

536
00:29:15,691 --> 00:29:20,662
models or having workstations that the
data stays on that device, if we're

537
00:29:20,662 --> 00:29:24,160
starting to see some customers start
to think about that into that space.

538
00:29:25,785 --> 00:29:28,570
Scott Slater: Yeah, I'll… Well,
I, I could start giving you just

539
00:29:28,571 --> 00:29:30,074
a quick little thought on it.

540
00:29:30,182 --> 00:29:33,755
I think it's important for us to
talk with customers about these new

541
00:29:33,764 --> 00:29:40,132
local models that exist and how large
models, some of them are being shrunk,

542
00:29:40,513 --> 00:29:46,687
to provide just as good capability
locally than on the cloud, right?

543
00:29:47,022 --> 00:29:51,572
And, a lot of these models you'll
see have this word flash at the

544
00:29:51,580 --> 00:29:53,110
end of the name of the model.

545
00:29:53,366 --> 00:29:54,135
That's mean

546
00:29:54,152 --> 00:29:57,694
that means that it's fast, it's
lightweight, it's built for

547
00:29:57,696 --> 00:30:01,616
speed and cost, and It's not
intended to be the biggest model.

548
00:30:02,031 --> 00:30:05,610
And, I like to, for customers
that don't understand this,

549
00:30:05,630 --> 00:30:07,181
to give them some analogies.

550
00:30:07,185 --> 00:30:11,150
So one that everybody understands
is, like, an image, right?

551
00:30:11,254 --> 00:30:17,279
a full-resolution photo versus,
like, a compressed JPEG, right?

552
00:30:17,308 --> 00:30:19,221
And I think most people understand that.

553
00:30:19,221 --> 00:30:23,451
And for most people, you look at that
photo, you won't know the difference.

554
00:30:23,476 --> 00:30:28,637
But if you zoom in on the compressed one,
it may be a little more blurry, right?

555
00:30:28,639 --> 00:30:31,511
But for most people, the
compressed photo is good enough.

556
00:30:31,910 --> 00:30:38,185
Or in music, it's like a master
studio file versus, a compressed

557
00:30:38,191 --> 00:30:40,258
file for streaming, right?

558
00:30:40,494 --> 00:30:43,283
For most people, you're not gonna
notice the difference when you're

559
00:30:43,296 --> 00:30:46,510
driving in your car and you're
listening to it in your car.

560
00:30:46,511 --> 00:30:50,484
But, maybe an audiophile audio
engineer, they'll know the

561
00:30:50,485 --> 00:30:52,323
difference, but most people wouldn't.

562
00:30:52,721 --> 00:30:57,235
And I think that's, a nice way to
compare, a cloud model that may have been

563
00:30:57,236 --> 00:31:00,169
shrunk to now be a local supported model

564
00:31:03,377 --> 00:31:05,900
Jacob: Yeah, I would just point out
that there's concern sometimes from

565
00:31:05,900 --> 00:31:10,116
folks going, "Well, doesn't it lose some
precision when you compress it?" And while

566
00:31:10,116 --> 00:31:15,251
yes, in theory it does, when you quantum
model down, it does lose some precision.

567
00:31:15,261 --> 00:31:19,209
But nowadays, even with frontier
models, all of the work that

568
00:31:19,209 --> 00:31:21,374
folks are doing should be checked.

569
00:31:21,767 --> 00:31:24,491
And honestly, it's often
checked by another model.

570
00:31:24,725 --> 00:31:26,817
Like that's built into
most workflows nowadays.

571
00:31:27,093 --> 00:31:30,590
And so you have one model that's shrunk
down that's doing some great work,

572
00:31:30,590 --> 00:31:34,004
heavy lifting, and all of that work
should be checked by another model.

573
00:31:34,169 --> 00:31:37,104
It might be another version of
that same model or a different one

574
00:31:37,106 --> 00:31:38,734
that's tuned for different things.

575
00:31:39,043 --> 00:31:42,273
And so together, those two smaller
models make an output that's

576
00:31:42,274 --> 00:31:45,356
better than even one larger model
could do, which is kind of cool.

577
00:31:45,529 --> 00:31:46,086
Scott Slater: models?

578
00:31:47,373 --> 00:31:48,704
Okay.

579
00:31:51,707 --> 00:31:55,827
Frankcx: I think that's pretty common
these days in writing applications.

580
00:31:55,941 --> 00:31:59,958
I'm writing an application right now
for a partner company of ours that

581
00:31:59,959 --> 00:32:03,752
does meeting room applications, and
I'm writing it with Codex, but I'm

582
00:32:03,760 --> 00:32:08,011
using Claude to do all of the check,
so that they balance each other.

583
00:32:08,280 --> 00:32:12,920
And then even this last week, I noticed
that, Google released Argon, and so

584
00:32:12,920 --> 00:32:16,902
I've been using Argon to be almost
a third-party reviewer because there

585
00:32:16,902 --> 00:32:19,096
are blind spots that these AIs create.

586
00:32:19,109 --> 00:32:22,989
And so I asked Argon to go through
And do a full scrub of the code base.

587
00:32:23,283 --> 00:32:25,737
And sure enough, it came up
with a couple, blind spots that

588
00:32:25,737 --> 00:32:27,536
Claude and Codex, didn't see.

589
00:32:27,644 --> 00:32:31,665
so it's amazing to watch, how you
can drive application development

590
00:32:31,877 --> 00:32:36,030
by trusting three agentic
coding agents to drive that, so

591
00:32:36,728 --> 00:32:39,931
Jacob: And by the way, I've got
three, different models running

592
00:32:39,932 --> 00:32:43,465
on one device next to me, an
RTX Spark Surface Laptop Ultra.

593
00:32:43,865 --> 00:32:47,629
they're all about f- 20 to 40 billion
parameter models all running at

594
00:32:47,630 --> 00:32:49,610
the same time in the same workflow.

595
00:32:49,636 --> 00:32:52,023
So th- like that, that
is totally possible and

596
00:32:52,153 --> 00:32:52,728
Robert: by the way.

597
00:32:52,767 --> 00:32:53,686
Frankcx: Four, yeah,

598
00:32:53,836 --> 00:32:54,826
Robert: low-key brags

599
00:32:54,947 --> 00:32:55,264
Frankcx: yeah

600
00:32:55,540 --> 00:32:57,660
Jacob: Hey, we're here to
talk about local AI, you

601
00:32:57,747 --> 00:32:58,659
Robert: ran my first,

602
00:32:58,750 --> 00:32:59,766
Jacob: as well, talk
about what we're doing

603
00:33:00,331 --> 00:33:02,327
Robert: model, a 32 billion

604
00:33:02,369 --> 00:33:02,379
Frankcx: Oh

605
00:33:02,628 --> 00:33:04,167
Robert: MoE two-bit model.

606
00:33:04,552 --> 00:33:05,642
and I was Very impressed.

607
00:33:05,903 --> 00:33:10,082
Actually, Jacob and I did a rogue
podcast, I think earlier this week, and

608
00:33:10,082 --> 00:33:12,243
we talked about models checking models

609
00:33:13,768 --> 00:33:14,992
Jacob: we did, yes.

610
00:33:15,165 --> 00:33:16,948
Scott Slater: of the
new Transformers model?

611
00:33:18,448 --> 00:33:19,210
Robert: Ooh, do tell.

612
00:33:19,235 --> 00:33:19,451
Do

613
00:33:19,698 --> 00:33:23,311
Scott Slater: So that's my way of
referring if you heard of NVIDIA's

614
00:33:23,335 --> 00:33:27,267
Nemotron model, it sounds like
Megatron, so I'm just calling

615
00:33:27,268 --> 00:33:28,731
it the new Transformers model.

616
00:33:28,832 --> 00:33:30,799
I think it will be a
fun way to talk about it

617
00:33:32,699 --> 00:33:32,821
Frankcx: It's

618
00:33:33,078 --> 00:33:36,121
Jacob: Yeah, I've been using
Nemotron or Transformer, whatever

619
00:33:36,122 --> 00:33:38,651
you wanna call it, 3.5 Bolt.

620
00:33:38,987 --> 00:33:46,534
It's a 70 billion parameter MoE model,
which fits into 28 gigs of VRAM, which

621
00:33:46,534 --> 00:33:50,687
is awesome because, you know, we've
been talking about the $6,000 device.

622
00:33:50,711 --> 00:33:54,134
Like, you don't need the $6,000
version to run that model.

623
00:33:54,144 --> 00:33:58,715
You can actually run it on a $4,000
version, and still get great results.

624
00:33:58,725 --> 00:33:59,024
So

625
00:34:00,008 --> 00:34:00,456
Frankcx: All right.

626
00:34:00,507 --> 00:34:00,978
Jacob: it's been cool.

627
00:34:01,448 --> 00:34:03,202
Frankcx: Let's finish up
the podcast this week.

628
00:34:03,204 --> 00:34:07,034
I know, Jacob, you gotta get going pretty
soon, but I'd love to go around the horn

629
00:34:07,035 --> 00:34:08,665
and talk about what we're working on.

630
00:34:09,006 --> 00:34:12,307
many of you, you know, on the podcast,
you're aware that I do a little bit of,

631
00:34:12,308 --> 00:34:15,786
Friday night and Saturday night Uber
driving so I can figure out what the cool

632
00:34:15,786 --> 00:34:18,006
kids are doing in my retirement age here.

633
00:34:18,394 --> 00:34:22,109
but one of the things that I love to be
able to do is that device that's literally

634
00:34:22,110 --> 00:34:27,067
sitting behind me, this workstation,
it collects twenty-three API feeds.

635
00:34:27,094 --> 00:34:30,753
It looks at FlightAware to understand
when planes are landing at SeaTac.

636
00:34:31,048 --> 00:34:34,596
It looks at when is my next bus,
which is an interesting API that can

637
00:34:34,597 --> 00:34:38,427
tell you when the next bus or train
is supposed to pull into a station.

638
00:34:38,854 --> 00:34:42,289
so I use it to understand, like,
what are trains coming from Seattle

639
00:34:42,525 --> 00:34:44,283
and, like, did a concert let out?

640
00:34:44,283 --> 00:34:45,956
Did a Mariners game let out?

641
00:34:45,960 --> 00:34:47,438
Did a Seahawks game let out?

642
00:34:47,448 --> 00:34:50,417
And I can actually pick up
people, you know, in that space.

643
00:34:50,635 --> 00:34:54,422
But it also looks at, there's an API
that understands what is traffic.

644
00:34:54,422 --> 00:34:58,464
Like, you know how when you look at, like
you're going to a restaurant, and Google

645
00:34:58,465 --> 00:35:02,846
can tell you that this is higher traffic
period or it's a lower traffic period?

646
00:35:03,119 --> 00:35:07,217
Well, I can understand by subscribing to
it what are all the traffic, spaces that

647
00:35:07,217 --> 00:35:11,033
I can understand where are all the people
and what restaurants are they going to.

648
00:35:11,034 --> 00:35:16,339
And so it all runs on, my phone, and
my phone can access my device that's

649
00:35:16,340 --> 00:35:20,783
behind me, but my device will look
across all that data, and it uses a

650
00:35:20,789 --> 00:35:25,765
QUEN model to scrub all that data that
understands where should I go next in

651
00:35:25,766 --> 00:35:29,547
order to see the highest propensity
of where riders could show up.

652
00:35:29,898 --> 00:35:33,066
And so I've been using it, and it's
actually starting to pay off a little bit.

653
00:35:33,110 --> 00:35:33,678
I don't know.

654
00:35:33,692 --> 00:35:37,056
It's hard to say whether it's, like,
better or worse, but it's kinda cool

655
00:35:37,057 --> 00:35:40,725
to have that intelligence to be able
to say, "Where should I be and where

656
00:35:40,726 --> 00:35:44,292
should I position myself to find
the highest propensity of riders?"

657
00:35:44,306 --> 00:35:45,316
So I'm

658
00:35:45,508 --> 00:35:46,931
Jacob: You're gonna
publish it and sell it,

659
00:35:47,037 --> 00:35:47,101
Frankcx: I

660
00:35:47,190 --> 00:35:47,693
Jacob: Frank?

661
00:35:47,990 --> 00:35:50,008
Frankcx: maybe it'll be an
app that'll be out there.

662
00:35:50,016 --> 00:35:52,822
It would have to run on a device,
whether that's in Azure or

663
00:35:52,822 --> 00:35:54,450
whether that's on my local device.

664
00:35:54,768 --> 00:35:59,466
But you could go out and buy a RTX
Spark or a DGX Spark device and set up

665
00:35:59,473 --> 00:36:03,482
Tailscale into your own environment and
be able to run it that way, where you're

666
00:36:03,490 --> 00:36:08,013
hosting it yourself rather than sharing
it as a shared app across many people.

667
00:36:08,411 --> 00:36:12,153
But what I like to point out is,
are there commercial instances where

668
00:36:12,419 --> 00:36:16,710
collecting a ton of API data and
then asking for an instant in time

669
00:36:16,711 --> 00:36:20,537
in order to make a decision with an
AI model, is that something that,

670
00:36:20,550 --> 00:36:24,717
Scott or Neil or Robert or even Jacob
would be like, do you see customers

671
00:36:25,021 --> 00:36:27,083
that could utilize cases like that?

672
00:36:27,114 --> 00:36:28,854
I'm not trying to push it as my service.

673
00:36:28,870 --> 00:36:33,356
I'm more trying to say like, you
know, instead of buying an RTX laptop,

674
00:36:33,394 --> 00:36:36,366
buying an RTX-based PC in the future.

675
00:36:36,659 --> 00:36:39,981
You know, do we see use cases like
that of being able to have an always-on

676
00:36:39,982 --> 00:36:43,974
device that is collecting and monitoring,
but you're using a more lower scale

677
00:36:43,975 --> 00:36:48,369
device in order to access that data,
even though that's all local AI?

678
00:36:48,531 --> 00:36:52,498
So curious input, but also, hey,
what's running on your device?

679
00:36:53,122 --> 00:36:55,570
Scott Slater: Frank, before we go
there, you just got me thinking.

680
00:36:55,581 --> 00:36:56,736
You might be onto something.

681
00:36:56,745 --> 00:37:00,550
Were you talking about
RTX Spark devices in cars?

682
00:37:00,887 --> 00:37:02,720
This may be a whole new market.

683
00:37:02,773 --> 00:37:08,905
Like, I think 16-year-old Scott wanted
a car stereo system with two 10-inch

684
00:37:08,909 --> 00:37:13,687
subwoofers in my trunk in my car, but
maybe now this is a new market where

685
00:37:13,688 --> 00:37:16,898
you want a RTX Spark device in the car.

686
00:37:17,177 --> 00:37:17,688
I don't know

687
00:37:17,952 --> 00:37:20,980
Frankcx: it makes you like a Waymo
car, it has to rely on its own

688
00:37:20,988 --> 00:37:23,255
local AI in order to drive, right?

689
00:37:23,279 --> 00:37:23,731
It can't

690
00:37:23,842 --> 00:37:27,045
Jacob: All cars nowadays
have GPUs in them.

691
00:37:27,418 --> 00:37:31,857
Not all of-- Like, I mean, Tesla famously
a decade ago was like, "Hey, you know,

692
00:37:31,857 --> 00:37:35,961
you can plug in your Xbox to this
because it's got a, it, it's got a, a

693
00:37:35,964 --> 00:37:40,656
GPU that's an Xbox level GPU at that
time." But like all cars are taking this

694
00:37:40,670 --> 00:37:44,052
information in from all these inputs,
especially if they have any self-driving

695
00:37:44,619 --> 00:37:47,044
or, or assisted driving capabilities.

696
00:37:47,099 --> 00:37:47,276
Neil: I'm,

697
00:37:47,480 --> 00:37:47,887
Frankcx: Ja

698
00:37:47,926 --> 00:37:48,697
Neil: some work with

699
00:37:48,977 --> 00:37:49,468
Jacob: It's me, thank you

700
00:37:49,552 --> 00:37:54,428
Neil: services firms, and this notion
of does everyone need a $6,000 device?

701
00:37:55,070 --> 00:37:59,702
Probably not, but exploring
the usage of RTX, DGX Spark,

702
00:37:59,971 --> 00:38:02,112
and scheduling specific jobs.

703
00:38:02,112 --> 00:38:04,010
So imagine like tax preparation.

704
00:38:04,488 --> 00:38:07,168
Johnny runs all his
clients on Monday morning.

705
00:38:07,391 --> 00:38:11,550
Tuesday-- Suzy gets access to run
jobs for her clients Tuesday night.

706
00:38:11,750 --> 00:38:14,771
I've been doing some
testing with RTX Spark.

707
00:38:14,775 --> 00:38:19,255
I threw 64 jobs at it at the same
time, and was able to reach a

708
00:38:19,297 --> 00:38:23,529
peak aggregate throughput of over
1,000 tokens per second, right?

709
00:38:23,534 --> 00:38:28,541
So pretty incredible as you think
about this notion of a token factory.

710
00:38:28,857 --> 00:38:34,229
We've got token factories on desks,
and for folks that may be using this

711
00:38:34,243 --> 00:38:39,145
once a week, once a month, that edge
server appliance can meet those demands.

712
00:38:39,152 --> 00:38:41,102
So very bullish on this

713
00:38:42,496 --> 00:38:46,233
Frankcx: I know for the novices out there,
like a token is like three quarters of a

714
00:38:46,459 --> 00:38:47,839
Robert: Yeah

715
00:38:48,577 --> 00:38:51,621
Frankcx: almost a
thousand words per second.

716
00:38:52,366 --> 00:38:55,160
that's the type of power that
these RTX Spark devices will bring.

717
00:38:55,268 --> 00:38:56,374
So pretty amazing.

718
00:38:56,550 --> 00:38:56,892
Scott Slater: And,

719
00:38:56,970 --> 00:38:59,226
Robert: changed our hybrid
intelligence story now.

720
00:38:59,226 --> 00:39:01,126
It's not just device and cloud.

721
00:39:01,139 --> 00:39:05,882
Now there's an edge component, and
customers are really rethinking

722
00:39:05,887 --> 00:39:08,169
how they go about this and where
they're making their investments

723
00:39:09,285 --> 00:39:12,337
Scott Slater: I think those DGX
stations are super interesting.

724
00:39:12,349 --> 00:39:15,696
Like, you know, where can
those fit for companies?

725
00:39:15,696 --> 00:39:19,869
And I was having a conversation with a
customer last week about it, and, the

726
00:39:19,869 --> 00:39:25,676
conversation got to maybe they could
be needed besides for, an AI team or

727
00:39:25,678 --> 00:39:30,774
a developer team or a research lab or
something like that, but where companies

728
00:39:30,774 --> 00:39:33,053
are constrained with capacity, right?

729
00:39:33,053 --> 00:39:37,600
Like, if they're building their own AI
infrastructure and they have capacity

730
00:39:37,607 --> 00:39:42,447
constraints, or if they can't rent
capacity from third parties, you could

731
00:39:42,457 --> 00:39:46,694
drop one of these in those scenarios
and not slow down your business.

732
00:39:46,740 --> 00:39:50,480
I think it's gonna be really
interesting for companies to have them

733
00:39:50,481 --> 00:39:54,344
available for those scenarios, ' cause
we know capacity's not unlimited

734
00:39:55,517 --> 00:39:58,885
Frankcx: Okay, you're just getting
me super fired up here because

735
00:39:59,543 --> 00:40:03,444
I'm gonna push it on back to you,
Scott, of imagine a device, Scott,

736
00:40:03,449 --> 00:40:06,371
that your sellers could sell that
costs a hundred thousand dollars.

737
00:40:06,835 --> 00:40:09,805
How many sellers do not wanna
sell a device like that?

738
00:40:10,598 --> 00:40:13,971
the flip side would be is how many
Surface laptops do you have to sell

739
00:40:13,997 --> 00:40:18,032
to make that type of number where
you could be positioning a DGX

740
00:40:18,040 --> 00:40:19,942
station device at a hundred grand?

741
00:40:20,238 --> 00:40:26,304
But this device can run in a private
environment, a frontier model that

742
00:40:26,304 --> 00:40:28,164
is just as capable of anything.

743
00:40:28,174 --> 00:40:32,329
Like, you could run that
MIA or, you know, MAI

744
00:40:32,405 --> 00:40:32,981
Jacob: And may I?

745
00:40:33,270 --> 00:40:36,866
Frankcx: on that device, and I
would say that that is probably

746
00:40:36,900 --> 00:40:39,297
Pavan and, S-Satya's, dream.

747
00:40:39,676 --> 00:40:44,044
But Scott, I'm gonna vote for you to build
an incubation team within Windows that

748
00:40:44,044 --> 00:40:47,744
goes out and sells a hundred thousand
dollar device 'cause I don't think there's

749
00:40:47,744 --> 00:40:51,453
a seller in the world that wouldn't
love to get an order for a few of those

750
00:40:51,736 --> 00:40:52,766
Robert: Hey, sign me up.

751
00:40:52,766 --> 00:40:54,703
I'll, I'll sell the hell out of those.

752
00:40:54,972 --> 00:40:55,557
Scott Slater: There we go

753
00:40:55,615 --> 00:41:00,011
Robert: said before, the ROI for
that, 100 grand one time opposed

754
00:41:00,011 --> 00:41:02,620
to every month for a branch?

755
00:41:02,749 --> 00:41:03,495
Frankcx: can make money.

756
00:41:03,533 --> 00:41:03,947
I mean, this

757
00:41:04,324 --> 00:41:05,078
Robert: Absolutely

758
00:41:05,342 --> 00:41:06,727
Frankcx: that just serves up data.

759
00:41:06,753 --> 00:41:07,950
It actually makes

760
00:41:08,259 --> 00:41:08,586
Robert: Yeah.

761
00:41:08,795 --> 00:41:09,367
Frankcx: in that way.

762
00:41:09,377 --> 00:41:09,553
It

763
00:41:09,614 --> 00:41:10,025
Robert: it's…

764
00:41:10,294 --> 00:41:11,229
Frankcx: factory of its

765
00:41:11,374 --> 00:41:12,112
Robert: There you go

766
00:41:12,561 --> 00:41:13,361
Scott Slater: Teñido Frank

767
00:41:13,805 --> 00:41:16,177
Jacob: I heard a rumor there'll
be one in San Francisco next

768
00:41:16,177 --> 00:41:18,216
week at GitHub's Universe event.

769
00:41:18,274 --> 00:41:20,249
It's just a rumor, unconfirmed.

770
00:41:20,941 --> 00:41:21,983
San Francisco, if you're there

771
00:41:22,631 --> 00:41:28,049
Scott Slater: from a client seller,
that $100,000 may seem like a lot.

772
00:41:28,472 --> 00:41:33,558
From an infrastructure seller,
that $100,000 is not a lot, right?

773
00:41:33,638 --> 00:41:36,748
So anybody that sold
infrastructure and data centers,

774
00:41:37,034 --> 00:41:37,271
Frankcx: Sure

775
00:41:37,719 --> 00:41:39,200
Scott Slater: that's a
very reasonable amount

776
00:41:41,090 --> 00:41:41,961
Frankcx: All right.

777
00:41:42,280 --> 00:41:45,105
Well, I think, that kind of
brings us as a good point.

778
00:41:45,105 --> 00:41:48,551
I think we're all excited for
this DGX Station device, what it

779
00:41:48,553 --> 00:41:50,916
can offer in a Windows ecosystem.

780
00:41:50,947 --> 00:41:55,649
I think that Surface personally tees
it up with Surface Laptop Ultra and,

781
00:41:55,649 --> 00:41:57,271
the dev boxes that are out there.

782
00:41:57,272 --> 00:42:02,372
But I think as customers look at how
they can climb into these spaces and be

783
00:42:02,372 --> 00:42:06,303
able to move into infrastructure like
that, and even colo facilities that

784
00:42:06,305 --> 00:42:11,175
are housing, NVIDIA and CUDA-based,
architecture, I think this is a good

785
00:42:11,176 --> 00:42:12,937
time, to be able to position that.

786
00:42:13,738 --> 00:42:17,247
any closing words before we
close out this week's pod?

787
00:42:18,830 --> 00:42:20,033
Robert: Great to have Scott on.

788
00:42:20,034 --> 00:42:20,631
Well done.

789
00:42:20,641 --> 00:42:21,026
Thank you,

790
00:42:21,076 --> 00:42:21,835
Neil: Thank you, Scott.

791
00:42:21,893 --> 00:42:23,029
Bringing in the heavy hitters.

792
00:42:23,479 --> 00:42:24,800
Scott Slater: thank you for having me.

793
00:42:24,828 --> 00:42:26,256
And, just one final question.

794
00:42:26,256 --> 00:42:29,701
Is Neil IM gonna take us
out with some outro music?

795
00:42:31,259 --> 00:42:31,800
Frankcx: We will.

796
00:42:31,814 --> 00:42:36,837
We, Neil, I c- I will, send me some
DJ, you know, exit music, and we'll put

797
00:42:36,837 --> 00:42:38,800
that on the, this week's pod as our exit

798
00:42:39,160 --> 00:42:39,914
Neil: Sounds great

799
00:42:39,963 --> 00:42:43,862
Robert: know that at one point at
a Surface Airlift, Neil actually

800
00:42:43,979 --> 00:42:46,104
was my live DJ during a session?

801
00:42:46,969 --> 00:42:47,698
Neil: Absolutely.

802
00:42:47,704 --> 00:42:49,278
Helped launch my DJ career.

803
00:42:49,296 --> 00:42:51,780
Always, always remember
and appreciate that moment.

804
00:42:51,841 --> 00:42:55,555
and if I recall, I actually, provided
walk-up music for you as well, Frank.

805
00:42:55,600 --> 00:42:57,990
I think it was Fool's Gold
was the track, if I recall.

806
00:42:58,030 --> 00:42:58,301
Frankcx: Yeah,

807
00:42:58,658 --> 00:42:59,018
Neil: Yep

808
00:42:59,273 --> 00:43:02,590
Frankcx: and I think maybe that's the
music unless we get sued, that maybe we

809
00:43:02,788 --> 00:43:03,473
Neil: There we go.

810
00:43:05,600 --> 00:43:06,326
Jacob: Aaron Carter?

811
00:43:07,709 --> 00:43:08,241
Frankcx: All right.

812
00:43:08,277 --> 00:43:11,239
Well, thanks everyone, and,
we'll see you again on the

813
00:43:11,239 --> 00:43:13,238
next edition of The Local Host

814
00:43:14,135 --> 00:43:14,623
Neil: Thank you

815
00:43:14,885 --> 00:43:15,475
Jacob: Thanks everyone

816
00:43:18,432 --> 00:43:19,948
Frankcx: There's no place like home.