July 1, 2025 • 32 mins
The Cloud Gambit is joining the Packet Pushers Network! Launched in 2023 as an independent podcast, The Cloud Gambit was created for engineers who lead, leaders who build, and founders who need both perspectives. We are thrilled at the opportunity to continue our podcasting journey on the Packet Pushers Network. To keep following the show, you will need to resubscribe on your favorite pod catcher. You can find updated links below!
New Podcast Links!
- The Cloud Gambit: https://packetpushers.net/podcast/the-cloud-gambit/
- Apple Podcasts: https://podcasts.apple.com/us/podcast/the-cloud-gambit/id1823741017
- Spotify: https://open.spotify.com/show/23UdyZ3ZwCKB7clwULXyeS
- Pocketcasts: https://pocketcasts.com/podcasts/43146260-380d-013e-acd1-0e87279210d5
- RSS Feed: https://feeds.packetpushers.net/thecloudgambit/
Other Links from the Show
- A2A to Linux Foundation: https://developers.googleblog.com/en/google-cloud-donates-a2a-to-linux-foundation/
- OpenAI Google Deal?: https://www.reuters.com/business/retail-consumer/openai-taps-google-unprecedented-cloud-deal-despite-ai-rivalry-sources-say-2025-06-10/
- Tech Giants Emission Surge: https://www.reuters.com/sustainability/climate-energy/tech-giants-indirect-emissions-rose-150-three-years-ai-expands-un-agency-says-2025-06-05/
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
William (00:00):
If I was a dancing person, I would dance right now,
but I don't dance.
Eyvonne (00:04):
Yeah, I'm not a dancer either.
We didn't even dance at ourwedding.
William (00:08):
The Cloud Gambit is getting adopted into the Packet
Pushers family of podcasts.
Welcome to another episode ofthe Cloud Gambit.
(00:29):
And you know what's crazy,yvonne, I started this podcast
back in August of 2023.
And I think right around 2024,and I think it was in August.
It was almost like exactly ayear.
Eyvonne (00:46):
You joined as co-host, thank you, by the way yeah, it's
been fun, and does it feel likeit's been almost a year?
William (00:54):
it doesn't really feel.
It's one of the few things inmy life right now that doesn't
feel like a year like everythingelse, like work and other stuff
.
It's like, oh, I feel like I'vebeen doing this for 20 years.
But yeah, it's been, it's beengood, I really enjoyed it and
it's so much nicer to have apodcast and not be alone, to
(01:16):
have someone that you could talkto.
Just different perspectives,different opinions.
You know it's been great.
Today's going to be kind oflike a news episode.
We have lots of news, by theway, both about the show and
with some major news headlinesas well.
I guess, first things first,what's going on in the world of
(01:40):
Yvonne?
Are you staying cool?
Eyvonne (01:44):
No, not cool.
Well, fortunately everybody'sfamiliar with the shed at this
point and it is air-conditionedand I even have some wonderful
humidity controls, so you knowworkday is pretty comfortable.
But we are in the midst of aneastern northeastern heat wave
so it's been pretty warm here inthe south midwest northeast so
(02:05):
so it's been pretty pretty warmhere in um in the south midwest
northeast, so it's been a littletoasty.
Um.
Lots of, lots of fun life stuffgoing on.
I've got a kid who, by thewho's getting married.
It may be done by the time thiscomes out.
So we've been in the throes ofof wedding planning and summer
(02:25):
fun and kid activities and got alot coming up in the month of
July, including with the podcast.
So we're excited.
William (02:34):
There's a lot going on.
Yeah, planning a wedding is noeasy feat.
That can be a very busy andvery stressful time.
Eyvonne (02:44):
And it is something I can do.
But it is not a skill set thatcomes naturally.
So it has been, but it's beenfun.
You know, it's kind of a oncein a lifetime thing, so we're
embracing it for the fun that itis, and we'll look at the
credit card bill later.
William (03:01):
Yeah, I love it.
That's smart.
So yeah, yesterday.
So I'm kind of on a quasivacation.
I'm working from the northeast,uh, where my in-laws are, and I
spent somewhere around likenine hours in the sun with my
kids yesterday at this amusementpark, hershey park in
pennsylvania, and if you can'ttell in the video, I'm, I'm, I'm
(03:23):
cooked, uh, the heat index waslike 105 and you know the sweat
and steam that I producedpersonally probably power my
home for a week.
It was hot.
Eyvonne (03:34):
Sunscreen is your friend, william, and wear a hat
Next time you may send me goingon vacation.
I'm screened.
Yeah, I'm gonna.
I going to have to start my uh,my worst sunscreen and wear a
hat, mantra.
William (03:48):
Yeah, I mean, it's pretty bad.
So my, my kid, he doesn'tcomplain.
He's nine years old and youknow, this is like.
The only thing he said all daywas like it was probably about
15 minutes in and we were justwalking through the park and he
looks at me and he's like dad in.
And we were just walkingthrough the park and he looks at
me and he's like dad.
(04:08):
This is like a different kindof hot.
This is really really hot, Iguess it is.
Let me tell you and it's.
Eyvonne (04:13):
it's been a very wet season, at least here, for us as
well, and so the humidity isalso, uh, through the roof, and
we're, of course, trying to tendto the yard, because there's an
outdoor wedding in our future,and, yeah, it's been a fun
summer.
William (04:28):
So the first big piece of news that we have is the
cloud game, but is gettingadopted into the Packet Pushers
family of podcasts.
Now, for those of you thatdon't know about Packet Pushers,
which it would be hard tobelieve, but they were founded
in I think it was 2010, by GregFarrow and Ethan Banks with the
(04:51):
goal of really creating aplatform where just technical
professionals could discusstechnology, share practical
knowledge like real-world stuffand really, I guess, explore
like industry type developments,and over the years, I think
they've solidified, you know,their reputation as a uh, a
(05:13):
leading voice in like enterpriseit media, if you will.
So their network of podcast hasalso evolved by adding, uh,
lots of new shows over the yearsthat specialize in different
areas.
So what does this mean, just atthe top, for the cloud gamut,
like what's going to happen?
(05:34):
So we will both stay aboard ashosts, yvonne and myself um, the
the transition, like the finaltransition, happens July 23rd.
So on July 23rd, that will bethe first episode under the
Packet Pushers family and theshow will continue to be
(05:58):
released on a bi-weekly basis.
It does move to oh boy,wednesday, yes, wednesday, okay,
so from Tuesday to Wednesday,thank you, yvonne, and the RSS
feed will change, so you willneed to resubscribe on your
preferred catcher when the timecomes, and we will send out a
(06:21):
reminder as well.
Am I missing anything?
Eyvonne (06:27):
I don't think so.
It's exciting for me becausePacket Pushers has been kind of
a stalwart technology source ofinformation for me since I
started in enterprise IT.
You know I started my firstenterprise role in 2010 and then
(06:49):
, you know, discovered thepodcast within a year or two of
that, met the Packet Pushersfolks out at Cisco Live and got
to be on the show back in theearly days and really that
community from you know, the2010s really helped me grow and
(07:10):
fuel my career, and so it'sreally exciting to to be part of
their network and to beassociated with everything from
you know network break.
They've got their heavynetworking podcast.
They also have strategy andleadership and operations
(07:30):
conversations.
So I think we you know ourcontent fills a unique slot
there.
We both have networkingbackgrounds, but we are also
both firmly entrenched in thecloud world these days, um and
so just super excited about uh,uh, yeah, join, joining that
(07:50):
network and being a part of thePacket Pushers world.
William (07:55):
Yes, If I was.
If I was a dancing person, Iwould dance right now, but I
don't dance.
Eyvonne (08:02):
I'm not a dancer either .
We didn't even dance at ourwedding.
Like we're just.
William (08:07):
I didn't either.
Eyvonne (08:08):
Yeah, I sometimes look at my friends and colleagues who
have a cultural background ofdancing and body movement with
envy and all.
But yeah, that's not the worldI came, came from and I'm too
old to fix it now yeah,sometimes I dance.
William (08:30):
I guess I dance and sing in the shower, um,
sometimes.
But yeah, that's a shower.
Time is reserved for my, mysinging antics.
Yeah, I try not to take itoutside of the you know privacy
of my own home definitely I'm anappreciator of music.
Eyvonne (08:45):
I'll say that.
Said that way.
William (08:48):
So, yeah, again, we're excited and we will send out
another reminder closer to thetime when the transition happens
.
And that is not the only bignews that has happened.
Um, recently there's actuallybeen a lot of stuff too much to
talk about.
I took three, I think, two orthree stories that I think are
(09:12):
uh, pretty, pretty noteworthy.
I guess I'll go through thefirst one.
This, basically and this is abig one this is actually
something I called a while back,which I never call anything
like lately, I feel like which Inever call anything Like lately
.
I feel like I don't callanything correctly because the
landscape is just very different.
I feel like it's a surpriseevery week, to be honest.
(09:34):
But Google Cloud has basicallydonated its agent to agent or 8A
protocol to the LinuxFoundation, which is a great
place.
I love this decision and itbasically it's creating an
independent project.
You know, that's backed byother big names, other big tech
(09:54):
giants, like you know, aws,cisco's, microsoft's, like
Salesforce's of the world, andto me this is noteworthy.
I mean, there's a lot ofreasons.
I could probably ramble.
I'm going to try not to ramble,um, but I think the first is it
really represents that crucialstep towards, uh, I'd say like
(10:15):
standardization, like in thisemerging crazy ai centric agent
based economy, ai-centricagent-based economy.
So this A2A protocol, basicallywhat it does is it provides a
single or just a common languagefor AI agents to discover
(10:36):
capabilities and exchangeinformation and then allows them
to coordinate very complextasks across different third
parties, platforms, vendors, etcetera, which is huge.
And I think the second thing isit really again it breaks down
(10:58):
the silos that I think havelimited AI's potential I'll just
say it that way Especially forenterprise adoption, because it
creates an open standard under aneutral governance.
So the industry isacknowledging that agent
interoperability is far tooimportant to be controlled,
(11:21):
obviously, by a single company.
And I think, third is probablythe timing is just very
strategic.
So with, I mean, I'd say likeover what?
Like 50, 100 companies nowsupporting the protocol, like
including AWS and Cisco, youknow as it's, you know more
(11:44):
recent validators, we're seeingthe industry coalesce around a
like a real standard, just asagent technology is beginning to
really proliferate.
So, yeah, I think it's thedonation to the Linux Foundation
ensures, you know, at thiscritical time, that this
critical component remains, youknow, vendor agnostic and
(12:05):
community driven.
Any thoughts from you, yvonne?
Eyvonne (12:10):
Well, I think it's a really important step and if you
look back through the historyof our technologies, you know if
you go all the way back to theInternet of our technologies,
you know if you go all the wayback to the internet we've
always had standards and publicorganizations that help dictate
and drive interoperability oftechnology.
(12:32):
And when that's happened,that's when we've seen the most
rapid adoption andubiquitousness of that
technology, whether you're goingback and talking about
networking protocols, ieee andthose types of organizations, or
(12:52):
whether you're talking aboutKubernetes and how it was open
sourced and I believe thathappened in 2014, if my notes
are correct.
Yes, the open source ofKubernetes and what we've seen
there with the containers, cloudnative technologies, container
(13:13):
orchestration and I think thething that we know is that even
when a technology is opensourced, there's plenty of room
in the marketplace for vendorsto create add value, support
that technology.
You know, we've seen that.
You know, in the decade plussince Kubernetes was released,
(13:37):
you know we've got every majorcloud provider has their own
Kubernetes engine.
We've got, you know,on-premises solutions for that
as well.
But also it provides afoundation for interoperability
and also gives enterprises afirm foundation of understanding
(14:04):
what that technology is goingto look like and it solidifies
it.
I think one of the things thatGoogle donating A2A to the Linux
Foundation is going to do isalso going to solidify that that
technology is here to stay,that it's a real thing, that it
can be relied on, and it willbecome the central focal point
(14:29):
for how AI agents and systemsinteract with one another, and
that will go a long way tohelping us advance the
technology in a way that folksdon't feel so locked in or
beholden to one particularvendor, because I think right
now that's a bit of what we'vegot right Like because the ages
(14:54):
don't interoperate well andthere's not yet an ecosystem.
I've said a lot of words, butthat's really what I'm getting
to is that this is going to helpan ecosystem build up around AI
technology and agents so thatthey can interoperate and
interact.
There'll be bumps and there'llbe challenges, but it's a great
(15:15):
step.
William (15:16):
Yeah, those are all great points and the
interoperability is so huge andI think before A2A and MCP
specifically, you basically hada scenario where you had to
create one-off If you wanted tointegrate.
It was like for each major LLMyou had single integrations for
whatever model, back to whatevertechnology.
(15:37):
You're working with guardrailsand the sort of locking things
down and keeping things safebecame very challenging and you
had so much custom stuff.
So these two really flip thescript there and make AI broadly
accessible and adoptable in asecure way that we expect to
(15:57):
have to be able to adopt themwith, I guess.
On to the next thing thing, andthis is another big one.
So on to the next news story.
Uh, I'm just gonna read theheadline hold on, this is a big
one.
So, um, open ai taps googleunprecedented cloud deal,
despite ai rivalry sources sayso.
(16:20):
I thought this was interestingand many, many would call this a
very surprising twist becauseyou know, open ai well, let's
just start here.
So open ai, you know, hasbasically, according to the news
, finalized a deal to use googlecloud like services, kind of in
the way that they usedMicrosoft Azure Compute for its
(16:43):
growing compute needs, andanybody that knows any of these
companies will know that Gemini,google, they're all kind of
competing, if you could call it.
I guess it is competing in theAI race right now.
So there is some intenserivalry between some of these
companies and this dealreportedly finalized in.
(17:07):
May and marks a significantshift in competitive dynamics in
the AI industry.
So why does this matter?
So just my thoughts on it is itreally highlights just this
staggering compute demand ofbuilding AI in 2025.
(17:27):
So OpenAI's decision to reallydiversify beyond Microsoft Azure
from its earlier days reallyunderscores how absolutely
massive these computingrequirements for training and
deploying AI models are sort ofreshaping business relationships
(17:51):
.
So even on OpenAI's annualizedrevenue seven numbers here in my
notes so it's surged to 10billion.
So its infrastructure needscontinue to basically outpace
what a single, you know cloudprovider can probably deliver in
the time that they need it, Iwould assume.
(18:12):
Uh, and that the second thing isum and this is just something
that comes top of mind is itkind of represents probably a
strategic pivot for OpenAI tosort of reduce dependency on
Microsoft at the end of the day?
Since ChatGVT launched late2022.
(18:33):
Openai has sort of been makingdifferent moves to kind of
diversify its infrastructure,like that computational
substrate, including the I thinkit was like 500 billion
Stargate data center project.
I think SoftBank and Oraclewere involved with that.
One Deals with companies likeCoreWeave and now Google Cloud.
(18:57):
So you know, that screams to mejust diversification, prevent
lock-in and just kind of liketrying to think ahead for future
capacity demands.
Any thoughts, yvonne?
Eyvonne (19:12):
Well, the winner in all this is NVIDIA.
You know they continue to winright now and it's.
You know we were talking earlyin the show about kind of our
career trajectories andtechnically kind of our career
(19:35):
trajectories and technically.
And one of the things that youknow, back in my very first job
at a mom and pop ISP again, Iwas, you know, 19, 20, they made
us build our own PCs, right,and so they ordered the hardware
and, like part of us gettingfamiliar with our environment
was like you're going toassemble your own hardware,
build your own PC.
(19:55):
And I remember, you know, therewas a time when it was and there
was a significant growth inhardware and hardware needs and
(20:18):
then we shifted into a point intime in the industry where we
didn't really care as much abouthardware.
You know everything was prettymuch standardized on x86.
You had some outliers there, youhad your AS400s, you had your,
you know, some customizedcompute systems, but we really
just sort of ordered the nextfastest thing.
The next time the refresh cyclecame around and it didn't
(20:40):
really impact what we werebuilding or how we were building
.
Well, now we've shifted thependulumist one back again and
now we really care abouthardware and right now, compute
processing power is really arate limiting factor for
(21:03):
everybody in the AI space, and Ithink you know what this move
signals is how important thatcompute capacity is and how it's
creating, maybe, relationshipsthat we would not have imagined,
you know, even 18 months ago,because there is such a
(21:25):
dependence on powerful computeand what that's going to mean
for the future of infrastructure.
What's it going to mean forpower generation, what's it
going to mean for data centers,tooling, all of those things
Like there's a.
You know there's a five to 10year trajectory, I think, where
(21:48):
physical plant is going to againbecome incredibly important,
where it's just been a thingyou've had to have for the last
decade.
William (21:57):
It's not going to be that way.
Moving forward.
Eyvonne (21:59):
And so there's some significant strategic advantage
to folks who figure that out andwho are innovating in that
particular space.
William (22:09):
Yeah, good points and I agree, nvidia is like one of
these big casinos at this point.
Eyvonne (22:16):
Yeah, they're just backing up the money truck and
that's going to continue for awhile, it appears.
William (22:24):
Yeah, I mean, that's one thing in common with all of
these different everybody that'srunning or in the like, if you
will like, if we're thinking OSImodel, like the layer one of AI
infrastructure, likeeverything's got the commonality
of needing GPUs.
Um, yeah, it's interesting tosee how that'll play out.
(22:45):
Last thing that we that I haveon on my list is so I just
randomly came across this when Iwas what are they called?
Dim scrolling the otheryesterday, I think.
Um, tech giants see 150emission surge in three years
(23:06):
amid ai boom, and all the linksto these will be in the show
notes if you want to go back andyou know, read the.
You know from the source whatI'm, what I'm talking about here
.
But basically, this new uh unreport basically goes into
operational carbon emissions forthe world's top uh tech
companies and then looks at,like, how it rose an average of
(23:28):
150 percent between 2020 and2023.
And, of course, when you thinkabout that much for rise, you're
thinking, okay, cloud and aidata centers, so like.
Aside from these numbers beingstaggering, it notes that
amazon's operational emissionsgrew 182 percent, microsoft's by
(23:49):
155 percent, meta by 145 andalphabet by 138.
Good on you, alphabet, um, andthat's just three years.
So the un's internationaltelecommunication Union I think
I'm getting that right Directlylinks this major uptick to
(24:14):
recent breakthroughs in AI anddemand for digital services like
cloud computing, yada, yada,yada.
So the big thing here, my bigtakeaway, is these data centers
are driving like anunprecedented electricity demand
surge.
You know they consumed let melook at my notes 415 terawatt
(24:37):
hours of electricity in 2024,about 1.5% of global power
demand.
At current growth rates, datacenter electricity consumption
will hit 945 terawatt hours by2030.
And that surpasses Japan'sentire annual electricity
(24:59):
consumption, if you want to usethat as a baseline.
So what do you think, yvonne?
How are we going to get morepower?
Eyvonne (25:22):
How are we going to get more power, advanced deals or
interest in startups that aredoing small form nuclear, which
scares a lot of folks initially,but when you hear about how it
works, it's incredibly safe.
There's also, you know, a lotof work going in on in solar.
(25:43):
I think those are the two bigareas that are being actively
explored nuclear and solar.
We're going to need more power.
I think that's a reality.
Also, I think even fororganizations that have had a
deep commitment to trying to begreen or carbon neutral, to run
(26:04):
environmentally conscious datacenters, the demand has just
been so great that we've notfigured out how to do that 100%
yet, and I think that willhappen over time.
I also think right now, we'rein a political environment that
is not as strongly incentivizing, you know, green initiatives,
(26:27):
and I think that's going to havean impact as well.
I don't think that'll be along-term reality, but I think
for the next four to eight yearswe're going to see that, and
and so I, you know, I know a fewyoung folks that are in this
technology arena power energyagain, like I was talking about
(26:51):
physical plant and I think it isan incredible area for growth
and expertise and innovationthat we're going to see because
there's a demand for it, likeit's a problem we've got to
solve.
If we don't solve it, then youknow the advancements we're
seeing are just not going to beable to continue.
(27:13):
And if there's anything that Ibelieve in, it is the ability of
capitalists to solve problemsthat are preventing them from
increasing the revenue.
So I think we will havesolutions and we will develop
them and then iterate them andimprove on them over time.
William (27:34):
Kind of feels like we're trying to figure out how
to power the plane, as we'retaking off on the wrong way,
right.
Eyvonne (27:42):
Yeah, I think so, but I think that's always how it is
with innovation.
That's always how it is withinnovation.
You know it's.
If you look at the periods ofrapid innovation that have
happened throughout history, youknow, from from the time the
Wright brothers flew that firstplane until we had commercial
(28:04):
airlines, it was 50, 75 years.
I mean it just wasn't that bigof a delta of time.
When you think about it,especially now looking at the
innovations we've had at thatspace in the last 50 years, you
know, yeah, there have beenincremental improvements, but
not the slope of the curve ofthe innovation that happened
(28:26):
initially.
William (28:27):
I think that's always the case.
Eyvonne (28:28):
You know, you hear, see this rapid period of expansion
and innovation, and then we kindof hit a plateau where we move
to incremental improvement.
We've seen that with cellphones, for example.
You know, mobile devicesexploded onto the scene with
Steve Jobs and the iPhone in the2000s, and then 2010s, 2000s
(28:53):
it's all running together and wesaw rapid expansion.
And then we saw, you know,second order organizations,
companies, be developed, youknow, uber, lyft, all of those
things.
And, yes, now we see the AIinnovation.
But there are these hugeexpansions and then more
(29:14):
incremental improvement, andthat's typically the history of
how innovation has happened, andI think we're going to see the
same kind of thing.
It's just right.
Now AI is in that period ofrapid expansion that requires
more data, more power, period ofrapid expansion that requires
more data, more power.
Five to ten years will will.
That will stop and we will bein the incremental improvement
(29:36):
arena again, and then there'llbe the next thing like that's uh
, it's kind of how these thingsgo super, ai, super intelligence
.
Well, we're gonna add super ontoit, I think the question is do
the robots come in this wave orthe next wave?
That's what I want to know,because that is I need.
I need a robot in my house todo things, mostly laundry dishes
(30:01):
, cooking.
If I could, yeah.
William (30:03):
But the laundry and the dishes are like, seem very
doable by some of thedemonstrations I've seen, both
out of some of the robotics inchina and even like, uh, like
tesla optimus, um, it's pretty,pretty wild what, what some of
these robots can do, um, andhonestly I think if you you can
(30:26):
buy a robot for 15 grand, thatdoes a lot, and it does a lot
very well.
That's huge.
It changes the game right, itdoes change the game.
It frees up a lot of time, butI don't know how I would feel
about a robot walking around inmy home.
It just seems kind of I don'tknow.
Eyvonne (30:47):
I'm sure it might change, well, I mean we've all
seen our robot, you know, and wedon't want to have to terminate
Vicky, I think that was thename of the hive mind computer
in that movie.
But I think we're we're a waysaway, although we've seen some
self-preservation out of ourexisting systems, reservation
out of our existing AI systems.
(31:08):
But you know, I am willing totake the risk if I can get a
robot to help me with housework.
William (31:17):
Frankly, it's worth it.
I would agree on that one.
Okay, all right.
Well, I guess we both have hardstops coming up and this is,
yeah, exciting times.
And remember to resubscribewhen we do switch over on July
23rd.
Anything else, yvonne, beforewe jump.
Eyvonne (31:36):
No, we'll be sure we share all those details on the
socials.
Please.
If you haven't checked out thePacket Pushers Network, go check
out their family of podcasts.
There's all kinds ofinteresting information over
there and great folks and
If I was a dancing person, I would dance right now,
but I don't dance.
Eyvonne (00:04):
Yeah, I'm not a dancer either.
We didn't even dance at ourwedding.
William (00:08):
The Cloud Gambit is getting adopted into the Packet
Pushers family of podcasts.
Welcome to another episode ofthe Cloud Gambit.
(00:29):
And you know what's crazy,yvonne, I started this podcast
back in August of 2023.
And I think right around 2024,and I think it was in August.
It was almost like exactly ayear.
Eyvonne (00:46):
You joined as co-host, thank you, by the way yeah, it's
been fun, and does it feel likeit's been almost a year?
William (00:54):
it doesn't really feel.
It's one of the few things inmy life right now that doesn't
feel like a year like everythingelse, like work and other stuff
.
It's like, oh, I feel like I'vebeen doing this for 20 years.
But yeah, it's been, it's beengood, I really enjoyed it and
it's so much nicer to have apodcast and not be alone, to
(01:16):
have someone that you could talkto.
Just different perspectives,different opinions.
You know it's been great.
Today's going to be kind oflike a news episode.
We have lots of news, by theway, both about the show and
with some major news headlinesas well.
I guess, first things first,what's going on in the world of
(01:40):
Yvonne?
Are you staying cool?
Eyvonne (01:44):
No, not cool.
Well, fortunately everybody'sfamiliar with the shed at this
point and it is air-conditionedand I even have some wonderful
humidity controls, so you knowworkday is pretty comfortable.
But we are in the midst of aneastern northeastern heat wave
so it's been pretty warm here inthe south midwest northeast so
(02:05):
so it's been pretty pretty warmhere in um in the south midwest
northeast, so it's been a littletoasty.
Um.
Lots of, lots of fun life stuffgoing on.
I've got a kid who, by thewho's getting married.
It may be done by the time thiscomes out.
So we've been in the throes ofof wedding planning and summer
(02:25):
fun and kid activities and got alot coming up in the month of
July, including with the podcast.
So we're excited.
William (02:34):
There's a lot going on.
Yeah, planning a wedding is noeasy feat.
That can be a very busy andvery stressful time.
Eyvonne (02:44):
And it is something I can do.
But it is not a skill set thatcomes naturally.
So it has been, but it's beenfun.
You know, it's kind of a oncein a lifetime thing, so we're
embracing it for the fun that itis, and we'll look at the
credit card bill later.
William (03:01):
Yeah, I love it.
That's smart.
So yeah, yesterday.
So I'm kind of on a quasivacation.
I'm working from the northeast,uh, where my in-laws are, and I
spent somewhere around likenine hours in the sun with my
kids yesterday at this amusementpark, hershey park in
pennsylvania, and if you can'ttell in the video, I'm, I'm, I'm
(03:23):
cooked, uh, the heat index waslike 105 and you know the sweat
and steam that I producedpersonally probably power my
home for a week.
It was hot.
Eyvonne (03:34):
Sunscreen is your friend, william, and wear a hat
Next time you may send me goingon vacation.
I'm screened.
Yeah, I'm gonna.
I going to have to start my uh,my worst sunscreen and wear a
hat, mantra.
William (03:48):
Yeah, I mean, it's pretty bad.
So my, my kid, he doesn'tcomplain.
He's nine years old and youknow, this is like.
The only thing he said all daywas like it was probably about
15 minutes in and we were justwalking through the park and he
looks at me and he's like dad in.
And we were just walkingthrough the park and he looks at
me and he's like dad.
(04:08):
This is like a different kindof hot.
This is really really hot, Iguess it is.
Let me tell you and it's.
Eyvonne (04:13):
it's been a very wet season, at least here, for us as
well, and so the humidity isalso, uh, through the roof, and
we're, of course, trying to tendto the yard, because there's an
outdoor wedding in our future,and, yeah, it's been a fun
summer.
William (04:28):
So the first big piece of news that we have is the
cloud game, but is gettingadopted into the Packet Pushers
family of podcasts.
Now, for those of you thatdon't know about Packet Pushers,
which it would be hard tobelieve, but they were founded
in I think it was 2010, by GregFarrow and Ethan Banks with the
(04:51):
goal of really creating aplatform where just technical
professionals could discusstechnology, share practical
knowledge like real-world stuffand really, I guess, explore
like industry type developments,and over the years, I think
they've solidified, you know,their reputation as a uh, a
(05:13):
leading voice in like enterpriseit media, if you will.
So their network of podcast hasalso evolved by adding, uh,
lots of new shows over the yearsthat specialize in different
areas.
So what does this mean, just atthe top, for the cloud gamut,
like what's going to happen?
(05:34):
So we will both stay aboard ashosts, yvonne and myself um, the
the transition, like the finaltransition, happens July 23rd.
So on July 23rd, that will bethe first episode under the
Packet Pushers family and theshow will continue to be
(05:58):
released on a bi-weekly basis.
It does move to oh boy,wednesday, yes, wednesday, okay,
so from Tuesday to Wednesday,thank you, yvonne, and the RSS
feed will change, so you willneed to resubscribe on your
preferred catcher when the timecomes, and we will send out a
(06:21):
reminder as well.
Am I missing anything?
Eyvonne (06:27):
I don't think so.
It's exciting for me becausePacket Pushers has been kind of
a stalwart technology source ofinformation for me since I
started in enterprise IT.
You know I started my firstenterprise role in 2010 and then
(06:49):
, you know, discovered thepodcast within a year or two of
that, met the Packet Pushersfolks out at Cisco Live and got
to be on the show back in theearly days and really that
community from you know, the2010s really helped me grow and
(07:10):
fuel my career, and so it'sreally exciting to to be part of
their network and to beassociated with everything from
you know network break.
They've got their heavynetworking podcast.
They also have strategy andleadership and operations
(07:30):
conversations.
So I think we you know ourcontent fills a unique slot
there.
We both have networkingbackgrounds, but we are also
both firmly entrenched in thecloud world these days, um and
so just super excited about uh,uh, yeah, join, joining that
(07:50):
network and being a part of thePacket Pushers world.
William (07:55):
Yes, If I was.
If I was a dancing person, Iwould dance right now, but I
don't dance.
Eyvonne (08:02):
I'm not a dancer either .
We didn't even dance at ourwedding.
Like we're just.
William (08:07):
I didn't either.
Eyvonne (08:08):
Yeah, I sometimes look at my friends and colleagues who
have a cultural background ofdancing and body movement with
envy and all.
But yeah, that's not the worldI came, came from and I'm too
old to fix it now yeah,sometimes I dance.
William (08:30):
I guess I dance and sing in the shower, um,
sometimes.
But yeah, that's a shower.
Time is reserved for my, mysinging antics.
Yeah, I try not to take itoutside of the you know privacy
of my own home definitely I'm anappreciator of music.
Eyvonne (08:45):
I'll say that.
Said that way.
William (08:48):
So, yeah, again, we're excited and we will send out
another reminder closer to thetime when the transition happens
.
And that is not the only bignews that has happened.
Um, recently there's actuallybeen a lot of stuff too much to
talk about.
I took three, I think, two orthree stories that I think are
(09:12):
uh, pretty, pretty noteworthy.
I guess I'll go through thefirst one.
This, basically and this is abig one this is actually
something I called a while back,which I never call anything
like lately, I feel like which Inever call anything Like lately
.
I feel like I don't callanything correctly because the
landscape is just very different.
I feel like it's a surpriseevery week, to be honest.
(09:34):
But Google Cloud has basicallydonated its agent to agent or 8A
protocol to the LinuxFoundation, which is a great
place.
I love this decision and itbasically it's creating an
independent project.
You know, that's backed byother big names, other big tech
(09:54):
giants, like you know, aws,cisco's, microsoft's, like
Salesforce's of the world, andto me this is noteworthy.
I mean, there's a lot ofreasons.
I could probably ramble.
I'm going to try not to ramble,um, but I think the first is it
really represents that crucialstep towards, uh, I'd say like
(10:15):
standardization, like in thisemerging crazy ai centric agent
based economy, ai-centricagent-based economy.
So this A2A protocol, basicallywhat it does is it provides a
single or just a common languagefor AI agents to discover
(10:36):
capabilities and exchangeinformation and then allows them
to coordinate very complextasks across different third
parties, platforms, vendors, etcetera, which is huge.
And I think the second thing isit really again it breaks down
(10:58):
the silos that I think havelimited AI's potential I'll just
say it that way Especially forenterprise adoption, because it
creates an open standard under aneutral governance.
So the industry isacknowledging that agent
interoperability is far tooimportant to be controlled,
(11:21):
obviously, by a single company.
And I think, third is probablythe timing is just very
strategic.
So with, I mean, I'd say likeover what?
Like 50, 100 companies nowsupporting the protocol, like
including AWS and Cisco, youknow as it's, you know more
(11:44):
recent validators, we're seeingthe industry coalesce around a
like a real standard, just asagent technology is beginning to
really proliferate.
So, yeah, I think it's thedonation to the Linux Foundation
ensures, you know, at thiscritical time, that this
critical component remains, youknow, vendor agnostic and
(12:05):
community driven.
Any thoughts from you, yvonne?
Eyvonne (12:10):
Well, I think it's a really important step and if you
look back through the historyof our technologies, you know if
you go all the way back to theInternet of our technologies,
you know if you go all the wayback to the internet we've
always had standards and publicorganizations that help dictate
and drive interoperability oftechnology.
(12:32):
And when that's happened,that's when we've seen the most
rapid adoption andubiquitousness of that
technology, whether you're goingback and talking about
networking protocols, ieee andthose types of organizations, or
(12:52):
whether you're talking aboutKubernetes and how it was open
sourced and I believe thathappened in 2014, if my notes
are correct.
Yes, the open source ofKubernetes and what we've seen
there with the containers, cloudnative technologies, container
(13:13):
orchestration and I think thething that we know is that even
when a technology is opensourced, there's plenty of room
in the marketplace for vendorsto create add value, support
that technology.
You know, we've seen that.
You know, in the decade plussince Kubernetes was released,
(13:37):
you know we've got every majorcloud provider has their own
Kubernetes engine.
We've got, you know,on-premises solutions for that
as well.
But also it provides afoundation for interoperability
and also gives enterprises afirm foundation of understanding
(14:04):
what that technology is goingto look like and it solidifies
it.
I think one of the things thatGoogle donating A2A to the Linux
Foundation is going to do isalso going to solidify that that
technology is here to stay,that it's a real thing, that it
can be relied on, and it willbecome the central focal point
(14:29):
for how AI agents and systemsinteract with one another, and
that will go a long way tohelping us advance the
technology in a way that folksdon't feel so locked in or
beholden to one particularvendor, because I think right
now that's a bit of what we'vegot right Like because the ages
(14:54):
don't interoperate well andthere's not yet an ecosystem.
I've said a lot of words, butthat's really what I'm getting
to is that this is going to helpan ecosystem build up around AI
technology and agents so thatthey can interoperate and
interact.
There'll be bumps and there'llbe challenges, but it's a great
(15:15):
step.
William (15:16):
Yeah, those are all great points and the
interoperability is so huge andI think before A2A and MCP
specifically, you basically hada scenario where you had to
create one-off If you wanted tointegrate.
It was like for each major LLMyou had single integrations for
whatever model, back to whatevertechnology.
(15:37):
You're working with guardrailsand the sort of locking things
down and keeping things safebecame very challenging and you
had so much custom stuff.
So these two really flip thescript there and make AI broadly
accessible and adoptable in asecure way that we expect to
(15:57):
have to be able to adopt themwith, I guess.
On to the next thing thing, andthis is another big one.
So on to the next news story.
Uh, I'm just gonna read theheadline hold on, this is a big
one.
So, um, open ai taps googleunprecedented cloud deal,
despite ai rivalry sources sayso.
(16:20):
I thought this was interestingand many, many would call this a
very surprising twist becauseyou know, open ai well, let's
just start here.
So open ai, you know, hasbasically, according to the news
, finalized a deal to use googlecloud like services, kind of in
the way that they usedMicrosoft Azure Compute for its
(16:43):
growing compute needs, andanybody that knows any of these
companies will know that Gemini,google, they're all kind of
competing, if you could call it.
I guess it is competing in theAI race right now.
So there is some intenserivalry between some of these
companies and this dealreportedly finalized in.
(17:07):
May and marks a significantshift in competitive dynamics in
the AI industry.
So why does this matter?
So just my thoughts on it is itreally highlights just this
staggering compute demand ofbuilding AI in 2025.
(17:27):
So OpenAI's decision to reallydiversify beyond Microsoft Azure
from its earlier days reallyunderscores how absolutely
massive these computingrequirements for training and
deploying AI models are sort ofreshaping business relationships
(17:51):
.
So even on OpenAI's annualizedrevenue seven numbers here in my
notes so it's surged to 10billion.
So its infrastructure needscontinue to basically outpace
what a single, you know cloudprovider can probably deliver in
the time that they need it, Iwould assume.
(18:12):
Uh, and that the second thing isum and this is just something
that comes top of mind is itkind of represents probably a
strategic pivot for OpenAI tosort of reduce dependency on
Microsoft at the end of the day?
Since ChatGVT launched late2022.
(18:33):
Openai has sort of been makingdifferent moves to kind of
diversify its infrastructure,like that computational
substrate, including the I thinkit was like 500 billion
Stargate data center project.
I think SoftBank and Oraclewere involved with that.
One Deals with companies likeCoreWeave and now Google Cloud.
(18:57):
So you know, that screams to mejust diversification, prevent
lock-in and just kind of liketrying to think ahead for future
capacity demands.
Any thoughts, yvonne?
Eyvonne (19:12):
Well, the winner in all this is NVIDIA.
You know they continue to winright now and it's.
You know we were talking earlyin the show about kind of our
career trajectories andtechnically kind of our career
(19:35):
trajectories and technically.
And one of the things that youknow, back in my very first job
at a mom and pop ISP again, Iwas, you know, 19, 20, they made
us build our own PCs, right,and so they ordered the hardware
and, like part of us gettingfamiliar with our environment
was like you're going toassemble your own hardware,
build your own PC.
(19:55):
And I remember, you know, therewas a time when it was and there
was a significant growth inhardware and hardware needs and
(20:18):
then we shifted into a point intime in the industry where we
didn't really care as much abouthardware.
You know everything was prettymuch standardized on x86.
You had some outliers there, youhad your AS400s, you had your,
you know, some customizedcompute systems, but we really
just sort of ordered the nextfastest thing.
The next time the refresh cyclecame around and it didn't
(20:40):
really impact what we werebuilding or how we were building
.
Well, now we've shifted thependulumist one back again and
now we really care abouthardware and right now, compute
processing power is really arate limiting factor for
(21:03):
everybody in the AI space, and Ithink you know what this move
signals is how important thatcompute capacity is and how it's
creating, maybe, relationshipsthat we would not have imagined,
you know, even 18 months ago,because there is such a
(21:25):
dependence on powerful computeand what that's going to mean
for the future of infrastructure.
What's it going to mean forpower generation, what's it
going to mean for data centers,tooling, all of those things
Like there's a.
You know there's a five to 10year trajectory, I think, where
(21:48):
physical plant is going to againbecome incredibly important,
where it's just been a thingyou've had to have for the last
decade.
William (21:57):
It's not going to be that way.
Moving forward.
Eyvonne (21:59):
And so there's some significant strategic advantage
to folks who figure that out andwho are innovating in that
particular space.
William (22:09):
Yeah, good points and I agree, nvidia is like one of
these big casinos at this point.
Eyvonne (22:16):
Yeah, they're just backing up the money truck and
that's going to continue for awhile, it appears.
William (22:24):
Yeah, I mean, that's one thing in common with all of
these different everybody that'srunning or in the like, if you
will like, if we're thinking OSImodel, like the layer one of AI
infrastructure, likeeverything's got the commonality
of needing GPUs.
Um, yeah, it's interesting tosee how that'll play out.
(22:45):
Last thing that we that I haveon on my list is so I just
randomly came across this when Iwas what are they called?
Dim scrolling the otheryesterday, I think.
Um, tech giants see 150emission surge in three years
(23:06):
amid ai boom, and all the linksto these will be in the show
notes if you want to go back andyou know, read the.
You know from the source whatI'm, what I'm talking about here
.
But basically, this new uh unreport basically goes into
operational carbon emissions forthe world's top uh tech
companies and then looks at,like, how it rose an average of
(23:28):
150 percent between 2020 and2023.
And, of course, when you thinkabout that much for rise, you're
thinking, okay, cloud and aidata centers, so like.
Aside from these numbers beingstaggering, it notes that
amazon's operational emissionsgrew 182 percent, microsoft's by
(23:49):
155 percent, meta by 145 andalphabet by 138.
Good on you, alphabet, um, andthat's just three years.
So the un's internationaltelecommunication Union I think
I'm getting that right Directlylinks this major uptick to
(24:14):
recent breakthroughs in AI anddemand for digital services like
cloud computing, yada, yada,yada.
So the big thing here, my bigtakeaway, is these data centers
are driving like anunprecedented electricity demand
surge.
You know they consumed let melook at my notes 415 terawatt
(24:37):
hours of electricity in 2024,about 1.5% of global power
demand.
At current growth rates, datacenter electricity consumption
will hit 945 terawatt hours by2030.
And that surpasses Japan'sentire annual electricity
(24:59):
consumption, if you want to usethat as a baseline.
So what do you think, yvonne?
How are we going to get morepower?
Eyvonne (25:22):
How are we going to get more power, advanced deals or
interest in startups that aredoing small form nuclear, which
scares a lot of folks initially,but when you hear about how it
works, it's incredibly safe.
There's also, you know, a lotof work going in on in solar.
(25:43):
I think those are the two bigareas that are being actively
explored nuclear and solar.
We're going to need more power.
I think that's a reality.
Also, I think even fororganizations that have had a
deep commitment to trying to begreen or carbon neutral, to run
(26:04):
environmentally conscious datacenters, the demand has just
been so great that we've notfigured out how to do that 100%
yet, and I think that willhappen over time.
I also think right now, we'rein a political environment that
is not as strongly incentivizing, you know, green initiatives,
(26:27):
and I think that's going to havean impact as well.
I don't think that'll be along-term reality, but I think
for the next four to eight yearswe're going to see that, and
and so I, you know, I know a fewyoung folks that are in this
technology arena power energyagain, like I was talking about
(26:51):
physical plant and I think it isan incredible area for growth
and expertise and innovationthat we're going to see because
there's a demand for it, likeit's a problem we've got to
solve.
If we don't solve it, then youknow the advancements we're
seeing are just not going to beable to continue.
(27:13):
And if there's anything that Ibelieve in, it is the ability of
capitalists to solve problemsthat are preventing them from
increasing the revenue.
So I think we will havesolutions and we will develop
them and then iterate them andimprove on them over time.
William (27:34):
Kind of feels like we're trying to figure out how
to power the plane, as we'retaking off on the wrong way,
right.
Eyvonne (27:42):
Yeah, I think so, but I think that's always how it is
with innovation.
That's always how it is withinnovation.
You know it's.
If you look at the periods ofrapid innovation that have
happened throughout history, youknow, from from the time the
Wright brothers flew that firstplane until we had commercial
(28:04):
airlines, it was 50, 75 years.
I mean it just wasn't that bigof a delta of time.
When you think about it,especially now looking at the
innovations we've had at thatspace in the last 50 years, you
know, yeah, there have beenincremental improvements, but
not the slope of the curve ofthe innovation that happened
(28:26):
initially.
William (28:27):
I think that's always the case.
Eyvonne (28:28):
You know, you hear, see this rapid period of expansion
and innovation, and then we kindof hit a plateau where we move
to incremental improvement.
We've seen that with cellphones, for example.
You know, mobile devicesexploded onto the scene with
Steve Jobs and the iPhone in the2000s, and then 2010s, 2000s
(28:53):
it's all running together and wesaw rapid expansion.
And then we saw, you know,second order organizations,
companies, be developed, youknow, uber, lyft, all of those
things.
And, yes, now we see the AIinnovation.
But there are these hugeexpansions and then more
(29:14):
incremental improvement, andthat's typically the history of
how innovation has happened, andI think we're going to see the
same kind of thing.
It's just right.
Now AI is in that period ofrapid expansion that requires
more data, more power, period ofrapid expansion that requires
more data, more power.
Five to ten years will will.
That will stop and we will bein the incremental improvement
(29:36):
arena again, and then there'llbe the next thing like that's uh
, it's kind of how these thingsgo super, ai, super intelligence
.
Well, we're gonna add super ontoit, I think the question is do
the robots come in this wave orthe next wave?
That's what I want to know,because that is I need.
I need a robot in my house todo things, mostly laundry dishes
(30:01):
, cooking.
If I could, yeah.
William (30:03):
But the laundry and the dishes are like, seem very
doable by some of thedemonstrations I've seen, both
out of some of the robotics inchina and even like, uh, like
tesla optimus, um, it's pretty,pretty wild what, what some of
these robots can do, um, andhonestly I think if you you can
(30:26):
buy a robot for 15 grand, thatdoes a lot, and it does a lot
very well.
That's huge.
It changes the game right, itdoes change the game.
It frees up a lot of time, butI don't know how I would feel
about a robot walking around inmy home.
It just seems kind of I don'tknow.
Eyvonne (30:47):
I'm sure it might change, well, I mean we've all
seen our robot, you know, and wedon't want to have to terminate
Vicky, I think that was thename of the hive mind computer
in that movie.
But I think we're we're a waysaway, although we've seen some
self-preservation out of ourexisting systems, reservation
out of our existing AI systems.
(31:08):
But you know, I am willing totake the risk if I can get a
robot to help me with housework.
William (31:17):
Frankly, it's worth it.
I would agree on that one.
Okay, all right.
Well, I guess we both have hardstops coming up and this is,
yeah, exciting times.
And remember to resubscribewhen we do switch over on July
23rd.
Anything else, yvonne, beforewe jump.
Eyvonne (31:36):
No, we'll be sure we share all those details on the
socials.
Please.
If you haven't checked out thePacket Pushers Network, go check
out their family of podcasts.
There's all kinds ofinteresting information over
there and great folks and
Popular Podcasts
Fudd Around And Find Out
UConn basketball star Azzi Fudd brings her championship swag to iHeart Women’s Sports with Fudd Around and Find Out, a weekly podcast that takes fans along for the ride as Azzi spends her final year of college trying to reclaim the National Championship and prepare to be a first round WNBA draft pick. Ever wonder what it’s like to be a world-class athlete in the public spotlight while still managing schoolwork, friendships and family time? It’s time to Fudd Around and Find Out!
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.