1 00:00:01,760 --> 00:00:03,759 Join us as we gather around the hedge, 2 00:00:03,759 --> 00:00:06,740 where we dig into technology, business, and culture 3 00:00:06,799 --> 00:00:09,139 with the finest minds in computer networking. 4 00:00:20,894 --> 00:00:24,195 Well, hello, Tom, again, and hello, Audrey, again. 5 00:00:25,134 --> 00:00:28,109 Hello, again. Hello, again. Third time's the charm. 6 00:00:28,510 --> 00:00:29,010 Yep. 7 00:00:29,390 --> 00:00:31,410 This is our third recording of the day. 8 00:00:31,550 --> 00:00:33,869 I might three d print, like, a mouth 9 00:00:33,869 --> 00:00:35,149 and put it back there like we were 10 00:00:35,149 --> 00:00:36,670 talking about last time. But it has to 11 00:00:36,670 --> 00:00:37,409 have teeth. 12 00:00:37,789 --> 00:00:38,690 Yeah. Yeah. 13 00:00:39,149 --> 00:00:40,909 Like Little Shop of Horrors. It has to 14 00:00:40,909 --> 00:00:42,850 have, like, teeth and be like, 15 00:00:47,174 --> 00:00:47,674 roar. 16 00:00:48,295 --> 00:00:50,454 Russ is doing sound effects now. I don't 17 00:00:50,454 --> 00:00:52,295 know what to think of that. Don't talk 18 00:00:52,295 --> 00:00:54,295 to Russ about spirit animals. Things start to 19 00:00:54,295 --> 00:00:54,795 happen. 20 00:00:59,390 --> 00:01:00,289 So and 21 00:01:00,670 --> 00:01:03,010 the third voice you hear there, of course, 22 00:01:03,390 --> 00:01:05,569 is Yvonne. Hey. Hello. 23 00:01:06,349 --> 00:01:07,629 I have to say I have a new 24 00:01:07,629 --> 00:01:08,129 plant. 25 00:01:09,310 --> 00:01:11,069 Okay. I don't know if you guys were 26 00:01:11,069 --> 00:01:13,010 around for the great HVAC, 27 00:01:13,415 --> 00:01:16,394 shed HVAC failure of twenty twenty six, but 28 00:01:16,694 --> 00:01:19,114 my different Vachea that was back there 29 00:01:19,734 --> 00:01:23,334 froze stiff. It was, like, 15 degrees out 30 00:01:23,334 --> 00:01:23,834 here. 31 00:01:24,295 --> 00:01:24,795 My 32 00:01:25,254 --> 00:01:26,474 HVAC quit. 33 00:01:27,319 --> 00:01:27,819 And, 34 00:01:28,599 --> 00:01:30,459 I've got to tell you that a 35 00:01:30,920 --> 00:01:33,719 a frozen stiff dip diphenbachia does not survive. 36 00:01:33,719 --> 00:01:35,399 I'm gonna show this. I know the rest 37 00:01:35,399 --> 00:01:36,920 of you can't see it, but I'm gonna 38 00:01:36,920 --> 00:01:39,640 show this to Russ and Tom. This is 39 00:01:39,640 --> 00:01:40,379 my other, 40 00:01:41,405 --> 00:01:43,825 peace lily. Oh. It actually 41 00:01:44,524 --> 00:01:46,844 survived the freeze. It was out here during 42 00:01:46,844 --> 00:01:48,284 the freeze, and I took it in the 43 00:01:48,284 --> 00:01:49,744 house and rehabbed it, 44 00:01:50,045 --> 00:01:52,064 and it's in good shape. So this time, 45 00:01:52,125 --> 00:01:54,045 we decided to get a peace lily back 46 00:01:54,045 --> 00:01:55,825 here, and maybe it will 47 00:01:56,880 --> 00:01:57,380 endure 48 00:01:58,400 --> 00:01:59,299 the temperature 49 00:02:00,079 --> 00:02:00,579 fluctuations 50 00:02:00,959 --> 00:02:03,359 that that happen out here in the grand 51 00:02:03,359 --> 00:02:04,959 old she shed. So I know that's more 52 00:02:04,959 --> 00:02:06,819 than you wanted to know about my houseplants, 53 00:02:06,959 --> 00:02:07,459 but 54 00:02:08,000 --> 00:02:08,900 there you go. 55 00:02:09,360 --> 00:02:11,300 No. It's all good. It's cool. Yeah. 56 00:02:11,935 --> 00:02:12,435 So 57 00:02:12,735 --> 00:02:14,415 we have a couple of topics to talk 58 00:02:14,415 --> 00:02:15,155 about today. 59 00:02:16,175 --> 00:02:18,514 We don't we're guestless. This is our roundtable 60 00:02:18,574 --> 00:02:19,074 episode. 61 00:02:19,615 --> 00:02:20,115 Guestless. 62 00:02:21,294 --> 00:02:22,655 I I don't know. That could be I 63 00:02:22,655 --> 00:02:23,474 don't know. Whatever. 64 00:02:23,939 --> 00:02:25,860 This is our roundtable episode. So, 65 00:02:26,500 --> 00:02:27,540 we'll start with 66 00:02:27,939 --> 00:02:29,479 there's an article in 67 00:02:30,099 --> 00:02:32,180 APNIC, and I just noticed this by George 68 00:02:32,180 --> 00:02:32,680 Michaelson, 69 00:02:33,139 --> 00:02:35,560 who's on the podcast on a regular basis. 70 00:02:36,715 --> 00:02:38,175 And he's talking about, 71 00:02:39,034 --> 00:02:41,455 why are we building all these data centers? 72 00:02:42,395 --> 00:02:44,235 And I think it's interesting, because I think 73 00:02:44,235 --> 00:02:46,474 that in IT, we kind of think we 74 00:02:46,474 --> 00:02:47,775 know the answer to that. 75 00:02:49,034 --> 00:02:49,534 Right? 76 00:02:50,189 --> 00:02:51,870 We think we think we know the answer 77 00:02:51,870 --> 00:02:54,370 because we need lots of processors for AI. 78 00:02:54,590 --> 00:02:57,150 We need lots of processors for all the 79 00:02:57,150 --> 00:02:59,010 stuff that we need processors for. 80 00:02:59,790 --> 00:03:01,490 But I think there's a lot of 81 00:03:02,205 --> 00:03:02,705 risk 82 00:03:03,004 --> 00:03:05,645 in this development as well and a lot 83 00:03:05,645 --> 00:03:06,224 of problems. 84 00:03:07,885 --> 00:03:08,705 For instance, 85 00:03:09,245 --> 00:03:10,764 what if in the end we actually don't 86 00:03:10,764 --> 00:03:12,944 need all these data centers to do this? 87 00:03:14,540 --> 00:03:16,800 What if someone really does create an LLM 88 00:03:17,020 --> 00:03:17,760 that doesn't 89 00:03:18,139 --> 00:03:18,639 need 90 00:03:19,020 --> 00:03:20,879 all as much processing power 91 00:03:21,260 --> 00:03:23,500 as we are trying to throw at these 92 00:03:23,500 --> 00:03:24,000 things? 93 00:03:24,700 --> 00:03:25,200 Or, 94 00:03:25,819 --> 00:03:28,425 you know, what if we find a way 95 00:03:28,425 --> 00:03:30,585 via power? Because everybody's very worried about power. 96 00:03:30,585 --> 00:03:31,865 I don't know if you've been reading up 97 00:03:31,865 --> 00:03:34,665 about ELS or whatever it is and co 98 00:03:34,665 --> 00:03:37,465 package topics and even co package copper. I'm 99 00:03:37,465 --> 00:03:39,064 doing a lot of research into those things 100 00:03:39,064 --> 00:03:40,925 right now, so that's what comes to mind 101 00:03:41,385 --> 00:03:41,885 about 102 00:03:42,419 --> 00:03:44,739 power savings. And maybe we don't need quite 103 00:03:44,739 --> 00:03:45,959 so much floor space 104 00:03:46,500 --> 00:03:48,840 for the compute power we even need. 105 00:03:49,379 --> 00:03:51,319 So I think there is some question 106 00:03:52,019 --> 00:03:52,519 there. 107 00:03:53,699 --> 00:03:56,180 I think I don't know. Tom, Yvonne, any 108 00:03:56,180 --> 00:03:58,155 any comments thus far? Because 109 00:03:58,955 --> 00:04:01,754 I remember I remember walking around when a 110 00:04:01,754 --> 00:04:03,675 company I worked for was doing an acquisition 111 00:04:03,675 --> 00:04:06,235 of a a data center shell in Salt 112 00:04:06,235 --> 00:04:07,134 Lake City, Utah. 113 00:04:07,435 --> 00:04:11,055 And the old century CenturyLink had, sold 114 00:04:11,360 --> 00:04:13,840 sold them this this data center. And it 115 00:04:13,840 --> 00:04:15,759 was it was hollowed out. It didn't have, 116 00:04:15,759 --> 00:04:17,439 like, finished white space and raised floor and 117 00:04:17,439 --> 00:04:19,120 stuff because they pulled all that stuff out. 118 00:04:19,120 --> 00:04:21,860 And I remember walking around. It's huge, gigantic 119 00:04:21,919 --> 00:04:22,419 building. 120 00:04:22,879 --> 00:04:24,639 It was a Coke bottling plant back in 121 00:04:24,639 --> 00:04:25,095 the day. 122 00:04:25,574 --> 00:04:27,975 And, I just remember walking around in there 123 00:04:27,975 --> 00:04:30,375 thinking, how are we ever gonna fill all 124 00:04:30,375 --> 00:04:31,814 of this? And I don't know what the 125 00:04:31,814 --> 00:04:33,334 state is now, but it's probably pretty close 126 00:04:33,334 --> 00:04:34,774 to full. And then that makes me think 127 00:04:34,774 --> 00:04:36,454 of other data centers I've been in where, 128 00:04:36,454 --> 00:04:38,235 you know, as the life cycle goes, 129 00:04:38,810 --> 00:04:40,889 lots of lots of racks of compute get 130 00:04:40,889 --> 00:04:42,730 condensed into one rack of compute, and you've 131 00:04:42,730 --> 00:04:43,970 got all this all this, 132 00:04:44,410 --> 00:04:46,490 all this floor space. And I'm not sure. 133 00:04:46,490 --> 00:04:48,750 I don't know. I'm sort of, waxing nostalgic 134 00:04:48,810 --> 00:04:50,490 for no purpose here, but this seems to 135 00:04:50,490 --> 00:04:53,264 have this is always happening. Like, this is 136 00:04:53,264 --> 00:04:55,444 always like, oh, we need all the space. 137 00:04:55,504 --> 00:04:57,345 Load it up. Oh, cleared out. It's all 138 00:04:57,345 --> 00:04:59,584 smaller now, and it seems like it'll go 139 00:04:59,584 --> 00:05:00,564 on probably forever. 140 00:05:02,064 --> 00:05:03,824 Yeah. I think there I think there is 141 00:05:03,824 --> 00:05:06,144 a cyclical nature to it. I think we're 142 00:05:06,144 --> 00:05:07,444 in the middle of 143 00:05:08,560 --> 00:05:09,939 an emerging technology, 144 00:05:10,879 --> 00:05:11,379 and 145 00:05:12,160 --> 00:05:13,300 we haven't gotten 146 00:05:13,680 --> 00:05:14,979 exactly to the optimization 147 00:05:15,279 --> 00:05:17,139 phase yet for that technology. 148 00:05:17,680 --> 00:05:20,180 Right? Like, for for example, when, 149 00:05:20,879 --> 00:05:23,699 servers and data centers were first a thing, 150 00:05:24,504 --> 00:05:25,964 We were entering 151 00:05:26,584 --> 00:05:28,125 this place where 152 00:05:29,464 --> 00:05:31,625 we were running out of floor space because 153 00:05:31,625 --> 00:05:33,485 every server was a physical box, 154 00:05:33,785 --> 00:05:35,704 and then we had virtualization. And then all 155 00:05:35,704 --> 00:05:36,845 of a sudden, like, 156 00:05:37,209 --> 00:05:37,709 this, 157 00:05:38,250 --> 00:05:40,729 you know, you could fit a 100 racks 158 00:05:40,729 --> 00:05:41,629 of servers 159 00:05:42,250 --> 00:05:44,430 in, like, a tenth of the space. 160 00:05:45,849 --> 00:05:48,490 I think I think, ultimately, we will see 161 00:05:48,490 --> 00:05:50,829 some more optimization around AI 162 00:05:51,370 --> 00:05:52,750 when we get there, 163 00:05:53,055 --> 00:05:55,154 but I don't think we're there yet. 164 00:05:55,455 --> 00:05:55,955 And 165 00:05:57,055 --> 00:05:58,355 depending on your workloads, 166 00:05:59,134 --> 00:05:59,795 you know, 167 00:06:01,694 --> 00:06:02,834 many, many, many 168 00:06:03,134 --> 00:06:06,014 of your data centers, especially your hyperscalers, are 169 00:06:06,014 --> 00:06:08,355 constantly battling capacity issues. 170 00:06:09,329 --> 00:06:10,949 And I even heard just today 171 00:06:11,810 --> 00:06:12,310 that, 172 00:06:12,769 --> 00:06:14,930 a customer had a twelve to eighteen month 173 00:06:14,930 --> 00:06:17,029 lead time for an interconnect 174 00:06:17,810 --> 00:06:19,669 Yep. Because they were struggling 175 00:06:20,290 --> 00:06:21,970 with and I and I don't know what 176 00:06:21,970 --> 00:06:23,634 the particular holdup was in this case. I 177 00:06:23,634 --> 00:06:24,834 don't know if it was an optic. I 178 00:06:24,834 --> 00:06:26,514 don't know if if it was just physical 179 00:06:26,514 --> 00:06:27,495 plant and construction. 180 00:06:27,954 --> 00:06:29,975 I don't know what the holdup was. But 181 00:06:30,834 --> 00:06:31,654 it it 182 00:06:32,354 --> 00:06:34,754 it's it's not I don't believe that we 183 00:06:34,754 --> 00:06:36,774 are building things to sit empty. 184 00:06:37,960 --> 00:06:39,660 That's not been my observation 185 00:06:39,960 --> 00:06:40,460 anyway. 186 00:06:42,120 --> 00:06:44,139 I I do think that there will 187 00:06:45,160 --> 00:06:48,360 be process improvements and consolidation that will you 188 00:06:48,360 --> 00:06:49,879 know, either we'll get more out of it 189 00:06:49,879 --> 00:06:51,685 or it'll render some of that space moot. 190 00:06:51,845 --> 00:06:53,205 But I don't think we're there right now. 191 00:06:53,205 --> 00:06:55,285 I think I think we need a lot 192 00:06:55,285 --> 00:06:57,285 more compute to run the workloads that we're 193 00:06:57,285 --> 00:06:58,105 running today. 194 00:06:58,965 --> 00:07:00,725 So it's interesting because this brings up the 195 00:07:00,725 --> 00:07:02,085 flip side of the problem. I'll give you 196 00:07:02,085 --> 00:07:03,605 a couple of quotes. I'm I'm doing a 197 00:07:03,605 --> 00:07:05,685 presentation in Anag, and these quotes come from 198 00:07:05,685 --> 00:07:06,939 that presentation. 199 00:07:07,399 --> 00:07:07,899 And 200 00:07:08,680 --> 00:07:11,399 the this is Corning. Somebody at Corning, one 201 00:07:11,399 --> 00:07:13,819 of the vice presidents at Corning, externally said, 202 00:07:14,439 --> 00:07:16,839 just this year, The United States will need 203 00:07:16,839 --> 00:07:21,785 to add 213,300,000 204 00:07:22,084 --> 00:07:23,064 fiber miles 205 00:07:24,404 --> 00:07:26,264 by 2029. 206 00:07:26,964 --> 00:07:30,725 213,000,000 207 00:07:30,725 --> 00:07:32,345 miles of new fiber 208 00:07:33,044 --> 00:07:34,985 within the next three years. 209 00:07:35,819 --> 00:07:39,759 Currently, The United States has about a 159,600,000 210 00:07:39,819 --> 00:07:40,639 fiber miles, 211 00:07:41,180 --> 00:07:46,479 and we need to increase by the 372,900,000. 212 00:07:47,259 --> 00:07:47,919 That is 213 00:07:48,824 --> 00:07:51,305 insane. And so if you're four x of 214 00:07:51,305 --> 00:07:53,144 what we have today when you do the 215 00:07:53,144 --> 00:07:55,144 when you do the math. Right? Yeah. You 216 00:07:55,144 --> 00:07:56,504 take the 100 we got now and the 217 00:07:56,504 --> 00:07:58,764 300 we need and yeah. That's 218 00:07:59,305 --> 00:08:02,740 in in four year. And that's interesting because 219 00:08:02,740 --> 00:08:04,279 that would have not been 220 00:08:04,740 --> 00:08:06,519 the constraint I would have chosen. 221 00:08:07,139 --> 00:08:10,199 Yeah. And there's another one. Yeah. Yeah. RVA 222 00:08:10,259 --> 00:08:12,339 LLC, which is a research company, and they're 223 00:08:12,339 --> 00:08:13,879 saying 92,000 224 00:08:14,019 --> 00:08:15,399 new fiber miles 225 00:08:15,860 --> 00:08:16,360 just 226 00:08:17,274 --> 00:08:18,254 for data center 227 00:08:18,555 --> 00:08:19,055 interconnect 228 00:08:20,235 --> 00:08:21,454 and scale across 229 00:08:22,954 --> 00:08:24,334 in the next year. 230 00:08:24,875 --> 00:08:26,954 That's interesting because it feels like it doesn't 231 00:08:26,954 --> 00:08:28,014 take into account 232 00:08:29,449 --> 00:08:31,149 multiplexing. Like, DWDM, 233 00:08:31,449 --> 00:08:32,649 you can get a lot out of a 234 00:08:32,649 --> 00:08:35,210 fiber. I sort of I'm skeptical about these 235 00:08:35,210 --> 00:08:37,610 numbers because, like, you can you can get 236 00:08:37,610 --> 00:08:39,289 a lot of capacity out of one pair 237 00:08:39,289 --> 00:08:40,590 of fiber tripling 238 00:08:41,049 --> 00:08:42,110 the fiber footprint. 239 00:08:42,634 --> 00:08:43,534 I don't know. 240 00:08:43,835 --> 00:08:46,235 I it seems as little far fetched to 241 00:08:46,235 --> 00:08:46,735 me. 242 00:08:47,434 --> 00:08:49,274 Well, but I think I think what's happening 243 00:08:49,274 --> 00:08:50,335 is is that 244 00:08:51,034 --> 00:08:54,174 data center builders are running into construction 245 00:08:54,554 --> 00:08:55,214 and power 246 00:08:56,360 --> 00:08:56,860 constraints. 247 00:08:57,399 --> 00:08:57,899 Right? 248 00:08:58,200 --> 00:08:59,879 There's only so large of a field you 249 00:08:59,879 --> 00:09:01,799 can put a data center in. And a 250 00:09:01,799 --> 00:09:03,799 lot of communities don't want data centers over 251 00:09:03,799 --> 00:09:06,539 a certain size. They're doing data center constraints. 252 00:09:06,600 --> 00:09:08,919 Right? Right. On the size and the and 253 00:09:08,919 --> 00:09:11,184 the power usage and stuff like that. And 254 00:09:11,184 --> 00:09:13,105 then there's just certain numbers of places you 255 00:09:13,105 --> 00:09:15,264 just can't get power. So some people are 256 00:09:15,264 --> 00:09:17,425 solving this. We'll just put out nukes. Right? 257 00:09:17,425 --> 00:09:19,264 We'll just put out little nuclear power plants, 258 00:09:19,264 --> 00:09:21,184 and that's a thing. And that's going to 259 00:09:21,184 --> 00:09:24,065 happen. People are doing it. Microsoft bought 3 260 00:09:24,065 --> 00:09:26,065 Mile Island. People think 3 Mile Island went 261 00:09:26,065 --> 00:09:28,580 away. It actually didn't. It's been operational all 262 00:09:28,580 --> 00:09:31,460 these years, and Microsoft bought the entire nuclear 263 00:09:31,460 --> 00:09:33,000 power plant so they could supply 264 00:09:33,460 --> 00:09:35,160 their electrical needs for, 265 00:09:35,940 --> 00:09:39,080 a massive build out of compute and stuff. 266 00:09:39,620 --> 00:09:40,120 And, 267 00:09:40,899 --> 00:09:41,384 I 268 00:09:41,785 --> 00:09:44,105 the thing is is that when you do 269 00:09:44,105 --> 00:09:46,985 that, when you start segregating things, you actually 270 00:09:46,985 --> 00:09:48,445 need to split your 271 00:09:49,065 --> 00:09:51,804 fabric across multiple buildings regionally. 272 00:09:52,665 --> 00:09:54,825 And you like, if you're actually gonna split 273 00:09:54,825 --> 00:09:56,860 a fabric, and I'm not not not not 274 00:09:56,860 --> 00:09:58,240 like traditional DCI, 275 00:09:58,620 --> 00:10:01,120 but, like, I'm gonna take two data centers 276 00:10:01,259 --> 00:10:03,420 with five stage fabrics or four of them, 277 00:10:03,420 --> 00:10:05,179 and I'm gonna make them into a nine 278 00:10:05,179 --> 00:10:06,960 stage regional data center. 279 00:10:08,214 --> 00:10:10,855 Like, the what the amount and you can 280 00:10:10,855 --> 00:10:12,694 you can do DWDM, but there's only a 281 00:10:12,694 --> 00:10:14,774 certain number of channels you're gonna get. And 282 00:10:14,774 --> 00:10:16,794 you're looking at 64 cross connects 283 00:10:17,334 --> 00:10:20,134 or, you know, a 120, whatever it is, 284 00:10:20,134 --> 00:10:22,429 however many cross connects this is that you 285 00:10:22,429 --> 00:10:24,370 start looking at the scale of these things. 286 00:10:24,589 --> 00:10:26,769 And, like, yeah, that's this is a problem. 287 00:10:27,230 --> 00:10:28,909 And so this is, like, the flip side. 288 00:10:28,909 --> 00:10:31,329 Right? Where are we getting this stuff? 289 00:10:32,669 --> 00:10:33,649 Like, seriously? 290 00:10:35,654 --> 00:10:37,274 Fuck. Who's who's building 291 00:10:38,054 --> 00:10:39,514 300,000,000 292 00:10:39,735 --> 00:10:41,914 miles of fiber optic cables 293 00:10:42,695 --> 00:10:44,154 or whatever it is 294 00:10:44,695 --> 00:10:46,315 in the next three years? 295 00:10:47,110 --> 00:10:49,529 Sure. Who who's creating that? 296 00:10:50,309 --> 00:10:52,629 Where are those plants? And those are truck 297 00:10:52,629 --> 00:10:54,570 rolls. Right? These are very 298 00:10:55,190 --> 00:10:56,730 physical. I mean, you're you're 299 00:10:57,590 --> 00:11:00,730 you're burying cable. You're you're rolling a trenching 300 00:11:00,790 --> 00:11:03,254 truck. You're it it is a it is 301 00:11:03,254 --> 00:11:04,075 a heavyweight 302 00:11:07,654 --> 00:11:08,154 endeavor 303 00:11:08,855 --> 00:11:10,615 Yeah. To to run fiber like that. You 304 00:11:10,615 --> 00:11:12,375 know, we're not we're not even talking about 305 00:11:12,375 --> 00:11:12,875 constructing 306 00:11:13,335 --> 00:11:15,654 a data center, a building in a place. 307 00:11:15,654 --> 00:11:16,794 We're talking about 308 00:11:17,220 --> 00:11:17,720 significant 309 00:11:18,179 --> 00:11:19,639 truck roles and infrastructure 310 00:11:20,019 --> 00:11:22,100 and and all of that. Yeah. It's it 311 00:11:22,100 --> 00:11:24,179 it I'm with Tom. It kinda boggles the 312 00:11:24,179 --> 00:11:27,159 mind a bit. Like, that that's like, 313 00:11:28,419 --> 00:11:30,259 you know, 1999 314 00:11:30,259 --> 00:11:33,115 kind of filled out numbers. You know? It 315 00:11:33,115 --> 00:11:34,254 it feels like 316 00:11:34,634 --> 00:11:36,335 that. Yeah. Yeah. 317 00:11:36,795 --> 00:11:38,955 Which which kinda drives more of the the 318 00:11:38,955 --> 00:11:41,035 talk of, you know, can we actually do 319 00:11:41,035 --> 00:11:43,054 this? What happens if we don't do this? 320 00:11:43,195 --> 00:11:46,014 Right? Does this does this hurt the bubble? 321 00:11:46,320 --> 00:11:47,779 Does this hurt the deployment 322 00:11:48,240 --> 00:11:50,399 in a serious way? Because we just physically 323 00:11:50,399 --> 00:11:52,480 can't build all this stuff. And and by 324 00:11:52,480 --> 00:11:54,259 the way, a funny story from this week, 325 00:11:55,440 --> 00:11:56,740 I think it's funny anyway, 326 00:11:57,200 --> 00:11:59,279 is that there used to be a shoe 327 00:11:59,279 --> 00:12:00,500 company called Allbirds. 328 00:12:02,154 --> 00:12:02,654 Yeah. 329 00:12:04,154 --> 00:12:05,774 They are now an AI company. 330 00:12:06,154 --> 00:12:07,835 I thought we were gonna talk about AI 331 00:12:07,835 --> 00:12:10,175 today. I know. We're trying not to. 332 00:12:10,875 --> 00:12:12,555 But it is but it is rather humorous 333 00:12:12,555 --> 00:12:14,254 to me. Right? Like, 334 00:12:14,820 --> 00:12:16,100 I don't know where all this stuff is 335 00:12:16,100 --> 00:12:16,920 coming from. 336 00:12:17,620 --> 00:12:19,139 And I don't know how we're gonna build 337 00:12:19,139 --> 00:12:20,740 all this data center space. I just don't 338 00:12:20,740 --> 00:12:22,580 know how we're gonna power it. I don't 339 00:12:22,580 --> 00:12:23,940 know how we're gonna water it. I don't 340 00:12:23,940 --> 00:12:25,460 know how I just don't know how we're 341 00:12:25,460 --> 00:12:28,205 gonna do it. And so that is that's 342 00:12:28,205 --> 00:12:29,904 like an issue to me. 343 00:12:31,804 --> 00:12:33,264 And communities are mad. 344 00:12:33,965 --> 00:12:36,365 Like, I've seen videos of people who live 345 00:12:36,365 --> 00:12:38,365 next to a data center that went up, 346 00:12:38,365 --> 00:12:40,044 and people say, well, it's not that loud 347 00:12:40,044 --> 00:12:42,144 over. And they just put their video camera 348 00:12:42,419 --> 00:12:43,159 on their 349 00:12:43,620 --> 00:12:45,700 porch. And you can't talk to anybody, each 350 00:12:45,700 --> 00:12:46,759 other in the yard 351 00:12:47,459 --> 00:12:49,000 because of the chilling towers 352 00:12:49,379 --> 00:12:51,079 and because of the the power. 353 00:12:51,620 --> 00:12:53,819 And it's like, I don't know. Like, I 354 00:12:53,860 --> 00:12:55,220 if I live next door to that and 355 00:12:55,220 --> 00:12:56,659 I couldn't go outside and play in the 356 00:12:56,659 --> 00:12:58,394 yard, I think I'd be pretty upset. 357 00:13:00,875 --> 00:13:02,095 So I don't know. 358 00:13:02,634 --> 00:13:04,315 It will there be a reaction to this? 359 00:13:04,315 --> 00:13:05,375 Will it slow down? 360 00:13:05,915 --> 00:13:07,434 So I don't know. Anything else on that 361 00:13:07,434 --> 00:13:08,335 one, Yvonne? 362 00:13:10,235 --> 00:13:11,835 I do believe that in the long term, 363 00:13:11,835 --> 00:13:13,649 we'll solve the problems. When 364 00:13:15,009 --> 00:13:17,089 I I struggle sometimes to understand the pushback 365 00:13:17,089 --> 00:13:18,929 of data centers because I live in a 366 00:13:18,929 --> 00:13:19,829 world where, 367 00:13:20,610 --> 00:13:22,950 in my state, we're known for auto manufacturing. 368 00:13:23,730 --> 00:13:24,789 And I think, man, 369 00:13:25,464 --> 00:13:28,024 a data center is nowhere near as disruptive 370 00:13:28,024 --> 00:13:29,705 as a car plant. You know what I 371 00:13:29,705 --> 00:13:31,784 mean? Like and so I have a I 372 00:13:31,784 --> 00:13:33,464 I, you know, I kinda struggle with some 373 00:13:33,464 --> 00:13:34,205 of the pushback, 374 00:13:35,304 --> 00:13:37,544 to to data centers and, you know, I've 375 00:13:37,544 --> 00:13:40,100 toured some big data centers that weren't loud 376 00:13:40,100 --> 00:13:40,759 like that. 377 00:13:42,100 --> 00:13:43,379 So I, you know, I I don't know 378 00:13:43,379 --> 00:13:45,139 what to make of all that pushback. I 379 00:13:45,139 --> 00:13:46,120 will say that. 380 00:13:47,300 --> 00:13:48,279 But but, 381 00:13:49,540 --> 00:13:51,860 yeah, I'm I'm like Tom. I'm flabbergasted by 382 00:13:51,860 --> 00:13:55,347 that much fiber being needed Yeah. For the 383 00:13:55,347 --> 00:13:58,002 next three years. It just seems crazy. I 384 00:13:58,002 --> 00:14:01,320 mean, there's just I don't know. Alright. So 385 00:14:01,320 --> 00:14:03,975 the second thing we're gonna talk about is 386 00:14:03,975 --> 00:14:06,629 the foreign made router ban in The US, 387 00:14:07,110 --> 00:14:08,410 which is fascinating 388 00:14:09,429 --> 00:14:11,669 because I'm not entirely certain what to make 389 00:14:11,669 --> 00:14:13,289 of it from a policy perspective. 390 00:14:14,070 --> 00:14:17,269 Right? Is this really security or is this 391 00:14:17,269 --> 00:14:17,769 really 392 00:14:18,884 --> 00:14:19,625 rent seeking 393 00:14:20,485 --> 00:14:22,725 in some which way or form? And I 394 00:14:22,725 --> 00:14:26,264 don't know. I've not seen enough 395 00:14:26,965 --> 00:14:28,345 about it to know 396 00:14:28,804 --> 00:14:30,965 what I think. I know the FCC has 397 00:14:30,965 --> 00:14:33,785 banned foreign made routers based on 398 00:14:35,299 --> 00:14:36,519 security concerns. 399 00:14:37,700 --> 00:14:40,120 And and in some ways, I actually honestly, 400 00:14:40,259 --> 00:14:42,759 in some ways, I don't actually blame them. 401 00:14:43,379 --> 00:14:45,879 Because many of these devices do have backdoors. 402 00:14:46,340 --> 00:14:49,399 People on the Internet have done put 403 00:14:49,754 --> 00:14:51,934 packet sniffers behind the home router, 404 00:14:52,715 --> 00:14:55,274 and just there's the amount of data that 405 00:14:55,274 --> 00:14:57,295 gets shipped to weird IP addresses, 406 00:14:57,754 --> 00:15:00,095 and nobody really knows what it's doing, 407 00:15:00,634 --> 00:15:03,134 is it's a pretty incredible amount of data. 408 00:15:03,274 --> 00:15:05,054 And so I I kinda get, 409 00:15:06,019 --> 00:15:08,579 and the possibility of their backdoors and stuff 410 00:15:08,579 --> 00:15:10,659 like that. So I kinda get all of 411 00:15:10,659 --> 00:15:11,159 that. 412 00:15:12,819 --> 00:15:14,100 So I don't know. What do you think, 413 00:15:14,100 --> 00:15:14,600 Tom? 414 00:15:15,220 --> 00:15:17,299 Any I, I bought a I remember I 415 00:15:17,299 --> 00:15:18,519 bought a super cheap 416 00:15:19,605 --> 00:15:20,105 network, 417 00:15:20,884 --> 00:15:21,384 connected 418 00:15:21,845 --> 00:15:22,345 PDU 419 00:15:22,644 --> 00:15:24,404 for my home lab, like, I don't know, 420 00:15:24,404 --> 00:15:26,404 ten years ago or something. Had three outlets 421 00:15:26,404 --> 00:15:28,404 on it, and you could and an Ethernet 422 00:15:28,404 --> 00:15:30,804 port and you could remotely power cycle. And 423 00:15:30,804 --> 00:15:32,940 then so I was just for fun. I, 424 00:15:33,340 --> 00:15:35,420 I I captured the packets coming off of 425 00:15:35,420 --> 00:15:36,720 that thing going outbound. 426 00:15:37,179 --> 00:15:39,100 And it was pretty it was pretty hilarious. 427 00:15:39,100 --> 00:15:41,899 It was a Chinese, device and and like 428 00:15:41,899 --> 00:15:44,059 several IPs in China. And what was it 429 00:15:44,059 --> 00:15:45,759 sending? I don't know. It was all HTTPS, 430 00:15:46,220 --> 00:15:48,355 like, but they were pretty big packets. I'm 431 00:15:48,355 --> 00:15:50,535 like, what are you saying about me 432 00:15:51,154 --> 00:15:53,554 and my and my my network and my 433 00:15:53,554 --> 00:15:56,434 power? And Yeah. So I I sort of, 434 00:15:56,434 --> 00:15:57,335 you know, I 435 00:15:57,794 --> 00:15:59,875 I sympathize. I think, like you, I think 436 00:15:59,875 --> 00:16:01,894 I'm like, okay. There there probably are 437 00:16:03,679 --> 00:16:05,139 legitimate security concerns. 438 00:16:06,079 --> 00:16:08,480 It seems like an interesting way to solve 439 00:16:08,480 --> 00:16:10,079 the problem, but are there other better ways? 440 00:16:10,079 --> 00:16:11,679 I don't know. I I don't spend too 441 00:16:11,679 --> 00:16:14,100 much time thinking about public policy and telecom, 442 00:16:14,240 --> 00:16:14,740 but 443 00:16:15,120 --> 00:16:15,620 Yeah. 444 00:16:16,024 --> 00:16:17,785 It it may be the only hammer they 445 00:16:17,785 --> 00:16:19,465 had to solve this problem. I have no 446 00:16:19,465 --> 00:16:19,965 idea. 447 00:16:20,585 --> 00:16:23,945 So, Yvonne, thoughts? You well, I mean, so 448 00:16:23,945 --> 00:16:26,504 we we we all remember, like, the DDoS 449 00:16:26,504 --> 00:16:28,605 attacks that were sparked by botnets, 450 00:16:29,450 --> 00:16:32,250 botnet armies that attacked, you know, either IoT 451 00:16:32,250 --> 00:16:35,149 devices or consumer grade routers that had, 452 00:16:35,769 --> 00:16:37,789 and and some of some of that 453 00:16:38,410 --> 00:16:38,990 is is 454 00:16:39,529 --> 00:16:41,850 you know, it's it's hard to blame the 455 00:16:41,850 --> 00:16:42,350 consumer 456 00:16:43,129 --> 00:16:44,509 for not changing their 457 00:16:44,985 --> 00:16:48,024 admin admin username password when they are, you 458 00:16:48,024 --> 00:16:48,924 know, a nontechnical 459 00:16:49,464 --> 00:16:51,245 consumer. Like so, 460 00:16:51,784 --> 00:16:53,804 I, you know, I think it's clear 461 00:16:54,585 --> 00:16:58,125 that there's absolutely has been some risk and, 462 00:16:59,299 --> 00:17:00,279 some misuse 463 00:17:00,659 --> 00:17:01,399 of those, 464 00:17:02,419 --> 00:17:03,480 consumer grade 465 00:17:03,779 --> 00:17:04,839 routers. Now, 466 00:17:05,299 --> 00:17:06,980 you know, we could ask the question, does 467 00:17:06,980 --> 00:17:09,539 does this actually solve the problem? I don't 468 00:17:09,539 --> 00:17:12,259 know. Is some of it security theater? I 469 00:17:12,259 --> 00:17:14,315 I feel like there's a lot of theater 470 00:17:14,315 --> 00:17:15,454 going on these days 471 00:17:15,914 --> 00:17:18,154 in the public policy space, so I I 472 00:17:18,154 --> 00:17:19,674 I wouldn't put it past that. But, I 473 00:17:19,674 --> 00:17:21,295 mean, I do think there was a 474 00:17:21,674 --> 00:17:22,174 legit 475 00:17:23,035 --> 00:17:23,535 problem. 476 00:17:25,595 --> 00:17:27,194 I do think we've solved it with a 477 00:17:27,194 --> 00:17:28,894 pretty blunt tool that 478 00:17:30,549 --> 00:17:33,130 is is not gonna completely solve the problem. 479 00:17:33,269 --> 00:17:33,769 But, 480 00:17:34,389 --> 00:17:37,190 you know, it it is gonna create some 481 00:17:37,190 --> 00:17:38,889 local markets, I think. 482 00:17:40,630 --> 00:17:42,389 You know, and and and I think it's 483 00:17:42,389 --> 00:17:44,484 also you know, we should be really clear 484 00:17:44,484 --> 00:17:47,365 this is consumer grade gear. This is not 485 00:17:47,365 --> 00:17:50,265 our enterprise corporate gear that we're talking about. 486 00:17:50,644 --> 00:17:53,044 Yeah. But, yeah, I I, you know, I 487 00:17:53,044 --> 00:17:54,025 think it's interesting. 488 00:17:55,044 --> 00:17:56,325 I I see the article. I don't know 489 00:17:56,325 --> 00:17:57,765 if it's been long enough for there to 490 00:17:57,765 --> 00:17:58,744 be any documented, 491 00:17:59,740 --> 00:18:01,659 you know, what are the impacts been if 492 00:18:01,659 --> 00:18:04,299 there are downstream impacts. Are there, you know, 493 00:18:04,299 --> 00:18:05,200 are there now 494 00:18:05,500 --> 00:18:07,759 is there a supply chain problem with consumer 495 00:18:07,819 --> 00:18:10,539 grade networking gear? You know? I I don't 496 00:18:10,539 --> 00:18:12,779 know any of those things, but I think 497 00:18:12,779 --> 00:18:13,099 it's, 498 00:18:13,984 --> 00:18:15,744 I mean, I understand why we did it, 499 00:18:15,744 --> 00:18:18,884 and I think, you know, the security implications 500 00:18:19,105 --> 00:18:22,224 are real, and we've seen them. Many of 501 00:18:22,224 --> 00:18:24,545 us have been on the, troubleshooting end of 502 00:18:24,545 --> 00:18:26,945 them. Yeah. Whether or not there was another 503 00:18:26,945 --> 00:18:29,205 solution, I have no idea. Whether or not, 504 00:18:29,299 --> 00:18:31,460 you know, what other factors came into play, 505 00:18:31,460 --> 00:18:32,900 but I agree with you. There there was 506 00:18:32,900 --> 00:18:34,900 a there is a problem there. I think 507 00:18:34,900 --> 00:18:36,740 one of the positive things I've seen out 508 00:18:36,740 --> 00:18:37,640 of this interestingly 509 00:18:38,099 --> 00:18:39,779 is I've seen a lot of videos of 510 00:18:39,779 --> 00:18:40,599 people saying, 511 00:18:40,900 --> 00:18:41,559 you know, 512 00:18:42,099 --> 00:18:44,839 just stop buying consumer grade routers. 513 00:18:45,674 --> 00:18:48,174 You can just go buy a Raspberry Pi 514 00:18:48,714 --> 00:18:50,494 or a little used box 515 00:18:51,115 --> 00:18:53,694 and make yourself make it into a router. 516 00:18:53,755 --> 00:18:56,014 Now that doesn't solve the Wi Fi problem. 517 00:18:56,154 --> 00:18:58,734 Right? That only solves the gateway problem, 518 00:18:59,269 --> 00:19:01,130 but you actually get more control 519 00:19:01,669 --> 00:19:02,649 over the device 520 00:19:03,269 --> 00:19:05,349 because there are open source packages you can 521 00:19:05,349 --> 00:19:07,669 download. You can use a Ubuntu or Debian 522 00:19:07,669 --> 00:19:09,829 or whatever and put it on a a 523 00:19:09,829 --> 00:19:13,210 lightweight version or or even, Apache or whatever. 524 00:19:13,275 --> 00:19:15,674 Put it on a lightweight not Apache, whatever 525 00:19:15,674 --> 00:19:17,934 it is. Anyway, and put it on a 526 00:19:18,394 --> 00:19:20,634 on a Raspberry Pi and load up a 527 00:19:20,634 --> 00:19:21,134 lightweight 528 00:19:21,914 --> 00:19:23,775 software package that does routing. 529 00:19:24,394 --> 00:19:24,894 And, 530 00:19:25,480 --> 00:19:27,740 like, there's not many routes on these things. 531 00:19:28,200 --> 00:19:29,500 There's no routing protocols. 532 00:19:30,279 --> 00:19:32,519 Right? It's just it's just a few static 533 00:19:32,519 --> 00:19:34,679 routes. It's a net box and a static 534 00:19:34,679 --> 00:19:36,440 routes. It's pretty much it. Yeah. Pretty much 535 00:19:36,440 --> 00:19:38,265 it. It's all it is. That is such 536 00:19:38,265 --> 00:19:38,924 an engineer's 537 00:19:39,305 --> 00:19:40,285 answer, though. 538 00:19:40,825 --> 00:19:42,845 I'm trying to imagine 539 00:19:43,785 --> 00:19:47,305 giving that answer to my parents who are 540 00:19:47,305 --> 00:19:49,724 deeply nontechnical and in their seventies. 541 00:19:50,700 --> 00:19:52,400 Right? And saying, hey. 542 00:19:52,940 --> 00:19:54,240 Just go on Amazon, 543 00:19:54,539 --> 00:19:56,240 order this Raspberry Pi, 544 00:19:57,660 --> 00:19:58,160 and 545 00:19:58,619 --> 00:20:00,079 do these 10 steps. 546 00:20:00,380 --> 00:20:03,404 Right? It it it's it's just not Just 547 00:20:03,404 --> 00:20:06,044 just configure IP tables, not a big deal, 548 00:20:06,044 --> 00:20:08,924 and choose from 50,000 Linux distros. You can 549 00:20:08,924 --> 00:20:10,144 do it. It's you know? 550 00:20:10,684 --> 00:20:12,284 Yeah. Yeah. I don't know. I think I 551 00:20:12,284 --> 00:20:13,884 think it's time for a trip to your 552 00:20:13,884 --> 00:20:15,724 parents' house, Yvonne. I think that's what you're 553 00:20:15,724 --> 00:20:16,224 saying. 554 00:20:16,950 --> 00:20:19,210 I'm gonna give them your cell phone number. 555 00:20:22,789 --> 00:20:24,710 I mean, I don't know. You know? But 556 00:20:24,710 --> 00:20:26,950 that is but that is an issue. Yeah. 557 00:20:26,950 --> 00:20:28,470 The the joke in my world is if 558 00:20:28,470 --> 00:20:30,869 my phone rings at 07:30 or at 10:30PM 559 00:20:30,869 --> 00:20:31,929 and it's my mother, 560 00:20:32,305 --> 00:20:34,865 either somebody in the family is critically ill 561 00:20:34,865 --> 00:20:37,265 or she's having a technology problem. It is 562 00:20:37,265 --> 00:20:38,484 one of those two things. 563 00:20:39,664 --> 00:20:41,105 And and I don't think I've ever been 564 00:20:41,105 --> 00:20:43,045 proven wrong there. So yeah. 565 00:20:43,424 --> 00:20:45,125 So the nice thing is my mother 566 00:20:45,664 --> 00:20:46,164 is 567 00:20:47,029 --> 00:20:49,509 not hardware savvy per se. Like, she can't 568 00:20:49,509 --> 00:20:51,109 tell you what the connectors are and stuff 569 00:20:51,109 --> 00:20:53,190 like that. But once you get the hardware 570 00:20:53,190 --> 00:20:54,649 up and running, get it booted, 571 00:20:55,429 --> 00:20:56,329 she's fine. 572 00:20:56,710 --> 00:20:58,789 Because she used to code bash strips and 573 00:20:58,789 --> 00:21:01,109 scrap bash scripts, and she knows Ubuntu and 574 00:21:01,109 --> 00:21:03,795 Debian. And, you know, she actually used to 575 00:21:03,795 --> 00:21:06,994 do lightweight shell scripting and stuff. And she 576 00:21:06,994 --> 00:21:09,075 did a lot of accounting work on Linux 577 00:21:09,075 --> 00:21:11,315 systems. And so she's you get her to 578 00:21:11,315 --> 00:21:13,335 the boot prompt, and she's like, I'm happy. 579 00:21:13,474 --> 00:21:15,174 You know? Yeah. She doesn't know 580 00:21:15,640 --> 00:21:17,099 HDMI versus VGA, 581 00:21:17,480 --> 00:21:19,819 right, any longer because all that has changed 582 00:21:20,359 --> 00:21:23,240 since she was in the doing anything with 583 00:21:23,240 --> 00:21:23,740 it. 584 00:21:24,119 --> 00:21:26,359 But yeah. It it it took my parents 585 00:21:26,359 --> 00:21:27,799 a while to come to the conclusion that 586 00:21:27,799 --> 00:21:29,880 I wasn't throwing my career away with these 587 00:21:29,880 --> 00:21:30,779 silly computers. 588 00:21:31,214 --> 00:21:34,494 So yeah. Yeah. They they finally got there, 589 00:21:34,494 --> 00:21:35,714 but it took a while. 590 00:21:37,775 --> 00:21:39,294 Yeah. I don't know about computers. I don't 591 00:21:39,294 --> 00:21:40,815 know if there's a future in any of 592 00:21:40,815 --> 00:21:42,115 this network engineering, 593 00:21:42,494 --> 00:21:42,994 computers. 594 00:21:43,855 --> 00:21:44,355 Now 595 00:21:44,750 --> 00:21:46,590 now robots and AI are gonna take the 596 00:21:46,590 --> 00:21:48,190 whole thing over, and we're just not even 597 00:21:48,190 --> 00:21:50,590 gonna know. And computers have nothing to do 598 00:21:50,590 --> 00:21:51,970 with those things. So 599 00:21:54,590 --> 00:21:56,690 I'm I'm I'm waiting to 600 00:21:57,684 --> 00:22:00,265 buy my own robot army. I am completely 601 00:22:00,325 --> 00:22:01,625 on board with that idea. 602 00:22:04,085 --> 00:22:06,164 Oh my goodness. I did see by the 603 00:22:06,164 --> 00:22:06,664 cloud. 604 00:22:07,285 --> 00:22:09,365 Yeah. I did see somebody the other day 605 00:22:09,365 --> 00:22:11,785 who was doing sheep herding with drones. 606 00:22:12,849 --> 00:22:15,089 She's quite clever. I mean, I prefer the 607 00:22:15,089 --> 00:22:17,250 dogs because the dogs are cuter, you know, 608 00:22:17,250 --> 00:22:20,769 and they're fun and everything. Some companionship and 609 00:22:20,769 --> 00:22:23,250 some emotional support and all of that. You 610 00:22:23,250 --> 00:22:25,109 know? Exactly. But okay. 611 00:22:25,484 --> 00:22:27,265 I guess what if it floats your boat? 612 00:22:27,484 --> 00:22:29,724 My husband has said if you get a 613 00:22:29,724 --> 00:22:32,365 robot, you cannot have it working while we'll 614 00:22:32,365 --> 00:22:33,585 sleep while we sleep 615 00:22:34,285 --> 00:22:36,045 because I will wake up in the middle 616 00:22:36,045 --> 00:22:37,825 of the night and do it harm. 617 00:22:38,779 --> 00:22:40,640 But if if I see a humanoid, 618 00:22:41,740 --> 00:22:42,480 you know, 619 00:22:43,579 --> 00:22:46,539 option moving around our house, I'm going I'm 620 00:22:46,539 --> 00:22:48,700 going to strike first and ask questions later. 621 00:22:48,700 --> 00:22:50,079 And I'm like, fair enough. 622 00:22:53,244 --> 00:22:55,325 Okay. So our third article is a little 623 00:22:55,325 --> 00:22:56,464 more comic relief. 624 00:22:57,565 --> 00:22:58,544 So I remember 625 00:22:58,924 --> 00:23:00,865 now this this goes back many years 626 00:23:01,244 --> 00:23:02,944 when cloud first came out, 627 00:23:03,325 --> 00:23:06,144 and I was working with a large enterprise, 628 00:23:06,444 --> 00:23:07,184 quote, unquote, 629 00:23:07,565 --> 00:23:08,544 large operator, 630 00:23:10,180 --> 00:23:11,559 non provider operator. 631 00:23:12,660 --> 00:23:14,279 And their vice president 632 00:23:14,580 --> 00:23:16,500 had said, come back off of some trip 633 00:23:16,500 --> 00:23:18,759 and said, we had all these presentations 634 00:23:19,059 --> 00:23:19,880 on cloud. 635 00:23:20,259 --> 00:23:22,660 What is our strategy for cloud? What are 636 00:23:22,660 --> 00:23:24,360 we doing for cloud? 637 00:23:25,424 --> 00:23:27,664 And so they went into all their network 638 00:23:27,664 --> 00:23:28,164 diagrams 639 00:23:29,184 --> 00:23:31,825 and renamed their data centers from data center 640 00:23:31,825 --> 00:23:33,284 one five whatever 641 00:23:33,744 --> 00:23:34,244 whatever 642 00:23:34,625 --> 00:23:35,365 to cloud 643 00:23:35,744 --> 00:23:38,404 with exactly the same numbers after them. 644 00:23:38,799 --> 00:23:40,740 Okay. And they said, 645 00:23:41,759 --> 00:23:43,220 we've already deployed it. 646 00:23:43,920 --> 00:23:46,160 Like, we don't we don't know we don't 647 00:23:46,160 --> 00:23:48,559 know what you're talking about. You know? You 648 00:23:48,559 --> 00:23:51,039 went to some conference about this. We've had 649 00:23:51,039 --> 00:23:53,294 this deployed for years. Like, I have no 650 00:23:53,294 --> 00:23:55,214 idea what you're talking about. We're ahead of 651 00:23:55,214 --> 00:23:55,875 the curve. 652 00:23:59,294 --> 00:23:59,794 Nice. 653 00:24:02,014 --> 00:24:03,774 Are you allowed to laugh at that joke, 654 00:24:03,774 --> 00:24:04,274 Yvonne? 655 00:24:05,054 --> 00:24:06,894 I can laugh at anything I want to. 656 00:24:06,894 --> 00:24:08,034 Thank you very much. 657 00:24:08,460 --> 00:24:08,960 So 658 00:24:10,940 --> 00:24:11,440 so 659 00:24:11,819 --> 00:24:13,980 yeah. So this article in the register, the 660 00:24:13,980 --> 00:24:16,240 third one we picked picked up for today, 661 00:24:16,460 --> 00:24:16,960 is 662 00:24:17,500 --> 00:24:18,000 about 663 00:24:20,380 --> 00:24:23,099 proving or showing your your managers or your 664 00:24:23,099 --> 00:24:25,904 upper level management that you have deployed AI 665 00:24:25,904 --> 00:24:27,125 in your network operations. 666 00:24:29,025 --> 00:24:31,025 And it's and it's very it's a brilliant 667 00:24:31,184 --> 00:24:32,644 I think it's absolutely brilliant. 668 00:24:33,105 --> 00:24:35,605 It's just call it a zero token architecture. 669 00:24:38,160 --> 00:24:40,079 My my favorite I just have to quote 670 00:24:40,079 --> 00:24:41,679 something from this because it made me laugh 671 00:24:41,679 --> 00:24:44,319 out loud. He's he's talking about continue using 672 00:24:44,319 --> 00:24:46,819 automation tools like Puppet, Ansible, and Chef, and 673 00:24:46,880 --> 00:24:48,740 and and put stuff in Cron. 674 00:24:49,115 --> 00:24:51,914 But just rename that to etsyagent.d, 675 00:24:51,914 --> 00:24:53,355 and you'll have all these agents doing all 676 00:24:53,355 --> 00:24:55,775 these automatic things using the zero token architecture. 677 00:24:57,994 --> 00:24:58,894 I love it. 678 00:25:01,355 --> 00:25:01,855 It 679 00:25:02,750 --> 00:25:04,750 it it would be less funny if it 680 00:25:04,750 --> 00:25:05,809 weren't so true. 681 00:25:06,190 --> 00:25:08,350 Right? But we have when you've been in 682 00:25:08,350 --> 00:25:11,230 this industry for a while, you've seen the 683 00:25:11,230 --> 00:25:12,769 the continual rename 684 00:25:13,150 --> 00:25:13,650 and, 685 00:25:14,269 --> 00:25:16,990 you know, buzzword chasing that goes on. And 686 00:25:16,990 --> 00:25:17,890 so, like, 687 00:25:18,345 --> 00:25:20,684 and for like, we we've been naming products 688 00:25:21,144 --> 00:25:23,224 AI for a long time that are that 689 00:25:23,224 --> 00:25:24,684 are just doing, like, 690 00:25:24,984 --> 00:25:27,325 automation or data analytics or, 691 00:25:28,105 --> 00:25:29,724 one of those things. But, 692 00:25:31,009 --> 00:25:33,429 zero token architecture is, 693 00:25:33,970 --> 00:25:35,990 I think, next level, and somebody 694 00:25:36,609 --> 00:25:37,109 deserves 695 00:25:37,649 --> 00:25:40,069 some clever points in my opinion. 696 00:25:41,889 --> 00:25:44,049 This this ought to be a presentation at 697 00:25:44,049 --> 00:25:46,585 every nug and every nog for the next 698 00:25:46,585 --> 00:25:48,444 two years. There ought to be a standardized, 699 00:25:49,065 --> 00:25:50,924 you know, like a standardized 700 00:25:51,224 --> 00:25:51,724 presentation 701 00:25:52,424 --> 00:25:54,444 on using zero token 702 00:25:54,904 --> 00:25:55,404 architectures. 703 00:25:57,865 --> 00:25:59,029 It's absolutely 704 00:25:59,650 --> 00:26:01,410 brilliant. But, I mean, it makes a good 705 00:26:01,410 --> 00:26:03,809 point, though. Right? Too. On the serious side, 706 00:26:03,809 --> 00:26:05,350 it makes a good point. Sometimes, 707 00:26:06,690 --> 00:26:08,769 you just need to, like, roll with the 708 00:26:08,769 --> 00:26:09,269 punches. 709 00:26:10,289 --> 00:26:11,109 Just like, 710 00:26:11,490 --> 00:26:13,944 you know, oh, that's the buzz buzzwords you 711 00:26:13,944 --> 00:26:16,044 wanna use this week? Okay. Well, 712 00:26:16,345 --> 00:26:17,944 I can Right. I can use that buzzword 713 00:26:17,944 --> 00:26:19,404 too. Like, whatever. 714 00:26:20,025 --> 00:26:23,224 Okay. It's And, and and this is a 715 00:26:23,224 --> 00:26:25,484 and this is a great like, the great 716 00:26:25,544 --> 00:26:26,525 Kelsey Hightower 717 00:26:28,559 --> 00:26:31,279 is is quoted here in coming up with 718 00:26:31,279 --> 00:26:32,740 this quip, which is 719 00:26:33,440 --> 00:26:34,579 even more fabulous. 720 00:26:34,880 --> 00:26:35,380 Right? 721 00:26:35,919 --> 00:26:37,539 But I I think it it 722 00:26:38,240 --> 00:26:39,700 there there is an underlying 723 00:26:40,000 --> 00:26:41,679 point here, and I think a lot of 724 00:26:41,679 --> 00:26:44,035 times when organizations start talking about the things 725 00:26:44,035 --> 00:26:46,195 that they could do with AI or exploring 726 00:26:46,195 --> 00:26:46,695 that, 727 00:26:47,154 --> 00:26:48,934 what they started to discover 728 00:26:49,235 --> 00:26:51,394 is that, oh, wait. We could do a 729 00:26:51,394 --> 00:26:53,815 lot of these things with traditional automation. 730 00:26:54,355 --> 00:26:56,295 Right? A lot of it is very deterministic. 731 00:26:57,059 --> 00:26:59,059 We just need to understand our workflows and 732 00:26:59,059 --> 00:27:00,200 what we wanna accomplish, 733 00:27:00,820 --> 00:27:02,840 and a lot of this is still 734 00:27:03,940 --> 00:27:04,440 standard 735 00:27:04,980 --> 00:27:07,400 automation. And a lot of those skills 736 00:27:07,940 --> 00:27:10,360 that we've needed to automate our environments, 737 00:27:12,255 --> 00:27:14,494 they're they're still very necessary. I mean, when 738 00:27:14,494 --> 00:27:15,555 you're talking about 739 00:27:15,934 --> 00:27:16,994 multistep agents, 740 00:27:18,174 --> 00:27:18,674 it's 741 00:27:19,295 --> 00:27:21,075 really just an automated workflow. 742 00:27:21,455 --> 00:27:24,095 Right? Yeah. Yeah. Yeah. You might be using 743 00:27:24,095 --> 00:27:26,029 natural language to to describe to the system 744 00:27:26,029 --> 00:27:27,890 what you want it to do instead of 745 00:27:27,950 --> 00:27:28,929 writing deterministic 746 00:27:29,230 --> 00:27:29,730 code, 747 00:27:31,069 --> 00:27:33,789 but it's still, like, step one, step two, 748 00:27:33,789 --> 00:27:34,529 step three 749 00:27:34,990 --> 00:27:36,289 that you are automating 750 00:27:36,670 --> 00:27:37,490 in a system 751 00:27:37,865 --> 00:27:40,345 that you're kicking off with a scheduler that 752 00:27:40,345 --> 00:27:41,884 we used to call Kron, 753 00:27:44,664 --> 00:27:46,365 and there's still some similarities 754 00:27:46,744 --> 00:27:47,244 there. 755 00:27:47,625 --> 00:27:49,865 Yeah. Yeah. And I think it's also interesting 756 00:27:49,865 --> 00:27:52,845 because sometimes when you wanna get something done, 757 00:27:53,570 --> 00:27:55,410 if you're a network operator and you've wanted 758 00:27:55,410 --> 00:27:57,250 to automate your network for the last ten 759 00:27:57,250 --> 00:27:59,329 years, and you could never get budget to 760 00:27:59,329 --> 00:27:59,990 do it, 761 00:28:00,529 --> 00:28:02,630 and you never get time to do it, 762 00:28:03,170 --> 00:28:05,269 do it through a I mean, use 763 00:28:05,884 --> 00:28:06,384 the 764 00:28:07,085 --> 00:28:09,884 use the flow. Use the hype to your 765 00:28:09,884 --> 00:28:10,384 advantage 766 00:28:11,484 --> 00:28:13,804 to get done what you've always needed to 767 00:28:13,804 --> 00:28:15,345 do or wanted to do 768 00:28:15,804 --> 00:28:17,565 and get the budget and get things that 769 00:28:17,565 --> 00:28:18,924 you need to do. And I think that's 770 00:28:18,924 --> 00:28:19,585 a fantastic, 771 00:28:20,570 --> 00:28:23,710 like, another fantastic way of of thinking about 772 00:28:24,089 --> 00:28:25,869 this renaming of stuff. 773 00:28:28,329 --> 00:28:28,829 Well, 774 00:28:29,609 --> 00:28:31,230 another great quote here. 775 00:28:32,404 --> 00:28:34,565 Kelsey says the agent will burn $2,000,000,000,000 776 00:28:34,565 --> 00:28:36,345 with the tokens and call it an API, 777 00:28:36,404 --> 00:28:38,724 he joked. He also shared a four letter 778 00:28:38,724 --> 00:28:40,744 acronym he recommends as an alternative, 779 00:28:41,365 --> 00:28:41,865 b 780 00:28:42,325 --> 00:28:44,964 a s h for the bash shell, which 781 00:28:44,964 --> 00:28:46,184 is, you know, 782 00:28:46,890 --> 00:28:49,150 it's it's it's quite clever. 783 00:28:52,490 --> 00:28:55,150 Yeah. But but his point is well taken 784 00:28:55,450 --> 00:28:55,950 that 785 00:28:58,250 --> 00:29:00,505 there are many solutions out there to our 786 00:29:00,505 --> 00:29:03,644 problems, and some of them don't require tokens. 787 00:29:04,984 --> 00:29:05,884 They require 788 00:29:06,265 --> 00:29:07,325 a a deterministic, 789 00:29:10,025 --> 00:29:12,025 bit of code that's gonna run on a 790 00:29:12,025 --> 00:29:12,525 schedule 791 00:29:13,220 --> 00:29:14,599 like a cron job 792 00:29:14,980 --> 00:29:16,920 that is going to accomplish tasks. 793 00:29:18,660 --> 00:29:19,799 Yeah. I love it. 794 00:29:20,820 --> 00:29:22,599 Yeah. It's really it's really cool. 795 00:29:23,059 --> 00:29:26,019 Alright. Well, I think that's all I have 796 00:29:26,019 --> 00:29:27,640 for articles to discuss. 797 00:29:28,095 --> 00:29:29,474 You have anything else, Tom? 798 00:29:29,934 --> 00:29:30,434 Nope. 799 00:29:31,134 --> 00:29:32,355 Anything else, Yvonne? 800 00:29:32,894 --> 00:29:33,634 All good. 801 00:29:34,095 --> 00:29:36,255 Alright. So, Tom, where can people find you 802 00:29:36,255 --> 00:29:37,234 if they want to? 803 00:29:40,180 --> 00:29:40,680 LinkedIn. 804 00:29:41,619 --> 00:29:43,779 I I really I think we should just 805 00:29:43,779 --> 00:29:45,559 do compression on that part 806 00:29:46,099 --> 00:29:46,920 of the show. 807 00:29:47,619 --> 00:29:49,380 What do you think, Yvonne? We miss we 808 00:29:49,380 --> 00:29:51,320 miss things. Something's lost 809 00:29:51,940 --> 00:29:54,440 when you do that. Yeah. Yeah. For us. 810 00:29:54,500 --> 00:29:55,240 You mean? 811 00:29:59,555 --> 00:30:02,195 So, Yvonne, where can people find you? Yeah. 812 00:30:02,195 --> 00:30:04,035 You can find me on LinkedIn at Yvonne 813 00:30:04,035 --> 00:30:04,515 Sharp. 814 00:30:05,234 --> 00:30:06,214 You can also 815 00:30:06,515 --> 00:30:08,595 find me at the Cloud Gambit, which is 816 00:30:08,595 --> 00:30:10,695 part of the Packet Pusher's Network, my 817 00:30:11,029 --> 00:30:11,929 other podcast, 818 00:30:12,710 --> 00:30:15,289 or you can find me. I'm more lurking 819 00:30:15,750 --> 00:30:17,750 on Twitter and x these days than I 820 00:30:17,750 --> 00:30:18,569 am posting, 821 00:30:19,190 --> 00:30:21,190 but I am over there at Sharp Network 822 00:30:21,190 --> 00:30:22,649 if you wanna reach out. 823 00:30:23,275 --> 00:30:25,194 Cool. Awesome. Alright. I'm Russ White. You can 824 00:30:25,194 --> 00:30:26,654 always find me here at The Hedge. 825 00:30:27,275 --> 00:30:29,375 My only podcast. I'm not Yvonne. 826 00:30:29,914 --> 00:30:31,055 I have, you know, 827 00:30:35,210 --> 00:30:36,990 Your one trick there. Says statement. 828 00:30:37,289 --> 00:30:37,789 Yeah. 829 00:30:39,130 --> 00:30:40,910 You have a day job too, Russ. 830 00:30:42,329 --> 00:30:44,329 And a double header soccer game to take 831 00:30:44,329 --> 00:30:46,430 a kid to tonight, just so you know. 832 00:30:47,394 --> 00:30:48,134 Just watch 833 00:30:48,515 --> 00:30:49,015 for 834 00:30:49,394 --> 00:30:51,095 the ball. Don't get hit, Yvonne. 835 00:30:51,714 --> 00:30:52,214 Yep. 836 00:30:52,674 --> 00:30:55,075 I'm gonna put my AirPods in and listen 837 00:30:55,075 --> 00:30:56,295 to my audiobook 838 00:30:56,674 --> 00:30:57,174 and 839 00:30:57,714 --> 00:30:59,654 clap at the designated times. 840 00:31:01,019 --> 00:31:01,519 Okay. 841 00:31:01,900 --> 00:31:03,740 Alright. So I'm Russ White. If I'm here 842 00:31:03,740 --> 00:31:04,940 at the hedge, you can find me at 843 00:31:04,940 --> 00:31:07,920 rool eleven dot tech. Find me on LinkedIn 844 00:31:08,460 --> 00:31:10,299 and on x from time to time. I 845 00:31:10,299 --> 00:31:12,140 don't know. I'm pretty easy to find. We 846 00:31:12,140 --> 00:31:14,059 know we live in an attention driven economy, 847 00:31:14,059 --> 00:31:15,920 and we thank you for spending the time 848 00:31:16,115 --> 00:31:17,795 listening all the way to the bitter end 849 00:31:17,795 --> 00:31:19,715 of this episode of The Hedge, and we 850 00:31:19,715 --> 00:31:20,855 will catch you next