The S-Curve and More – DevOps Experience 2024

Hi, my name is Mark Callahan. I'm the founder and CEO of Cloud Canaries. I wanna do a presentation to talk to you about DevOps in the intelligence era.

It's here. We're gonna talk about new technology and how that affects and creates new culture and new attitudes. Ultimately, it's learning how to love the s-curve and more.

Little bit of a little bio for me. You can certainly pull down the slides and take a look at it. Um, I like to, uh, uh, crush pillars.

I like the clouds. I like canaries. I love Boston.

I love to row, I love beacons and beacons are workload data. Got a degree from MIT and so on. At the end of the day, when is it time to let new technology out of the Faraday cage?

In this tech case, it's a robot, and there's an 85% chance if you let the robot out. 0, 1% chance that it'll take over the world. The question is, what's your decision?

What are you going to do? Are you gonna take a pass or you're gonna embrace it? And sometimes when you do embrace this, there's always risk, but there's always opportunity too.

And rewards, sometimes it's unexpected. So let's go to our next slide here. One of the things that when you look at new technologies, you look for something called the inflection point.

And, and that h helps you to stay ahead of the scur. And we'll talk about the s-curve as well. Basically, it's when a new technology dramatically changes the trajectory of a business, of an industry or an economy.

But beware new technology is disruptive. This is especially true for DevOps. New technology can create new culture, new attitudes, a new way of solving problems, but also can disrupt and it will disrupt the status quo.

Eliminating old approaches, business models, and require new insights. Be prepared. The technology era will do all of these new technology can bring new risks.

Here's a little slide about, oh boy, the robots replaced you too. It's kind of funny. You notice the obsolete goals, uh, stamp on the, uh, the robot standing in line, uh, for un belong benefits.

However, new technology can bring new rewards. I know everybody, everybody's commute became much more pleasant since driverless cars. But this is just too much.

And if you notice, there's someone in a driverless car swimming. And the point is, is that sometimes the new rewards are unexpected. How can that happen?

Completely surprise. So again, be prepared for that. So let's, uh, jump into the, uh, um, how to love the S-curve.

The S-curve is a, basically a product lifecycle where it shows you how a new product gains strength, gains adoption becomes, uh, um, has a high level of growth and eventually stabilizes. Beware of s-curves, multiple s curves and, and multiple technologies and how they interact. The classic example that I learned at, uh, at MIT was schooner wind based freighters, hundreds of years.

They were used to, uh, transport freight all over the world, very efficient. They used the wind. Um, they were limited by weather.

Uh, but the technology for sales was incredibly advanced. And then it happened. Technology B came out.

The steamboat steam engines and coal for fuel new freighter designs bigger, could, could move more freight shipping's not limited by weather. The technology for steam engines advanced very quickly. Again, this, this, uh, chart on the, on the right shows technology A Skinner, and then technology B, which was a steam engine.

So, uh, which technology was at the end of its curve? Well, it was sales. You know, sales, sales sailboats still exist.

Uh, but with, for freight, for moving freight, within a decade or two, most of those schooner were replaced by steam engine based freighters. It was at the beginning of its s-curve, while cell technology was at the end of its. So, new technology replaces old technology.

It has an impact on how you, how you do your, your daily work. It has a impact in culture. It has a impact on, on new products and, and, and how they are successful to help at the end of the day, customers.

So let's dive into the scur. There's different phases for the S group. And, and you can use this, there's two kind of views of it.

I have a, an emotional, uh, the emotional phases in the economic phases. I like the emotional ones at the bottom. Uh, okay, it's still early, but where's the traction?

And then the next one is, this thing is not going anywhere. And then you see this little inflection point where the, the curve like goes up and, and the, the emotional response is, okay, maybe it's not hopeless. And then it goes up, uh, uh, skyrockets, it goes, uh, gains speed, uh, dramatically.

And the comment is, we're all gonna be rich. And then it kind of levels off. And, and the comment is, what just happened?

And at the end, we are doomed. So that's a, a typical s-curve, and you'll see that in many different technologies. You s saw that with schooner, and you saw that with, uh, steam engines and, uh, steamboats.

You see that with televisions, you'd see it, you know, um, audio equipment, you name it. One technology, uh, has a, a market lead, and it's replaced by another technology, um, that evolves in its its own scur. Um, the, the economic phases are, um, you know, you probably, if you've been to business school, you probably all, uh, have seen these.

But you know, you search for a solution, proof of concept, early adapters, system integration, and the market expansion. Again, different phases of that scur. So now you know what an scur is.

The, the new technology of the intelligence era. We believe at Cloud canaries is data, AI and compute. We're gonna go into each little piece, um, in a second here.

And we also believe we're at an inflection point, uh, in the scur. You know, that little space right before for, oh my gosh, is this ever gonna work? To, hey, it's working, and then it takes off.

And I, we're at that point, and it's really, as I mentioned, three pieces. It's the data and it's ai, but I'm gonna call it commoditized AI in the sense that it's no longer a research tool for research labs. You can now find it in the production floor, even in the cloud.

Um, and numerous vendors are developing their own set of tools, and they're widely available. And, um, anyone can, can really use them and actually build models that generate forecast and insights. The third piece here is, and is critical or commoditized.

Anything ai in this case, you need something that, that always pushes costs down. And Moore's law, which isn't really a law, it's more of an observation, is about how the density of integrated circuits double every two years, two and a half years. It kind of varies.

It's varied over the last 40 years. But the bottom line is computers are always gonna get faster. They're always gonna have more memory, and they're always going to the price point on, on that level of compute and memory is always gonna go down.

So what that's creates is this new technology where you can build models, you can do forecasts, and you can reduce costs. So take a look at your organization. Uh, what stage, uh, best describes your organization and current adoption of ai.

Take a, you know, think about that. No investment. We have a little bit over 20% of planning about using AI solutions of some type exploration.

A little bit more. 24% pilot projects are still under 15% partial integration. You know, AI's being used today, 20% and full integration still kind of like at 7%.

So there's a ways to go and we kind of fit right into that inflection point. So, um, it's only gonna get, there's only gonna be more. It's going to speed up and, uh, really have an impact on everyone's lives, but it's also gonna have an impact on, on, on DevOps.

One of the things I, I did wanna drill into a little bit is commoditized ai. Again, this is a situation where the tools and the technology has developed to a point where I call it the average Joe and Jane developer can actually easily get tools, build models, and actually use them to generate forecasts and, um, insight. And a, again, it's really driven by the availability of data, quality data, uh, the organization of that data and vast quantities, but also the cost of compute.

And as we've seen, the cost of compute is going down, there's more tools available. So there, it's easier to pick something that really matches your, your, um, solution area and, and data, right? Um, the rocket fuel for AI is data in compute.

You kind of mentioned that it's, you need data to build your models, but you also need compute. And the more compute, the faster, the bigger, the better. The, the easier it is to actually build AI models that actually produce, you know, unexpected amazing results.

And there's a little, uh, blurb here from, uh, Steve Brown, which is a very in interesting individual. Um, data properly cleaned and organized is the rocky full fuel of tomorrow's powerful AI solutions, smart services, new customer experience, and the AI powered tools and intelligent agents that will augment your employees and give them superhero level capabilities and simultaneously boost their job at satisfaction. Um, Google him, he's, he's, uh, he has a lot of other really cool things to say.

Um, data. Data and more data key attributes of data for the intelligence era is quality. How that organ, how that data is organized, and the quantity of data.

So a lot of these, you know, the, the, the quality in organization, it might be a little bit different than what you, uh, would, uh, need if you were to actually use the data. So the data has to be structured in a way that can be easily loaded into, into, uh, or fed into, um, tools that are generating models. And, and for the most part, more the merrier.

We mentioned this before. Um, compute is the driver of cost along with data and, and how that data is organized and cleaned. But with it, it will always, the cost of a AI based solutions will always be going down in the future.

'cause compute increases, storage increases, it's just easier to do things. In some ways. Brute force been using what I call fancy algorithms that over a few years become irrelevant because computers compute is just, makes 'em irrelevant.

This all brings, comes together in, in the data AI to compute lifecycle. And this is really cool. These, you collect data, um, from some source for us at Cloud Canaries, it's workload data.

Um, you take that DA collected data and you create a model. You use that model to generate forecast data and in and forecasted insights. Then you can use actual data to actually validate your forecast data and modify your model where appropriate based upon the validation.

And you can continue that. So your model is always gonna get better. Your insights are always gonna get better, and your forecasts are always gonna get better in this new intelligence zero.

And that's gonna have, again, have dramatic impact on DevOps and how you run your business. So, um, new technology requires new DevOps habits. Now, a lot of these are old op habits, however, it creates a new opportunity to really employ these habits in a way that really help, um, DevOps and your teams and the organization as a whole, you know, manage the cloud as your most important asset.

Become an interpreter of cloud insights. Collect lots of cloud data for modeling, forecast, cross lines the business to help, help and offer, uh, new solutions. Be proactive and ready with the new approaches and solutions.

Present your insight to your, uh, enterprise business decision makers using cloud data. Negotiate with insights from cloud data. Understand and participate in critical corporate initiatives.

Speak your com company's business language that's really important for DevOps. Pilot, evaluate, imagine new solutions. New technology requires new DevOps habits.

New technology gives you the opportunity to do this as well. One example for us at Aries is intelligent canaries is we use, uh, workload data, AI and compute, um, as a background and intelligent canaries are microsurfaces, we use billions of workloads each with data. We use artificial intelligence to create models.

And these models are sometimes so sophisticated. Even the neural scientists will go, well, they'll do the AI shrug. I don't know, how did it come up with that?

How did it come up with that insight? Well, you can figure it out, but it might take years and, and compute, um, Moore's law or observation that everything gets faster, everything gets larger, everything goes down in cost, data, AI and compute technology. And the S group.

So we're gonna take this one step further. Intelligent canaries are active observers. And, and, and, and this is how you know, technology A gets replaced by technology.

B, we have a, a solution, you know, a marketplace for observability and, uh, observability in the past has acquired instrumentation. Uh, if you observability in the future will not require observant instrumentation technology. B.

So over time, solutions will be replaced by newer solutions that again, will change the way DevOps works, the way DevOps thinks and, and will create new opportunities and some amazing unexpected results. As I just mentioned, um, existing observability solutions are obsolete. And you can, you can replace observability with others, uh, other solutions.

Um, because of this new technology, it completely changes the way that you will work and your culture and the way you interact with the business as a whole. Um, you know, log and trace data. Spend a look at workload, don't lie.

NEUR networks, compute observability, canaries without instrumentation. Um, what intelligent canaries can do, forecast visibility, troubleshoot effectiveness, SLA compliance, align alignment of metrics, both business metrics and cloud metrics. Um, the intelligent era will, uh, present, uh, new opportunities.

AI can be applied to the entire digital delivery lifecycle. Insight analysis portfolio and back. Um, backlog, continuous integration, continuous testing, continuous delivery.

AI will affect all of those. And again, it's, it will have impacts on, on, on DevOps as a whole. Remember that first slide, you have to make a decision whether you're gonna embrace new, new technology or not.

Um, I think it's better to pilot, evaluate, and then if it looks, makes sense, embrace, um, new technology, new culture, new attitude, pick the next cloud solution. And the bottom line here is that the, uh, this new era will basically replace well, all the solutions that you're currently using with a new set of solutions that are based on new technology. And we believe data, AI, and compute.

And we believe most legacy cloud solutions are obsolete or will be with this new technology. So pull a canary from a hat. Here are some, uh, uh, solutions that we believe in.

The intelligence zero will be replaced with AI based solutions that uses data and compute, digital experience, contract negotiation, crisis response, market analysis, sales forecasting, competitive analysis, buyer behavior monitoring, cost reduction, risk management, buyer behavior, continuous monitoring, relationship management, and many, many more. It will change DevOps in a way that will be interesting and will give DevOps new opportunities. So be prepared.

Are you ready to release or to open the door of the Faraday Cage and embrace the possible risk, but also the opportunity and rewards. Thank you very much. And let us know if you're want to open that cage.