May 7, 2025 • 26 mins
The relentless race for cloud dominance continues as AWS reports 16.9% growth year-over-year—a number that would thrill most companies but falls short of expectations for the cloud giant. Tim and Chris dig into what's behind these numbers and why Microsoft and Google continue gaining ground with stronger-than-anticipated growth rates. Microsoft's massive 53% increase in capital spending signals their aggressive push into AI infrastructure, raising questions about whether AWS can maintain its leadership position.
Cybersecurity threats are evolving in concerning ways according to Threat Labs' 2025 Phishing Report. While overall phishing attempts have declined, attacks have become more targeted and sophisticated. The hosts explore disturbing trends including the rise of cryptocurrency scams with fake wallets and an increase in job-related phishing that exploits today's challenging employment market. Most alarming is how threat actors are capitalizing on AI hype, creating fraudulent agent websites that mimic legitimate platforms to steal credentials. This exposes a critical vulnerability in emerging technologies like Multi-agent Collaboration Protocol systems that currently lack robust security frameworks.
The conversation shifts to more positive developments with Kubernetes 1.33 "Octarine" release, which brings sidecar containers to stable status—a significant improvement for managing service mesh implementations. The hosts break down how this and other updates like in-place resource resizing make Kubernetes more flexible and easier to manage for enterprise deployments.
The episode closes with a somber discussion of Intel's announcement of potential layoffs affecting up to 20% of its workforce following substantial quarterly losses. Tim and Chris challenge the new CEO's assertion that requiring more office days will make the company "lean, fast and agile," questioning whether return-to-office mandates have more to do with real estate investments than actual productivity improvements.
What tech news matters most to you? We'd love your feedback on our news format and what topics you'd like us to cover in future episodes!
Purchase Chris and Tim's new book on AWS Cloud Networking: https://www.amazon.com/Certified-Advanced-Networking-Certification-certification/dp/1835080839/
Check out the Fortnightly Cloud Networking News
https://docs.google.com/document/d/1fkBWCGwXDUX9OfZ9_MvSVup8tJJzJeqrauaE6VPT2b0/
Visit our website and subscribe: https://www.cables2clouds.com/
Follow us on BlueSky: https://bsky.app/profile/cables2clouds.com
Follow us on YouTube: https://www.youtube.com/@cables2clouds/
Follow us on TikTok: https://www.tiktok.com/@cables2clouds
Merch Store: https://store.cables2clouds.com/
Join the Discord Study group: https://artofneteng.com/iaatj
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Tim McConnaughy (00:15):
Hey there and welcome back to another episode
of the Cables to CloudsFortnightly News.
I am here, your host this week,tim, and with me, as always, is
Chris, my co-host, who you mayremember from previous episodes
of this podcast.
So yeah, with the fortnightlynews, what we do is we go, we
(00:35):
take a look around, we try tosee what in the last two weeks
has happened that's interesting.
We roll it up and share it andthen comment on it and see if we
find it interesting what we'llsay.
So let's just jump right intoit.
This one just came out fromcloudcompetingnewsnet and the
title of the article is AmazonCloud Revenue Misses Estimates
(00:58):
Again as Rivals Pull Ahead.
I feel like and I know this isobviously I feel like this is
done quarterly anyway, but itseems like every other episode,
we're talking about the race tothe bottom, as it were, of cloud
money, essentially.
So what's interesting here, ofcourse, is that in this article,
they point out that Amazon hadvery slow first quarter results,
(01:21):
even though they stillincreased right first quarter
results, even though they stillincreased right.
So they had a 16.9% increaseyear over year, which, of course
, in any normal business wouldbe really good numbers, but this
being cloud and AWS, apparently.
That's terrible and everyoneshould panic.
So I mean, it's just it's notas fast.
It's not as fast as waspredicted, as they wanted it to
(01:43):
be.
What's interesting, though, inthis article is that they
mentioned that both Microsoftand Google or Alphabet, I guess
had stronger than expectedgrowth, while AWS had lower than
expected growth.
Now we've been doing this awhile now, and I swear to God,
it seems like every time wereport on this stuff, we feel
(02:04):
like this is the point whereMicrosoft is going to overtake
Amazon or AWS as the leader, andit still has yet to happen, but
we do continually report that,quarter over quarter, it seems
like AWS is slowing down.
Microsoft and Alphabet areheating up.
(02:26):
So now put that in perspective.
Of course, aws is the 800-poundgorilla.
Aws is in the lead and has beenin the lead pretty much since
the beginning.
So take all of that with agrain of salt, right.
Microsoft and Google have a lotmore upward trajectory than
they have downward, and AWS ismore or less the reverse, where
it's a lot easier for them tolose ground than it is for them
(02:47):
to gain it once you get to acertain point.
So Andy Jassy said Amazon hasnot yet observed a significant
rise in average selling prices.
Oh, this is sorry.
As part of this article also,they talk about Amazon a little
bit because of the tariffs thing.
So, even though AWS and Amazonare technically tracked
(03:08):
differently, a lot of times whenwe get to earnings calls, for
some reason the reporting onboth of the numbers.
I guess that's because, ofcourse, I guess they report them
both at the same time.
So they talk about third-partysellers and all this other stuff
and tariffs potentiallyimpacting the Amazon side of the
house potentially impacting theAmazon side of the house.
So meanwhile, like I said,microsoft's getting a lot more
(03:29):
growth.
There was something in herespecifically about the spending.
There it is.
Capital spending at Microsoftrose to $16.75 billion, a 53%
year-over-year increase.
So Microsoft is buildinghardware for AI.
Basically that's what it is, sopretty interesting.
(03:49):
Google is, of course, on therise as well.
So there's a lot of numbers inthis article.
I don't want to sit here andregurgitate them, but generally
AWS has slower growth, microsoftand Google are taking off and
there's a ridiculous amount ofmoney being spent on AI.
So you took the article.
Anything to add, chris?
Chris Miles (04:09):
No, I kind of echo your sentiments here.
Whereas it's funny, that evenincreases year over year of this
percentage, which would bephenomenal for any other company
on the earth, seems to causeconcern for investors when it
comes to Amazon or Microsoft orAlphabet, which is just
(04:29):
hilarious.
The thing is, if you're saying,just to do a call to what's
trending on social media rightnow, if Amazon is the 100-pound
gorilla or 800-pound gorilla,sorry to say and there's 100
individual small cloud providers, could they beat Amazon?
Tim McConnaughy (04:48):
That's the question.
Oh my God, so topical, sotopical, yeah, I guess that
remains to be seen.
Yeah, we should ask thatquestion on social media,
actually, and see what peoplesay.
Chris Miles (05:02):
We'll get two answers and they'll both say no,
Probably All right.
Next up, we have an articlehere which is basically a report
, a 2025 phishing report puttogether by Threat Labs that's
Threat Labs with a Z Very, verycool having a Z in there which
(05:24):
is a part of Zscaler, I believe,Basically doing a report on the
effects of AI on scams likephishing over the past year,
Right?
So we've talked a bit on thepodcast about how we think AI is
probably going to maybe notincrease the number of
sophisticated cybersecurityattacks, but at least increase
(05:47):
the kind of more targeted typeof attacks and things like that.
So I wanted to put this in hereto see if we can actually glean
what has happened over the lastyear and there's been some
major call-outs in here.
So basically, ThreatLabs did ananalysis of over 2 billion
(06:09):
blocked phishing transactionsacross their Zero Trust Exchange
platform, which is a cloudsecurity platform.
So this is dated from Januaryto December of last year Some
interesting call-outs in here.
So phishing, while it's down,it's become overly more targeted
as of recent.
So there are more targetedattacks where they've been
(06:34):
highly successful, things likethat.
The U S uh, it says the UnitedStates fishing has declined, but
it still remains number one onthe list of um targeted areas,
which is, uh, is not really ahuge surprise.
Um, interestingly, there was acall out here that crypto scams
are on the rise with fakewallets.
Um, which is uh, which is kindof funny.
(06:59):
I don't know if this hadanything to do with the incoming
administration being kind ofmore, you know, the pro-crypto
party type thing, and not toeven get into the politics of it
.
I'm just wondering if that iswhy there's been an increased
kind of traction in thisparticular area.
And then another thing herewhich is really sad to see but
tech support and job scams arethriving and you know, as we all
(07:23):
exist in this world today, weknow that the job market is
pretty shit right now, so I cansee that being an area of
opportunity for, you know,threat actors and things like
that.
I'm trying to think whatactually else was in here that
was interesting.
I'm trying to think whatactually else was in here that
was interesting.
There were certain call-outshere about phishing.
(07:47):
Sites are now starting to putcaptures on the front pages of
them to kind of mimic theircredibility and get past some
existing security tools that arekind of scanning these things
like that.
So that was an interestingcall-out as well like that, so
that was an interesting call outas well.
All in all, I don't see a hugerise of kind of threats and
things like that in this reportthat lead me directly to AI's
(08:08):
involvement.
To be honest with you, I'mexpecting to see probably more
of that come out in the next,probably next year's, report, so
I look forward to reading thatone.
But I did see some points inhere about vishing taking over,
which vishing is kind of voicevishing, where you know they
kind of are able to impersonatesomeone and mimic their voice in
(08:30):
real time and basically stealvital information.
So I could see that onerelating to AI and like deep
fakes and things like that.
But, yeah, interesting article.
You can read the full report.
You can download it directlyfrom them.
So if you want to have a closerlook, take a look in the show
notes.
Anything else to add?
Tim, how did you feel about it?
Tim McConnaughy (08:50):
Yeah, it was good.
So I did see something.
So I like this one.
Phishing targets, ai hype.
I thought that was great, right.
Fraudulent AI agent websitesthat mimic real platforms are
exploiting the growing trust inAI to steal user credentials and
payment details, and we've beentalking about this for a while
off the show and on the show.
But these new federated systems, if you will, probably gives a
(09:16):
little too much credit, like MCPand some of the agent-to-agent
type of behavior that's growingreally popular in AI workloads.
A lot of that stuff's notsecure.
There's no real security to it.
Nobody's authenticating agents,nobody's doing anything to
(09:36):
validate or encrypt the data ordecrypt the data, like I mean.
It's just it's all very wildwest.
And because nobody has builtsecurity frameworks for these AI
interactions, because the AIinteractions are, like you know,
days old or weeks old, at thispoint it's really ripe for
threat actors to essentially,you know, mimic legitimate you
(09:59):
know legitimate MCP serversBecause, remember, the whole
point of the MCP server is likean exchange, if you will Like an
agent, I believe, an agentexchange or something where
these services can find eachother.
So there's nothing stopping oneof these, or even what used to
happen all the time withsoftware is it would be
(10:20):
legitimate software and thensomebody would abandon it.
You know the maintainer wouldabandon it, and so I would come
along and pick it up and build amalware into it and just keep
the train rolling right so thatstuff can happen Like there's no
validation that I've seen onany of this.
So that, I thought, was a veryinteresting insight from this
report.
I say insight.
(10:40):
I say insight I mean it shouldbe obvious to anybody who knows
has been looking atcybersecurity for a period of
time.
But it's cool to see itdocumented that like hey, yeah,
this is happening right.
Like this is something peopleare going to have to deal with
as AI becomes more and more partof our work life.
Chris Miles (10:55):
Yeah, I mean, I think about the timing of it and
obviously, with the cutoffbeing December of last year,
that's when I feel like we werejust starting to talk a lot more
about AI agents and now we'vemoved on from agents to MCP.
Now MCP is the new rage andit's funny because you hear the
kind of counterpoint a lot oftimes that adding security can
(11:18):
kind of stifle the innovation orthe agility of new technology
and things like that.
But, like, when I look at mcpand like look, I, I'm not an
expert, have a very rudimentaryunderstanding of it, it seems
very easy to build some basicsecurity into that, like with
common protocols and things likethat that are in use on the
internet, very much so today, um, and it's not there yet.
(11:40):
But that that at the same time,because it's not in, you know,
such a wide adoption just yet,then threat actors aren't going
to be targeting that, so to say,just yet.
I'm hoping by the time therubber meets the road and MCP is
kind of that like standardframework that everyone uses,
then maybe you know the uh, thesecurity will have already been
(12:03):
kind of baked in and but, uh,you know, I've I've I've heard
the joke several times alreadyand I think it's too funny to
not repeat.
But um, people are saying thatthe S in MCP stands for security
, so I think that's uh, veryrelevant right now All right, Uh
, so this next one is from Ithink that's very relevant right
now.
Tim McConnaughy (12:20):
All right, so this next one is from, I think
it's actually.
Oh, this is from InfoQ.
Yeah, so InfoQcom basically hasan article that talks about
essentially the release notesfrom the new version of
Kubernetes that has beenreleased, called Octarine, which
apparently is an homage toTerry Pratchett's Discworld.
(12:42):
I don't know how or why,because it's my wife who is the
Discworld junkie.
I am not.
I've listened to maybe one ortwo audiobooks from Terry
Pratchett on a road trip, yeah,and watched.
Was it the one they had onAmazon?
Was it Amazon the one withDavid Tennant and Michael Sheen?
(13:02):
Oh, shoot, I forgot the onewith the devil and the angel,
zerphil and Crowley.
Chris Miles (13:09):
I was going to say, man, if it's between me and you
, you're the resident experthere.
I've not ventured into it.
Tim McConnaughy (13:15):
Yeah, I completely forgot the name of it
all of a sudden.
That's okay, I'm going to kickmyself when I remember, but
anyway.
So yeah, new version ofKubernetes released.
One of the most anticipatedfeatures in Kubernetes 1.33 is
the promotion of sidecarcontainers to stable status.
So this is interesting and I'mstill not a Kubernetes expert.
(13:35):
I've been working on it.
I don't know if I'll ever be aKubernetes expert, but I have
been learning a lot more aboutKubernetes.
So the idea here basically issidecar containers are an add-on
basically to the wholeKubernetes environment and it
usually has been used forservice mesh.
So service mesh providers likeIstio or Linkerd, they deploy a
(14:01):
sidecar, what's called acontainer.
It's a container inside a podwhose sole purpose of being in
the pod is to facilitatecommunication with other
containers that have theirsidecars.
Now the whole thing aboutsidecar containers is that up
until now it hasn't been part ofthe native Kubernetes workflow
or Kubernetes lifecycle.
It's been kind of managed asits own thing, kind of like CNI
(14:24):
is my understanding.
So this now is saying basicallyokay, so now we've got sidecars
as part of the Kubernetessolution.
Essentially it's not just anadd-on, a bolt-on, a vendor hack
.
Now this is actually part ofthe Kubernetes lifecycle, so you
(14:44):
can actually manage sidecarswith Kubernetes, manifests and
all of that.
So that's, I think, a step inthe right direction.
You have expansion of yoursolution to include things that
a vendor is adding functionalitythat wasn't originally built
into the product.
So they've seen widespreadadoption.
(15:07):
They see the value of bringingit in-house essentially or be
part of the actual basicsolution.
So that's the big one.
There's a few more here I like.
This too actually Promotesin-place resource resizing for
vertical scaling of pods.
Now, this is a beta feature.
This isn't a stable feature yet, but it's interesting that we
(15:27):
can, because we didn't have thisbefore.
You have to blow away your podsand do manifest changes and
spin up new pods.
I'm curious the use case and Iwould love to hear if somebody
knows, by the way, what the usecase.
And I would love to hear ifsomebody knows, by the way, what
the use case is for verticallyresizing your pods.
I guess it's just so you don'tblow away the container, right,
(15:48):
so you don't have to kill thecontainer to do it.
There's obviously some value inthat.
But then again, kubernetes waskind of predicated on the idea
of microservices and we couldjust continue.
Cattle, not pets, right, wedon't care, essentially.
So I am curious what the realvalue here is, and I think that
maybe that's why it's a betafeature they're trying to see
like are people using it?
Do they need this?
Is this something that this isa problem that needs to be
(16:10):
solved?
So that's a pretty interestingone.
Also, enhanced support forservice account tokens.
So tokens are something Ihaven't really played with a lot
at all, but the idea, of course, is it's a security feature.
So talk about so we're justtalking about security features
here.
Here's one too, right, so wehave the ability for our pods
and applications to pass tokensfor communication purposes.
(16:30):
And then there was one more thatlooked interesting.
Um, they are changing the waythat service ips are allocated.
Um, so before you just prettymuch had cluster ip, that would
would, uh, is what you would usefor kind of inter-cluster
services, because the idea hereis that you expose the service
and the service is what you hit.
Is that north star that you canhit, so that you can hit the
(16:53):
back-end application containers?
Because those containers areliving, dying and being reborn
all the time.
Essentially, you don't want touse them as a target.
So the idea of the cluster IPin the service is that here's a
static spot, if you will, thatwe can hit, and then on the back
end it's almost like a loadbalancer.
Chris Miles (17:12):
Well, it is really a load balancer, just like a
virtual IP, is basically what itis.
Tim McConnaughy (17:16):
Yeah, virtual IP with a load balancer that
points it back in containers.
So they've changed it now toadd this new thing called
service CIDR and IP addressImplementation enables cluster
administrators to dynamicallyexpand the IP address pool
available for cluster IP bycreating additional service CIDR
objects.
I'm assuming that before that,if you wanted to expand that out
(17:37):
, you had to blow some stuffaway and restart.
Now, containers should not besomething that would be
impactful, but I couldabsolutely see where blowing
away cluster IPs could beimpactful to applications, right
, because that's how thosemicroservices are communicating
with each other.
So pretty interesting stuff.
Yeah, it's pretty exciting.
I have to read up on some ofthe newer things and still try
(17:59):
to understand how they fit here.
But yeah, dude, you gotanything to add here.
I don't know where you're atwith your Kubernetes studies.
Chris Miles (18:08):
Yeah, not a ton to add.
I mean, the sidecar thing hasobviously, I think, been the
major call out, from what I'veseen people commenting on this.
So, like you know, like whenI've talked to you know
customers and people atconferences and things like that
, you know, the idea of runninga sidecar operation can be a
(18:29):
pain sometimes.
I'm curious to know from peoplethat are working with this on a
daily basis, like is the whenit comes to sidecar?
Um, is the?
Is the pain point made mainlythe orchestration of it, which
would, I'm assuming, be solvedby this update or kind of you
know, be a much less of a burdenby this update?
(18:52):
Or is it more kind ofoperational, uh type thing that
that comes into play there whereit's you know kind of order of
operations around, when thesidecar does what, et cetera,
versus the existing services, etcetera?
So I'd be curious to know, youknow, how much are people seeing
this as like a, like a?
(19:13):
Are they sighing a breath ofrelief at this point?
Or or is there other majorproblems that come in with the
sidecar that are still going toremain?
So I'd be curious to hear yeah,definitely, all righty.
And lastly, we'll round out withanother bit of a sad one here.
So we have an article from USAToday about our good friend at
(19:33):
Intel.
So Intel their new CEO, which Ibelieve is Lipbutan is how you
pronounce that name hasbasically announced that, or
alluded to layoffs coming, asthe company has reported about
821 million in losses over Q1 ofthis year.
(19:57):
So it looks like they could beanticipating laying off up to
more than 20% of the workforceand, as we just talked about,
this job market is not in a goodstate.
So this is not gonna be goodfor a lot of those people, which
is very sad to hear.
You know, he's apparently hadsent out an email initially
(20:21):
saying that we are seen as tooslow, too complex and too set in
our ways and we need to change.
So he's kind of been sowing theseeds that this is coming.
It seems like he's coming inand kind of doing the whole
thing.
You know like we've beenrunning inefficiently.
We need to kind of, you know,cut waste, cut things and do
more with less type thing.
(20:43):
Thankfully, intel has kind ofbeen on the camp where they've
been bailed out on the tariffs.
You know, saying that they'rebeing bailed out on something
that shouldn't exist at all inthe first place is kind of a
funny sentence, but nonetheless,that's the world we're living
in.
So I'm assuming they'reprobably anticipating for when
the tariffs are actually goingto come back in full force and
(21:05):
things like semiconductors willnot be excluded from the
existing tariffs.
So this is just the latest inIntel's decline.
It's kind of a sad state to seethem continually on this kind
of downward trend.
I'm hearing a lot of peoplepredicting that it's going to be
sold off for parts in theupcoming future.
(21:27):
So yeah, it's sad to see, butthat's the state we're working
in.
Anything to add, tim?
Tim McConnaughy (21:35):
Yeah, so I mean , yeah, this is not as well.
On this article he specificallysays he's calling for a return
to work policy in which someemployees, who currently spend
three days per week on site,would be spending four days on
site.
And specifically he says ourcompetitors are lean, fast and
agile and that's what we mustbecome to improve our execution
(22:00):
execution.
So apparently being lean yeah,apparently being lean, fast and
agile means adding one more dayto your in office.
I'm not sure what those twohave to do with each other, but
apparently that's why I'm not aCEO.
Guys, I'm sorry.
And then also he says thesecritical changes will reduce the
size of our workforce, butobserve that Intel must balance
our reductions with the need toretain and recruit key talent,
Because there's nothing that keytalent loves more than going to
(22:22):
companies that lay off 20% oftheir workforce.
Chris Miles (22:25):
Yeah, exactly.
Tim McConnaughy (22:26):
They're just waiting in line to do that.
I mean, I don't know what elseto say except to point out that
this entire article is full ofboilerplate.
We have to make layoffs andchallenging market and blah,
blah, blah, and it's literally.
You just change the names andit would be any tech company
that you know is, quote unquotefacing headwinds.
(22:48):
There's nothing new here.
Tech continues to implode.
So yeah, and of course, forsome reason, the answer to tech
imploding all of these peoplethink the answer to tech
imploding is to get everybodyback in the office.
Because, as we, know, duringCOVID, tech burned to the ground
and nobody was able to get anywork done.
(23:09):
So the only way to get workdone now is to get people back
in the office.
Chris Miles (23:15):
Yeah, I don't think there is an Amazon office in
Chandler Arizona.
But if they were upset aboutthe four days back in the office
, if you could stomach one more,maybe you could go to AWS and
just do the full five days.
Tim McConnaughy (23:33):
I still struggle with the idea that the
reason people aren't beingsuccessful is because they're
not five days in the office.
Trends should show this to bethe case.
You should be able to look atthe last five years now or three
, let me say, yeah, the lastfive years and draw a very easy
line that says our productivityand value and ability to get
(23:55):
shit done has steadily declinedsince people have not been in
the office.
Where's that graph?
It doesn't exist, right, itdoesn't exist.
So this, all of this bullshit,is really about real estate.
It's always been about realestate and tax incentives.
Real estate and tax incentivesis what it's always been about.
So you're not fooling anybody,dude.
That's all I have to add onthat one.
(24:16):
All right, let's go ahead andwrap this up today.
So I hate to end on a sour notethere, but this is the world
we're living in and it's onlygetting more interesting.
And when I say more interesting, I mean, like in the ancient
Chinese curse, like, may youlive in interesting times, like
(24:36):
that, that style.
So, yeah, we do have a questionfor you guys.
We would love to know if youenjoy the news um more less, uh,
than you know.
Is there a particular formatthat you like?
Is there more uh stuff that youwould like to see do for the
news?
Uh, different?
Would you like to expand out toinclude different topics?
(24:59):
It would be really great tounderstand.
Basically, is the news helpfulto you and, if so, how does it
help?
How do you consume it and whatare we missing?
Chris Miles (25:13):
so if you'd like to consume it.
If there's a way you'd like toconsume it in a certain way that
we're not doing today, let usknow.
I'm just open to hearing somemore ideas, because we know
there's a way you'd like toconsume it in a certain way that
we're not doing today.
Let us know.
Just open to hearing some moreideas, because we know there's a
lot of tech news out there.
It's not a unique thing tooffer that by any means.
So we hope that you're cominghere because you like Tim and
I's personality, but we're notalways the best people, so we're
(25:40):
just curious to hear from youguys what you'd like us to
change or if you think it'sworking, or open to any and all
feedback.
Tim McConnaughy (25:46):
Yeah, appreciate it All right.
Well, we'll go ahead and cut itthere, everybody, we'll see you
next week.
Bye, thanks for watching.
Hey there and welcome back to another episode
of the Cables to CloudsFortnightly News.
I am here, your host this week,tim, and with me, as always, is
Chris, my co-host, who you mayremember from previous episodes
of this podcast.
So yeah, with the fortnightlynews, what we do is we go, we
(00:35):
take a look around, we try tosee what in the last two weeks
has happened that's interesting.
We roll it up and share it andthen comment on it and see if we
find it interesting what we'llsay.
So let's just jump right intoit.
This one just came out fromcloudcompetingnewsnet and the
title of the article is AmazonCloud Revenue Misses Estimates
(00:58):
Again as Rivals Pull Ahead.
I feel like and I know this isobviously I feel like this is
done quarterly anyway, but itseems like every other episode,
we're talking about the race tothe bottom, as it were, of cloud
money, essentially.
So what's interesting here, ofcourse, is that in this article,
they point out that Amazon hadvery slow first quarter results,
(01:21):
even though they stillincreased right first quarter
results, even though they stillincreased right.
So they had a 16.9% increaseyear over year, which, of course
, in any normal business wouldbe really good numbers, but this
being cloud and AWS, apparently.
That's terrible and everyoneshould panic.
So I mean, it's just it's notas fast.
It's not as fast as waspredicted, as they wanted it to
(01:43):
be.
What's interesting, though, inthis article is that they
mentioned that both Microsoftand Google or Alphabet, I guess
had stronger than expectedgrowth, while AWS had lower than
expected growth.
Now we've been doing this awhile now, and I swear to God,
it seems like every time wereport on this stuff, we feel
(02:04):
like this is the point whereMicrosoft is going to overtake
Amazon or AWS as the leader, andit still has yet to happen, but
we do continually report that,quarter over quarter, it seems
like AWS is slowing down.
Microsoft and Alphabet areheating up.
(02:26):
So now put that in perspective.
Of course, aws is the 800-poundgorilla.
Aws is in the lead and has beenin the lead pretty much since
the beginning.
So take all of that with agrain of salt, right.
Microsoft and Google have a lotmore upward trajectory than
they have downward, and AWS ismore or less the reverse, where
it's a lot easier for them tolose ground than it is for them
(02:47):
to gain it once you get to acertain point.
So Andy Jassy said Amazon hasnot yet observed a significant
rise in average selling prices.
Oh, this is sorry.
As part of this article also,they talk about Amazon a little
bit because of the tariffs thing.
So, even though AWS and Amazonare technically tracked
(03:08):
differently, a lot of times whenwe get to earnings calls, for
some reason the reporting onboth of the numbers.
I guess that's because, ofcourse, I guess they report them
both at the same time.
So they talk about third-partysellers and all this other stuff
and tariffs potentiallyimpacting the Amazon side of the
house potentially impacting theAmazon side of the house.
So meanwhile, like I said,microsoft's getting a lot more
(03:29):
growth.
There was something in herespecifically about the spending.
There it is.
Capital spending at Microsoftrose to $16.75 billion, a 53%
year-over-year increase.
So Microsoft is buildinghardware for AI.
Basically that's what it is, sopretty interesting.
(03:49):
Google is, of course, on therise as well.
So there's a lot of numbers inthis article.
I don't want to sit here andregurgitate them, but generally
AWS has slower growth, microsoftand Google are taking off and
there's a ridiculous amount ofmoney being spent on AI.
So you took the article.
Anything to add, chris?
Chris Miles (04:09):
No, I kind of echo your sentiments here.
Whereas it's funny, that evenincreases year over year of this
percentage, which would bephenomenal for any other company
on the earth, seems to causeconcern for investors when it
comes to Amazon or Microsoft orAlphabet, which is just
(04:29):
hilarious.
The thing is, if you're saying,just to do a call to what's
trending on social media rightnow, if Amazon is the 100-pound
gorilla or 800-pound gorilla,sorry to say and there's 100
individual small cloud providers, could they beat Amazon?
Tim McConnaughy (04:48):
That's the question.
Oh my God, so topical, sotopical, yeah, I guess that
remains to be seen.
Yeah, we should ask thatquestion on social media,
actually, and see what peoplesay.
Chris Miles (05:02):
We'll get two answers and they'll both say no,
Probably All right.
Next up, we have an articlehere which is basically a report
, a 2025 phishing report puttogether by Threat Labs that's
Threat Labs with a Z Very, verycool having a Z in there which
(05:24):
is a part of Zscaler, I believe,Basically doing a report on the
effects of AI on scams likephishing over the past year,
Right?
So we've talked a bit on thepodcast about how we think AI is
probably going to maybe notincrease the number of
sophisticated cybersecurityattacks, but at least increase
(05:47):
the kind of more targeted typeof attacks and things like that.
So I wanted to put this in hereto see if we can actually glean
what has happened over the lastyear and there's been some
major call-outs in here.
So basically, ThreatLabs did ananalysis of over 2 billion
(06:09):
blocked phishing transactionsacross their Zero Trust Exchange
platform, which is a cloudsecurity platform.
So this is dated from Januaryto December of last year Some
interesting call-outs in here.
So phishing, while it's down,it's become overly more targeted
as of recent.
So there are more targetedattacks where they've been
(06:34):
highly successful, things likethat.
The U S uh, it says the UnitedStates fishing has declined, but
it still remains number one onthe list of um targeted areas,
which is, uh, is not really ahuge surprise.
Um, interestingly, there was acall out here that crypto scams
are on the rise with fakewallets.
Um, which is uh, which is kindof funny.
(06:59):
I don't know if this hadanything to do with the incoming
administration being kind ofmore, you know, the pro-crypto
party type thing, and not toeven get into the politics of it
.
I'm just wondering if that iswhy there's been an increased
kind of traction in thisparticular area.
And then another thing herewhich is really sad to see but
tech support and job scams arethriving and you know, as we all
(07:23):
exist in this world today, weknow that the job market is
pretty shit right now, so I cansee that being an area of
opportunity for, you know,threat actors and things like
that.
I'm trying to think whatactually else was in here that
was interesting.
I'm trying to think whatactually else was in here that
was interesting.
There were certain call-outshere about phishing.
(07:47):
Sites are now starting to putcaptures on the front pages of
them to kind of mimic theircredibility and get past some
existing security tools that arekind of scanning these things
like that.
So that was an interestingcall-out as well like that, so
that was an interesting call outas well.
All in all, I don't see a hugerise of kind of threats and
things like that in this reportthat lead me directly to AI's
(08:08):
involvement.
To be honest with you, I'mexpecting to see probably more
of that come out in the next,probably next year's, report, so
I look forward to reading thatone.
But I did see some points inhere about vishing taking over,
which vishing is kind of voicevishing, where you know they
kind of are able to impersonatesomeone and mimic their voice in
(08:30):
real time and basically stealvital information.
So I could see that onerelating to AI and like deep
fakes and things like that.
But, yeah, interesting article.
You can read the full report.
You can download it directlyfrom them.
So if you want to have a closerlook, take a look in the show
notes.
Anything else to add?
Tim, how did you feel about it?
Tim McConnaughy (08:50):
Yeah, it was good.
So I did see something.
So I like this one.
Phishing targets, ai hype.
I thought that was great, right.
Fraudulent AI agent websitesthat mimic real platforms are
exploiting the growing trust inAI to steal user credentials and
payment details, and we've beentalking about this for a while
off the show and on the show.
But these new federated systems, if you will, probably gives a
(09:16):
little too much credit, like MCPand some of the agent-to-agent
type of behavior that's growingreally popular in AI workloads.
A lot of that stuff's notsecure.
There's no real security to it.
Nobody's authenticating agents,nobody's doing anything to
(09:36):
validate or encrypt the data ordecrypt the data, like I mean.
It's just it's all very wildwest.
And because nobody has builtsecurity frameworks for these AI
interactions, because the AIinteractions are, like you know,
days old or weeks old, at thispoint it's really ripe for
threat actors to essentially,you know, mimic legitimate you
(09:59):
know legitimate MCP serversBecause, remember, the whole
point of the MCP server is likean exchange, if you will Like an
agent, I believe, an agentexchange or something where
these services can find eachother.
So there's nothing stopping oneof these, or even what used to
happen all the time withsoftware is it would be
(10:20):
legitimate software and thensomebody would abandon it.
You know the maintainer wouldabandon it, and so I would come
along and pick it up and build amalware into it and just keep
the train rolling right so thatstuff can happen Like there's no
validation that I've seen onany of this.
So that, I thought, was a veryinteresting insight from this
report.
I say insight.
(10:40):
I say insight I mean it shouldbe obvious to anybody who knows
has been looking atcybersecurity for a period of
time.
But it's cool to see itdocumented that like hey, yeah,
this is happening right.
Like this is something peopleare going to have to deal with
as AI becomes more and more partof our work life.
Chris Miles (10:55):
Yeah, I mean, I think about the timing of it and
obviously, with the cutoffbeing December of last year,
that's when I feel like we werejust starting to talk a lot more
about AI agents and now we'vemoved on from agents to MCP.
Now MCP is the new rage andit's funny because you hear the
kind of counterpoint a lot oftimes that adding security can
(11:18):
kind of stifle the innovation orthe agility of new technology
and things like that.
But, like, when I look at mcpand like look, I, I'm not an
expert, have a very rudimentaryunderstanding of it, it seems
very easy to build some basicsecurity into that, like with
common protocols and things likethat that are in use on the
internet, very much so today, um, and it's not there yet.
(11:40):
But that that at the same time,because it's not in, you know,
such a wide adoption just yet,then threat actors aren't going
to be targeting that, so to say,just yet.
I'm hoping by the time therubber meets the road and MCP is
kind of that like standardframework that everyone uses,
then maybe you know the uh, thesecurity will have already been
(12:03):
kind of baked in and but, uh,you know, I've I've I've heard
the joke several times alreadyand I think it's too funny to
not repeat.
But um, people are saying thatthe S in MCP stands for security
, so I think that's uh, veryrelevant right now All right, Uh
, so this next one is from Ithink that's very relevant right
now.
Tim McConnaughy (12:20):
All right, so this next one is from, I think
it's actually.
Oh, this is from InfoQ.
Yeah, so InfoQcom basically hasan article that talks about
essentially the release notesfrom the new version of
Kubernetes that has beenreleased, called Octarine, which
apparently is an homage toTerry Pratchett's Discworld.
(12:42):
I don't know how or why,because it's my wife who is the
Discworld junkie.
I am not.
I've listened to maybe one ortwo audiobooks from Terry
Pratchett on a road trip, yeah,and watched.
Was it the one they had onAmazon?
Was it Amazon the one withDavid Tennant and Michael Sheen?
(13:02):
Oh, shoot, I forgot the onewith the devil and the angel,
zerphil and Crowley.
Chris Miles (13:09):
I was going to say, man, if it's between me and you
, you're the resident experthere.
I've not ventured into it.
Tim McConnaughy (13:15):
Yeah, I completely forgot the name of it
all of a sudden.
That's okay, I'm going to kickmyself when I remember, but
anyway.
So yeah, new version ofKubernetes released.
One of the most anticipatedfeatures in Kubernetes 1.33 is
the promotion of sidecarcontainers to stable status.
So this is interesting and I'mstill not a Kubernetes expert.
(13:35):
I've been working on it.
I don't know if I'll ever be aKubernetes expert, but I have
been learning a lot more aboutKubernetes.
So the idea here basically issidecar containers are an add-on
basically to the wholeKubernetes environment and it
usually has been used forservice mesh.
So service mesh providers likeIstio or Linkerd, they deploy a
(14:01):
sidecar, what's called acontainer.
It's a container inside a podwhose sole purpose of being in
the pod is to facilitatecommunication with other
containers that have theirsidecars.
Now the whole thing aboutsidecar containers is that up
until now it hasn't been part ofthe native Kubernetes workflow
or Kubernetes lifecycle.
It's been kind of managed asits own thing, kind of like CNI
(14:24):
is my understanding.
So this now is saying basicallyokay, so now we've got sidecars
as part of the Kubernetessolution.
Essentially it's not just anadd-on, a bolt-on, a vendor hack
.
Now this is actually part ofthe Kubernetes lifecycle, so you
(14:44):
can actually manage sidecarswith Kubernetes, manifests and
all of that.
So that's, I think, a step inthe right direction.
You have expansion of yoursolution to include things that
a vendor is adding functionalitythat wasn't originally built
into the product.
So they've seen widespreadadoption.
(15:07):
They see the value of bringingit in-house essentially or be
part of the actual basicsolution.
So that's the big one.
There's a few more here I like.
This too actually Promotesin-place resource resizing for
vertical scaling of pods.
Now, this is a beta feature.
This isn't a stable feature yet, but it's interesting that we
(15:27):
can, because we didn't have thisbefore.
You have to blow away your podsand do manifest changes and
spin up new pods.
I'm curious the use case and Iwould love to hear if somebody
knows, by the way, what the usecase.
And I would love to hear ifsomebody knows, by the way, what
the use case is for verticallyresizing your pods.
I guess it's just so you don'tblow away the container, right,
(15:48):
so you don't have to kill thecontainer to do it.
There's obviously some value inthat.
But then again, kubernetes waskind of predicated on the idea
of microservices and we couldjust continue.
Cattle, not pets, right, wedon't care, essentially.
So I am curious what the realvalue here is, and I think that
maybe that's why it's a betafeature they're trying to see
like are people using it?
Do they need this?
Is this something that this isa problem that needs to be
(16:10):
solved?
So that's a pretty interestingone.
Also, enhanced support forservice account tokens.
So tokens are something Ihaven't really played with a lot
at all, but the idea, of course, is it's a security feature.
So talk about so we're justtalking about security features
here.
Here's one too, right, so wehave the ability for our pods
and applications to pass tokensfor communication purposes.
(16:30):
And then there was one more thatlooked interesting.
Um, they are changing the waythat service ips are allocated.
Um, so before you just prettymuch had cluster ip, that would
would, uh, is what you would usefor kind of inter-cluster
services, because the idea hereis that you expose the service
and the service is what you hit.
Is that north star that you canhit, so that you can hit the
(16:53):
back-end application containers?
Because those containers areliving, dying and being reborn
all the time.
Essentially, you don't want touse them as a target.
So the idea of the cluster IPin the service is that here's a
static spot, if you will, thatwe can hit, and then on the back
end it's almost like a loadbalancer.
Chris Miles (17:12):
Well, it is really a load balancer, just like a
virtual IP, is basically what itis.
Tim McConnaughy (17:16):
Yeah, virtual IP with a load balancer that
points it back in containers.
So they've changed it now toadd this new thing called
service CIDR and IP addressImplementation enables cluster
administrators to dynamicallyexpand the IP address pool
available for cluster IP bycreating additional service CIDR
objects.
I'm assuming that before that,if you wanted to expand that out
(17:37):
, you had to blow some stuffaway and restart.
Now, containers should not besomething that would be
impactful, but I couldabsolutely see where blowing
away cluster IPs could beimpactful to applications, right
, because that's how thosemicroservices are communicating
with each other.
So pretty interesting stuff.
Yeah, it's pretty exciting.
I have to read up on some ofthe newer things and still try
(17:59):
to understand how they fit here.
But yeah, dude, you gotanything to add here.
I don't know where you're atwith your Kubernetes studies.
Chris Miles (18:08):
Yeah, not a ton to add.
I mean, the sidecar thing hasobviously, I think, been the
major call out, from what I'veseen people commenting on this.
So, like you know, like whenI've talked to you know
customers and people atconferences and things like that
, you know, the idea of runninga sidecar operation can be a
(18:29):
pain sometimes.
I'm curious to know from peoplethat are working with this on a
daily basis, like is the whenit comes to sidecar?
Um, is the?
Is the pain point made mainlythe orchestration of it, which
would, I'm assuming, be solvedby this update or kind of you
know, be a much less of a burdenby this update?
(18:52):
Or is it more kind ofoperational, uh type thing that
that comes into play there whereit's you know kind of order of
operations around, when thesidecar does what, et cetera,
versus the existing services, etcetera?
So I'd be curious to know, youknow, how much are people seeing
this as like a, like a?
(19:13):
Are they sighing a breath ofrelief at this point?
Or or is there other majorproblems that come in with the
sidecar that are still going toremain?
So I'd be curious to hear yeah,definitely, all righty.
And lastly, we'll round out withanother bit of a sad one here.
So we have an article from USAToday about our good friend at
(19:33):
Intel.
So Intel their new CEO, which Ibelieve is Lipbutan is how you
pronounce that name hasbasically announced that, or
alluded to layoffs coming, asthe company has reported about
821 million in losses over Q1 ofthis year.
(19:57):
So it looks like they could beanticipating laying off up to
more than 20% of the workforceand, as we just talked about,
this job market is not in a goodstate.
So this is not gonna be goodfor a lot of those people, which
is very sad to hear.
You know, he's apparently hadsent out an email initially
(20:21):
saying that we are seen as tooslow, too complex and too set in
our ways and we need to change.
So he's kind of been sowing theseeds that this is coming.
It seems like he's coming inand kind of doing the whole
thing.
You know like we've beenrunning inefficiently.
We need to kind of, you know,cut waste, cut things and do
more with less type thing.
(20:43):
Thankfully, intel has kind ofbeen on the camp where they've
been bailed out on the tariffs.
You know, saying that they'rebeing bailed out on something
that shouldn't exist at all inthe first place is kind of a
funny sentence, but nonetheless,that's the world we're living
in.
So I'm assuming they'reprobably anticipating for when
the tariffs are actually goingto come back in full force and
(21:05):
things like semiconductors willnot be excluded from the
existing tariffs.
So this is just the latest inIntel's decline.
It's kind of a sad state to seethem continually on this kind
of downward trend.
I'm hearing a lot of peoplepredicting that it's going to be
sold off for parts in theupcoming future.
(21:27):
So yeah, it's sad to see, butthat's the state we're working
in.
Anything to add, tim?
Tim McConnaughy (21:35):
Yeah, so I mean , yeah, this is not as well.
On this article he specificallysays he's calling for a return
to work policy in which someemployees, who currently spend
three days per week on site,would be spending four days on
site.
And specifically he says ourcompetitors are lean, fast and
agile and that's what we mustbecome to improve our execution
(22:00):
execution.
So apparently being lean yeah,apparently being lean, fast and
agile means adding one more dayto your in office.
I'm not sure what those twohave to do with each other, but
apparently that's why I'm not aCEO.
Guys, I'm sorry.
And then also he says thesecritical changes will reduce the
size of our workforce, butobserve that Intel must balance
our reductions with the need toretain and recruit key talent,
Because there's nothing that keytalent loves more than going to
(22:22):
companies that lay off 20% oftheir workforce.
Chris Miles (22:25):
Yeah, exactly.
Tim McConnaughy (22:26):
They're just waiting in line to do that.
I mean, I don't know what elseto say except to point out that
this entire article is full ofboilerplate.
We have to make layoffs andchallenging market and blah,
blah, blah, and it's literally.
You just change the names andit would be any tech company
that you know is, quote unquotefacing headwinds.
(22:48):
There's nothing new here.
Tech continues to implode.
So yeah, and of course, forsome reason, the answer to tech
imploding all of these peoplethink the answer to tech
imploding is to get everybodyback in the office.
Because, as we, know, duringCOVID, tech burned to the ground
and nobody was able to get anywork done.
(23:09):
So the only way to get workdone now is to get people back
in the office.
Chris Miles (23:15):
Yeah, I don't think there is an Amazon office in
Chandler Arizona.
But if they were upset aboutthe four days back in the office
, if you could stomach one more,maybe you could go to AWS and
just do the full five days.
Tim McConnaughy (23:33):
I still struggle with the idea that the
reason people aren't beingsuccessful is because they're
not five days in the office.
Trends should show this to bethe case.
You should be able to look atthe last five years now or three
, let me say, yeah, the lastfive years and draw a very easy
line that says our productivityand value and ability to get
(23:55):
shit done has steadily declinedsince people have not been in
the office.
Where's that graph?
It doesn't exist, right, itdoesn't exist.
So this, all of this bullshit,is really about real estate.
It's always been about realestate and tax incentives.
Real estate and tax incentivesis what it's always been about.
So you're not fooling anybody,dude.
That's all I have to add onthat one.
(24:16):
All right, let's go ahead andwrap this up today.
So I hate to end on a sour notethere, but this is the world
we're living in and it's onlygetting more interesting.
And when I say more interesting, I mean, like in the ancient
Chinese curse, like, may youlive in interesting times, like
(24:36):
that, that style.
So, yeah, we do have a questionfor you guys.
We would love to know if youenjoy the news um more less, uh,
than you know.
Is there a particular formatthat you like?
Is there more uh stuff that youwould like to see do for the
news?
Uh, different?
Would you like to expand out toinclude different topics?
(24:59):
It would be really great tounderstand.
Basically, is the news helpfulto you and, if so, how does it
help?
How do you consume it and whatare we missing?
Chris Miles (25:13):
so if you'd like to consume it.
If there's a way you'd like toconsume it in a certain way that
we're not doing today, let usknow.
I'm just open to hearing somemore ideas, because we know
there's a way you'd like toconsume it in a certain way that
we're not doing today.
Let us know.
Just open to hearing some moreideas, because we know there's a
lot of tech news out there.
It's not a unique thing tooffer that by any means.
So we hope that you're cominghere because you like Tim and
I's personality, but we're notalways the best people, so we're
(25:40):
just curious to hear from youguys what you'd like us to
change or if you think it'sworking, or open to any and all
feedback.
Tim McConnaughy (25:46):
Yeah, appreciate it All right.
Well, we'll go ahead and cut itthere, everybody, we'll see you
next week.
Bye, thanks for watching.
Popular Podcasts
Cold Case Files: Miami
Joyce Sapp, 76; Bryan Herrera, 16; and Laurance Webb, 32—three Miami residents whose lives were stolen in brutal, unsolved homicides. Cold Case Files: Miami follows award‑winning radio host and City of Miami Police reserve officer Enrique Santos as he partners with the department’s Cold Case Homicide Unit, determined family members, and the advocates who spend their lives fighting for justice for the victims who can no longer fight for themselves.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.