¶ Cloud Buyers Regret
All right , before I hit stop , I'll just say this real quick while it's recording and then , if we want to use this as the cold open , we can . But I just want to preemptively apologize to everyone listening today because Tim and I are quite sick .
Tim is probably in the throes of it as I'm coming down from it , so if our energy is low and we sound like absolute death , we apologize and we promise we're very happy and we really enjoyed doing this for you today , so please be nice to us .
Welcome to the Cables to Clouds podcast , your one-stop shop for all things hybrid and multi-cloud networking .
Now here are your hosts Tim , chris and Alex networking .
Now here are your hosts Tim , Chris and Alex . Hello and welcome back to your Cables to Clouds fortnightly news episode . My name is Alex Perkins . I am at Bumps in the Wire on socials . I am joined today by two sick co-hosts Chris Miles at BGP Maine and Tim McConaughey at Juan Golbez .
We are going to jump right in and I will try to keep the energy high for you guys . So let's get started with the mandatory AI news . We've got an article here from Reuters called Open AI . Co-founder Sutzgever's New Safety-Focused AI Startup , SSI , Raises $1 Billion Now . Sutzgever is Ilya Sutzgever . We've talked about him quite a bit on this show .
He was one of the OpenAI co-founders . He was on the board of OpenAI .
He was instrumental in all the same Altman drama when he got fired and then brought back , and he eventually left the super alignment team , which is basically a safety team for artificial intelligence , and ended up founding this new company called SSI , which I believe stands for safe super intelligence .
They just closed a series A for $1 billion Absolutely ridiculous . I saw some estimates . They didn't give out a valuation . The quote in this article is that people close to the matter in quotations is they're valued around $5 billion currently . This company has 10 employees right now , 10 . Okay , so this is absolutely nuts .
And we talked about this I don't know , maybe probably , I think , within the last month about all these new safety regulations that are going to be popping up , and there was something I called out about a requirement to have a third party company be used for verifying like models and stuff like this , and I to me I'm guessing that's what this whole thing is
about is that they're kind of getting a very early jump on being that third party for all these different AI companies , so they could make a lot of money , I guess . Who knows , with AI , it's been a lot of fun watching these markets go , so I will stop talking here . What do you guys think ?
This might be just my pure ignorance , but obviously they're coming in as super safe intelligence or super safe or sorry , safe super intelligence , kind of backwards safe super intelligence .
It's funny to me to think that they're getting evaluations this high and getting funding this high when it seems to me like their whole model is to just kind of counteract the other side of it , like AI is obviously pushing innovation and changes with you know the way it's doing things , and then they're coming in just say we're actually going to halt a bit of
that and slow it down . And so it's crazy that there's that much funding in going in the opposite direction when this thing hasn't really even shown that it has true ROI yet . We've seen that from many other companies that are even pulling out of their ai adoption so far .
So , like I said , that might just be my pure ignorance on the matter and I don't know where this fits in , but it's , it seems um , like almost like it's almost like it's getting funding to like kind of stop it in its tracks in a way before it can even really become profitable . Yeah , it's a good point .
I don't know . I mean a five billion dollar valuation for a company in its tracks in a way , before it can even really become profitable . Yeah , that's a good point . I don't know . I mean a $5 billion valuation for a company that doesn't have any sort of product . It's more like an idea . Right ?
We're going to create a product that will in some way enforce some kind of safety framework .
They don't even have compute .
Yeah , there's no detail of any kind about what they're going to do . How will they make models safer ? They raised a billion dollars . I mean me . I would be like you know what . There's 10 of us , 100 million per person . Let's go ahead and close up shop and move on .
You know , like I wouldn't be done .
That'd be good enough for me , man . Sorry , we couldn't make it work , guys , Goodbye .
Yeah , I mean I don't know what to say to that $5 billion evaluation is absolutely insane . There's no product , there's no details . The $5 billion evaluation is just on the guy , pretty much . Yeah , it's on Ilya , it's all me . 1 billion . The plan is to uh , what did it say ?
Basically to hire top talent and start building out their compute that they need , but again , like you don't have anything in , like there's nothing , there's no plan , there's no , there's .
They don't have anything currently , so it's just well I'll tell you based on their current funding , they can . If anyone from there is listening , they can hire me for a measly 10 million dollars out of all their funding . So hit me up . I'll put my phone number in the show notes .
The only other thing before we move on . I wanted to call this out the investors . There's something really interesting here . So there's this group called , or an investment partnership called , NFDG . It's run by the guy who is CEO of SSI . How is this legal ? I would immediately think that that can't be legal .
Is it a VC firm ? Because if so , that actually kind of makes sense right .
It's an investment partnership .
It sounds like a VC , but maybe , yeah , who knows what that actually looks like . It sounds like , basically , he went to whoever else was in his partnership and say hey , invest in this . So yeah , all right . Yeah , invest in us . Yeah , invest in this . So yeah , so all right .
Yeah , I thought it was pretty crazy . All right Up next we got an article from the register titled cloud buyers regret . Now you can imagine just from the title what this is about , and the headline of the article says admins wonder if the cloud was such a good idea after all , it looks like . So this report was .
This report was published by uk cloud outfit sivo , which I don't know a whole lot about them . I kind of looked at the their page real quick and it seemed like they are very much in the game of cloud finances . So of course I don't .
I don't want to say that it's going to be a super biased report , but it's going to be a little bit biased towards , you know , the information , the information that these guys are trying to find Financial information .
Yeah .
I mean , Tim , you got any thoughts here ?
I know we've been talking about cloud costs forever , so I mean how long ago was it now that we did this , the episode about where all the cloud savings ?
I feel like it's been in over a year , yeah Right .
And it's the same thing . It was when we said it that long ago this article , by the way , they get into the details of basically how much . I forget how they do it . They put together a bundle of compute and block storage and whatnot . They say basically it costs like $1,000 or $1,200 . And basically the price keeps going up , outstripping even inflation .
So , despite the fact that everybody was told the cloud would become cheaper as it becomes more commoditized and more people using it means they can spread the savings and all that , it's actually only getting more expensive .
And I think that goes back to the whole thing we were talking about and the same thing we came to , the same thing we thought about in that episode , which is really the cloud providers can't actually make it cheaper , right , because they're just going to end up DDoSing everybody .
Right , if you make it so cheap , then you'll have everybody doing it and they don't have the capacity . So I think they are artificially keeping the numbers high .
Yeah , and for reference , they said the configuration they used is a three-node cluster with 200 gig persistent volume and 5 terabytes data transfer
¶ Tech Industry News and Analysis
. Chris , you got any thoughts here Was it just in oh , that was in Azure .
Yeah , they did all of them , the big three , they did the big three Gotcha . Yeah , I mean it's . I'm glad they factored into like a real world scenario . That seems good , but like I don't understand . Like SIVO , like the article calls out the beginning that you know , the survey unsurprisingly paints SIVO in a flattering light .
Like , what do they have to gain in this situation ? Did you guys extract that ?
I didn't really understand that either .
I'm like what is this ?
FinOps company maybe , I'm not sure . Well , yeah , that's what doesn't make sense , right ?
If people leave the cloud , isn't that their business model ? To help companies save money in the cloud ? So , yeah , I don't quite get that one . Yeah , that's a little weird of a sentence , but it's funny because a lot of these they said they used the CSP's calculators for this . Right , right , but yeah not much to add .
I think , like I said , we kind of called this probably close to a year ago now that the savings were kind of a not necessarily like a bait and switch , but it's just like now . They're at the point , like you said , with capacity and things like that . I just don't think they can do it . It'll run itself over very quick .
Yeah , they'll end up DDoSing everybody else as they get more and more customers if the price comes down .
Like Azure keeps doing in there I mean that's .
Yeah , that's true . I just I just saw an email from one of our customers saying uh had requested a quota increase for a certain um family in europe and uh , yeah , it got denied , you know . So they can't even use that , that family in europe now because of the capacity problems .
All right , so we got another uh story , this one's from broadcom , our favorite uh whipping boy . Um , so this one is uh .
So broadcom's just had its earnings , just finished the earnings report , and apparently the shares are kind of slumping because the revenue , the , the forecast , basically the revenue forecast from broadcom com is down , um , and of course , when you're a company of like broadcom size , any movement ends up giving you like some sticker shock here .
But the the real reason for this ? Um , I think the article says , like it's lost 73 billion dollars in value , which you know that looks great on paper but nobody's actually out there spending 73 billion dollars .
Um , the reason is because the forecast for like kind of the demand and delivery of their AI chips that they're building they have certain AI ASICs and chips that they're building . I guess the demand for it isn't as high as they kind of initially thought that it would be , and this I've actually seen .
This is just another facet of something I've seen , which is a lot of companies , you know , after jumping into the AI thing , you know try like oh no , we have to do this to be competitive , are kind of realizing that , like actually , no , we kind of , you know , chasing , kind of chasing nothing here .
So I don't know , what do you guys think about this one ?
Yeah right , so they . They forecasted an increase in ai chips of a billion dollars over the fiscal year and that just wasn't enough to please the investors . So that's why they sold off .
They lost 10 of their , their shares , share value yeah 10 just crazy . Yeah , you're their size when you're , when you're Broadcom's size , a billion dollars is probably a drop . It's like Cisco , right , it's like , but it's it means nothing , right yeah , they .
I mean what Broadcom's like a like 700 billion dollar company like it's they have to be .
If they , yes , they have to be over 700 billion yeah , yeah , 10% .
73 billion , yeah yeah .
So it's , you know it's . I mean that's , we're talking about 73 billion that they lost in market value , right , so that's all perceived value . It's not even not even real at this point . But I mean , broadcom's not necessarily known for being , you know , a risk , a risk taker , so I could see them making some , some cuts in certain areas because of this .
But , um , yeah , I mean , I think this just kind of leans back to what we're talking about the the ai , the ai roi thing . Um , nobody's really seeing it come to fruition yet . Um , and you know , if anyone's gonna , uh , probably swing the axe , I'd say it'd be broadcom .
But , like we were talking about this before , we started hitting record like tim , like , yes , this , this caused a slump , their , you know their stock price probably dropped a bit , but it'll come right back like it's , yeah , it's a stock market only goes up , yep , exactly , all right , um . So next up , we have um , an announcement .
We don't necessarily have an article here , but we do have an announcement from HashiCorp that they have finally launched the first version , I believe , of their . I'm assuming this means they're going to have multiples , but there's a Terraform professional certification that's coming out .
We wanted to cover this because I think when we had Chris Williams on the show he was talking about this a bit . So it's kind of cool to see this finally coming to fruition .
So they've put out this certification called the Terraform Authoring and Operations Professional , and I believe this one is specifically targeted at AWS , so it covers a domain and this is a lab-based exam , still online proctored , but we're talking about four-hour tests here with a 15 minute break .
So we're thinking you're really going to be in the weeds like building and , as it says in the name , authoring Terraform . So you're probably building things from scratch . You know , you're probably using the Linux shell and building YAML and JSON , et cetera , right , so this seems like a pretty all in comprehensive test .
So I'm so I'm actually kind of excited to see something like this on the market from them , because I think we talked about this on the show the Terraform Associate . I think all of us have probably taken it here on the show , or Alex hasn't , but me and . Tim have . I think Tim will probably agree with me . It was a little too easy , if I'm being honest .
I felt like it was if you had any day-to-day operations with Terraform . It was pretty simple , but pretty cool to see them put this out , and I think the early registration is already up for it , and I believe if you register now , I think it includes a free retake as well . So it might be actually a good time to get in and try it .
But yeah , how do you ?
guys feel about this . I think it's awesome , man . I love lab-based exams because it's not trivia , like you need to know it right , but like you don't need to sit there and memorize timers and and ABCD , like questions right , like you just have to get in there and do the thing . So yeah , a four hour exam though that's crazy , like I love it .
I mean , I haven't taken an exam that long since I did my CCIE right the mini ccie of so it's half half the time .
Right , it's half the time but still four hours is four hours , though they do not provide you jason's deli with oh , this , oh , is that what you had you ? Had I don't know if you know I took it many times it was jason's deli 90 of the time yeah , I uh .
Well , in rtp you had , um , we had , uh , you're gonna at least once in rtp , right ? Yeah , david , we had david blair Blair there and he had like the hot sauce collection and they had it catered from the cafeteria or whatever .
No , I think we had Indian food like one time in RTP , but every other time it was Jason's Deli . It was abysmal , it was so depressing .
Oh man , I also agree . This is awesome . As someone who has not taken any kind of long lab exam like this , I love the idea , because what better way to show your skills than an actual practical , hands-on lab exam ? And , chris , like you were saying , there is a blueprint on this page and it's exactly that .
Like you're literally you're building out a module , you're like configuring remote state I'm just looking at some of the basic stuff in here , but it's like it's even saying you know you Terraform apply , terraform destroy . Like you're actually going in and building out an entire environment .
So yeah , it's very cool .
Yeah , I think this is awesome . The only thing I don't like is that there's no to renew this . You have to retake the exam . That's it , you have to retake the exam . That's it . Yeah , that's the only thing that I think they need to look into changing .
Oh , did they say ? How long is it ?
Two years , two years , yeah , I don't know , two years you have two years to give feedback and get them to change that before you have to take it again .
Yeah , that is a long exam to have to take every two years , but I like the idea of doing it at least once , obviously .
I mean to be honest . I bet , if you're that in the weeds with Terraform and you're doing that stuff every day , I bet that would be easy for you like building a module .
Four hours is very generous .
Yeah , those were the things I felt was missing from the first exam . So I feel like it'd be easy to show that you're a . You're a you know , adequate terraform user and that's I mean , that's a .
That's a good point . That's a very good point . Maybe it doesn't . Maybe the actual task list would not take somebody who would you know is the candidate to pass that exam , would not actually take them four hours well , I mean chris you , when you retook the associate , you just went and took it right .
It wasn't like yeah , I didn't study .
Yeah , so maybe it'll be similar . I mean it's $295 , I think $300 . So that's not bad for .
Not for a four hour lab exam , exactly .
I mean , Cisco specialty exam is more than that .
That's less than $100 per hour .
That's where you get your money's worth there .
All right , less than a hundred dollars per hour , that's that's . Uh , there you go . Money's worth there , all right , uh , last one for today . Um , so , as always , we try to squeeze in a new fun article that won't piss anyone off . Um , so today we are offering a new industry term that has come to the light . Um , we are now talking about micro clouds .
Um , apparently so . We have have an article here today from IT Pro Today bringing up this concept of micro clouds being the next big thing in cloud computing , or just another edge strategy . I'm not necessarily shitting on the article very much , it's just like I do find it crazy that we're talking about another term , especially like this .
So once they break down , like , what is a micro cloud ? It's really ready-to-use server clusters outside of traditional data centers , and the more you read about it you're like , why is this called something different ? This is just edge computing , right ? This is pretty much cut and dry edge computing .
And then later in the article , it says if micro clouds sound a lot like edge computing . And then , light in the article , it says if micro , if micro cloud sound a lot like edge computing , that's because they are , um , so they they kind of relate back to , I guess canonical actually launched a product that was called micro cloud .
Um , so that's , I don't know if they're they're leading in on that branding or what have you , but , um , yeah , how do you guys feel how , how soon before we we switch to a , you know , a micro cloud networking podcast
¶ Confusion Over Micro Cloud Definitions
?
I hate this so much . I yeah , this is the worst article I've read in a long time about the cloud . Uh , why we don't need a new term . And and if ? If micro cloud was a thing , this this isn't it . Like this is , it doesn't even make sense . Like you said it's . Eventually , the whole point is , if this sounds like edge computing , it's because it is .
This is nothing new . It's somewhere . I saw that maybe they offer some kind of software to interface with this prebuilt cluster , but what is that software ? Because I doubt that it's like a cloud management interface like AWS console or Azure . You know it's nothing like that , or they would have called that out in the article .
So I hate this article and I wish it was never written .
I don't think it's the fault of the people that wrote the article necessarily , because I think honestly , they might have . The more I read the article , the more I think that maybe they have the same feeling .
I think they're just reporting on , like this thing called that we're calling micro cloud with a lowercase m as opposed to micro cloud with the uppercase m , which is the canonical product . But , yeah , I mean .
So you look at , I went through this entire article like three times , trying to understand what made this micro cloud versus just edge computing or really just you know , like an MSP type of offer where a managed service provider just builds your servers for you and then you can consume them . It's basically the same shit , right ?
There's literally nothing in here that says this is a cloud , this is a this , this is a cloud product , because X , y , z , just like you can do in cloud . Like there's none of that , right , there's nothing . There's no interface , like you said . There's no consumption model .
I can't even see the consumption model being the same , like I don't know where they get off using the word cloud . It's just edge computing , or somebody else building your servers , pre-built servers for you , like MSP . So , yeah , this is ridiculous . We're getting too many damn marketing terms .
Yeah , I mean the way it's written . I question if at some point it's like pre-packaged Kubernetes , clusters that are running specific types of services , and maybe I could , I could maybe see that being something that's consumable in a in a very quick way . That's different than edge computing . I don't know if it's worthy of calling it a completely different .
You know , product necessarily um or environment , even in that case , um , but I don't even know that . That's what I can extract from this .
That's just me trying to trying to see glass half full , I guess , but yeah well , and even if you take that uh chick-fil-a has in all their stores , they have deployed kubernetes clusters . Is that a micro cloud ? That was by a different company .
Yeah , go in there and order the micro cloud .
Yeah , I just don't , I don't get it , the micro clouds with cheese , please .
And they'll say what do the Chick-fil-A people say ?
My pleasure , I haven't been inside a Chick-fil-A in so many years . Dude , oh dude , you don't know about the my pleasure thing I don't , or I don't remember it's been probably 10 years since I went inside of a Chick-fil-A .
I think it's like a rule Every time you say thank you , they have to say MyPleasure . Oh , got it . They must get like 30 lashings if they don't , because they do it every time yeah .
I'm sorry , the only own it . I guess that's the idea is that you like an is play and this is more like a pass play where they're gonna build the servers for you like . That's the only thing that I can see .
That's even slightly different but yeah , and then if you look at this from the opposite direction , like the csps as they push out , like aws has outposts , that's not a micro cloud , that's an extension of AWS , right ?
Cloud light .
Cloud light .
It's smaller cloud . Yeah , we need a stratification . We need like eight different sizings of cloud before we're done .
Yeah , somebody needs to get a diagram .
Yeah , mega diet cloud to go Super , super size me . I can't wait till they start doing the monopoly game with all the with the cloud offering so you can get all the little pieces you know . Oh geez .
I love it .
All right . Well , with that , I think we are going to wrap it up . Thanks for tuning in today . If you liked what you heard or saw , please , you know , subscribe like . Heard or saw . Um , please , you know , subscribe like .
Share this around to all your friends , tell them about how awesome the show is and how Chris and Tim stuck it out , even though they were both sick . Uh , so , thanks , thanks , guys , and uh , we will talk to you guys next week . Bye , hi everyone . It's Alex and this has been the cables to clouds podcast . Thanks for tuning in today .
If you enjoyed our show , please subscribe to us and your favorite podcatcher , as well as subscribe and turn on notifications for our YouTube channel to be notified of all of our new episodes . Follow us on socials at Cables to Clouds . You can also visit our website for all of the show notes at Cables to Clouds dot com .
Thanks again for listening and see you next time .