2844: A Conversation About AI, Zero Trust, and Public Sector Transformation with Intel

[00:00:00] How do we navigate the rapidly evolving landscape of artificial intelligence, cyber security

[00:00:07] and the continuously evolving technology trends in a world that never sleeps?

[00:00:13] Well today here on Tech Talks Daily I'm privileged to welcome back to the show Steve Auren,

[00:00:19] federal CTO at Intel. And he's going to be shedding light on some of the most pressing questions

[00:00:25] that you might have been thinking of recently. And our journey today will take us deep into the

[00:00:29] heart of zero trust architecture. The transformative power of AI across every sector and by that I also

[00:00:37] mean the public sector, including government and the pivotal strategies that can drive successful

[00:00:43] innovation. And as we peer into the horizon of emerging technologies such as that concept of

[00:00:50] an AI PC and bringing that to life and bringing AI acceleration to the edge. I also want to

[00:00:56] explore Intel's strategic focus on scaling AI and security to meet the demands of these

[00:01:03] increasingly diverse environments. So today I invite you to join me as I unravel the complexities

[00:01:09] of these topics guided by Steve's insights and also talk about surrounding oneself with a team

[00:01:15] of smart individuals and that relentless pursuit of knowledge to stay ahead in the technology curve.

[00:01:20] So the only question I have left for you is are you ready to embark on this

[00:01:24] enlightening journey and discover how these innovations are not just shaping our digital landscape

[00:01:30] but revolutionising the way we address security, efficiency and indeed technological advancement?

[00:01:38] Hopefully you said yes. Before we get today's guests I'll know I need to give a quick shout out

[00:01:42] and a thank you to the sponsors of TechTalks daily this month. They are kite works and in a digital

[00:01:48] age where the landscape of remote work is ever expanding, the security and efficiency of your

[00:01:54] managed file transfer solution are paramount. This is where kite work sets a new benchmark for surpassing

[00:02:00] legacy MFT tools with its unparalleled security measures and user centric design. They've even been

[00:02:06] awarded the prestigious Fed Ramp Moderate Authorisation, a recognition that is not easily obtained and

[00:02:12] they've held it since 2017 by the Department of Defence. So what I'm trying to say here is don't let

[00:02:18] outdated technology dictate the safety and efficiency of your business operations and you can start

[00:02:23] your journey towards uncompromise security and unparalleled functionality today by visiting kiteworks.com

[00:02:29] that's kiteworks.com but now it's time to get on with the show and invite today's guest on

[00:02:36] so buckle up and hold on tight as I beam your ears all the way to the state side where Steve is going

[00:02:41] to join us once again to share his insults. So a massive warm welcome back to the show we last spoke

[00:02:48] 18 months ago, I don't know what happened there or how that happened but can you just remind

[00:02:53] the listeners that maybe missed our first conversation with just a little about who you are and

[00:02:58] what you do. Thank you Neil and it's great to be back here. I'm Steve Warren, I'm the Federal

[00:03:03] CTO for Intel Corporation and in that role I help the federal government and the public sector adopt

[00:03:10] technology and innovations to help their mission enterprise needs and also help Intel to best

[00:03:16] understand the requirements and the use cases that the government has. And speaking of use cases one

[00:03:23] of the huge topics right now is zero trust architecture. There's a few myths and misconceptions

[00:03:28] around it some people think that it is way too complicated or the same it's just something you

[00:03:33] cannot not have in the current climate so just to set the same for our conversation today could

[00:03:39] you just walk me through the initial steps organizations should maybe think about taking

[00:03:45] right at the beginning of their journey towards a zero trust architecture and also how does Intel

[00:03:50] approach that transition because it's one of those areas that people might not associate with

[00:03:55] Intel too. That's a really good question and there's a lot of confusion today around what does

[00:04:01] it mean to do zero trust and I think the starting point is understanding that it's not an end game

[00:04:06] it's not a product or technology deploy magically you have zero trust it is in fact a journey

[00:04:12] and it's an ongoing journey as we evolve and mature our security controls and our enterprise

[00:04:19] and capabilities to have a zero trust approach to the way we do things. And so the first step is

[00:04:25] to know that there are many paths to get on that journey and it starts with understanding

[00:04:31] your infrastructure. So step one is having a plan understanding what you need to do what your

[00:04:36] zero trust journey is going to look like and to get that ball rolling there are a couple of key

[00:04:41] recommendations that I often give to CIOs and CIOs are sort of where to begin. One of the first

[00:04:47] steps is asset inventory and understanding both the assets that you're managing and deploying

[00:04:53] and the resources or services that you're providing whether it be internally to your

[00:04:59] customers and employees or services out to your business partners understanding what those key

[00:05:05] assets and resources are because you can't secure what you don't know. And as it's core this is not

[00:05:12] daunting task. Most organizations have this data today they may be distributed across multiple

[00:05:17] databases and most multiple different business units but your your finance department knows what

[00:05:22] you spent money on either capitalizing all your hardware that your finance department knows what

[00:05:26] cloud services you're taking advantage of your business units in your operational environments know

[00:05:31] what services they're providing and so this information exists it's about gathering it together

[00:05:36] in order to be able to do that assessment that inventory to know what you need to protect what

[00:05:41] transactions need to be protected and authorized what devices and data sets and things that you have

[00:05:47] to control. And so step one is just getting that that asset inventory and understanding what you

[00:05:52] have and what the risk of those things are. From there a lot of organizations quickly move to

[00:05:58] multi-factor authentication and that is a key part of the story. You need to be able to do

[00:06:03] multi-factor authentication or two-factor authentication to be able to authenticate both the user

[00:06:08] and the user at that moment in time to achieve those zero trust tenants of continuous authorization

[00:06:14] default deny and continuous monitoring. But that's not where you stop but I think that's a mistake

[00:06:20] that a lot of organizations make is they get their multi-factor rolled out and they say but mission

[00:06:25] accomplished were done. And multi-factor is one key aspect but there's a lot of other things to go

[00:06:30] along for the ride in order to be able to achieve zero trust. In many organizations have already

[00:06:35] started deploying two-factor or multi-factor authentication. So the question is what do I do next?

[00:06:41] And there it's really looking at the data and understanding how do you protect the data throughout

[00:06:46] its life cycle at rest on systems in transit between systems and end users and while it's being

[00:06:53] operated upon in in use and being able to put together a plan and a strategy for protecting data

[00:06:58] through all of its stages from creation all the way through a deletion. Another key area that

[00:07:06] seems awfully daunting is the continuous monitoring. That's a scary one but it's not as scary

[00:07:12] as people often think. The key is to be able to do to change the way you approach your auditing,

[00:07:18] your monitoring, your logging and your event correlation from a moment in time or a point so whether

[00:07:24] it's six months every six months we audit our systems or in the development process we do some

[00:07:29] audits and then at deployment we do some audits and then like set every six months to changing that

[00:07:35] dynamic to an ongoing continuous monitoring. And one of the key tens of zero trust is this notion

[00:07:41] of a dynamic approach to risk. When I'm transacting with the system I need to be able to transact

[00:07:48] with that system but I'm doing it at a moment in time from a given system on a given network with

[00:07:52] a given authority. And being able to assess that at this moment in time and apply the appropriate

[00:07:58] authorizations based on the risk environment I'm currently at as opposed to a decision that was made

[00:08:04] six months ago or ten months ago when I first got access to that system it's changing that capability

[00:08:11] to be able to make that determination at the time of transaction and that is one of the key things

[00:08:15] that will help evolve the zero trust journey for organizations is switching from that point in time

[00:08:21] you know and then you know sort of every regular cadence of audits and monitoring to an

[00:08:26] ongoing continuous and applying that risk calculus at the time of transaction. And so those are

[00:08:34] some of the key building blocks there's a lot of guidance out there about where to begin and how

[00:08:37] to structure your plans. And I think the last advice before I talk about what Intel does is get out

[00:08:42] of the analysis phase and into the implementation phase. And we've seen a great example of the

[00:08:47] US government and the Department of Defense. The first executive order said you must have a plan

[00:08:51] and then the next piece of that is you must have a date to start implementing and then the next

[00:08:55] piece of that is start implementing with a goal of getting a certain level of maturity by a given

[00:09:01] date. And so they actually pushed to get out of the analysis phase because a lot of these kind of

[00:09:06] revolutionary or evolutionary approaches to security you can spend two to three years

[00:09:10] are analyzing the problem and creating the perfect risk scores and that's not the goal of zero

[00:09:15] trust. The goals of trust is to get that ball moving to increase your security,

[00:09:20] increase your risk mitigations and the only way to do that is to actually start applying it to real world

[00:09:26] systems and getting it out into your environment. Now Intel you know as everyone knows is a chip

[00:09:31] manufacturer we also provide produce a lot of software a lot of systems are hardware and software

[00:09:38] exist all across the the environment from clients to network boxes to the cloud to edge systems

[00:09:46] and one of the ways that we're helping is we're providing key security capabilities and

[00:09:49] controls into our hardware exposed to our software and ecosystem that organizations can take advantage

[00:09:56] of as they start their zero trust journey. And some of them are the obvious things like encryption

[00:10:00] acceleration or confidential computing for data confidential for while it's in use. And some

[00:10:06] of them are less obvious like accelerating the security controls that you need to do. One of the

[00:10:12] key benefits we've seen is not been in security realm at all it's in the data movement and IO

[00:10:17] side of the camp but when you're doing continuous monitoring and you're having to process you know

[00:10:22] terabytes of data to be able to do that continuous monitoring event. Having the right architecture

[00:10:27] that can move that data at speed to not be a bottleneck actually improves your security posture.

[00:10:33] And so Intel is providing these key ingredients and these key capabilities and services

[00:10:38] that didn't organizations can adopt as part of their zero trust journey. And so one way if you

[00:10:43] look at the NIST SP800-207 which is the zero trust architecture guidance one of the things they show

[00:10:50] is the all the different kinds of ways that the controls and implement are implemented by the

[00:10:55] the policy orchestrator and the policy administrator all those capabilities that you have from all

[00:11:00] your security ecosystem providers from your platform providers and hardware providers are basically

[00:11:05] controls that can be implemented in a zero trust way and that's where Intel strategy has been

[00:11:10] is taking our security capabilities and allowing organizations to implement them with you know

[00:11:16] tenants like default and I continuous authorization at a station so they can build them into the

[00:11:22] zero trust architecture. And when we last spoke in September 2022 the world was a very very different

[00:11:30] place we were blissfully unaware that the world was about to go crazy for Gen AI and

[00:11:35] AI although it had been around for many many years was about to enter the mainstream and businesses

[00:11:40] in just about every industry we're trying to find their place in the AI narrative and plus 4

[00:11:46] to 2024 with AI's integration into various sectors. The battleground of AI threats though and risks

[00:11:54] are ever expanding so a question I've got I asked from a security background here is what are the

[00:12:00] pressing AI threats that you see today and how are you Intel working to mitigate some of these

[00:12:06] risks. So Neil you hit it on the head the you know the AI craze has happened and whether it be

[00:12:11] a large language models and gendered AI or just the scaling of the classic deep learning algorithms

[00:12:18] into every aspect of our lives with every revolution comes really understanding that there are

[00:12:24] risk and threat and AI is no different every day we see new articles and new publications of threats

[00:12:30] and attacks. I think the way to look at it is that you look at this sort of the AI life cycle and

[00:12:35] understand there's the what's hot right now which is AI prompt injection and manipulation so these

[00:12:40] targeted attacks at large language models fuzzing them and attacking them to be able to get either

[00:12:47] exposed IP about what the data was what the AI was trained on or getting them to provide false

[00:12:53] answers or forget information they've learned with the hallucination and catastrophic for getting

[00:12:59] which I think is one of the best terms we've come up with as an industry is catastrophic for getting

[00:13:06] but those kind of targeted attacks at the front end of AI is where everyone is focused on and those

[00:13:11] are serious attacks but I also think the things that are often overlooked is corruption and biases

[00:13:17] and poisoning that can happen to the data sets that those AI's are trained on or especially in

[00:13:22] the case of large language models are always learning so every query is an opportunity for the AI

[00:13:29] to learn and so putting in the corrupted learning injections or queries being able to do poisoning

[00:13:38] attacks that can poison the data and poison the algorithm are also of key concern and lastly is how

[00:13:45] do I protect that model once it's deployed so when you think about a lot of the applications AI

[00:13:50] where there be an autonomous vehicle or an edge sensor in that device there's a set of algorithms

[00:13:57] that have been deployed to whether being a card avoid the tree or be able to recognize the tree

[00:14:02] in versus the shadow so that you don't run into it in a camera being able to detect if someone's trying

[00:14:07] to break into your house or any of those kind of use cases how do I protect that model from an

[00:14:13] adversary that wants to either corrupt it or spoof it or give it false information and so when we

[00:14:20] look at sort of how do we protect AI we have to think about it from the all aspects of the AI

[00:14:25] life cycle but just to hit to that key point I think right now the hot area of research and we

[00:14:31] there've been papers published there was a whole section of the DEF CON conference this past summer

[00:14:36] dedicated to attacking AI a lot of that focuses on this what you call the injection attacks and

[00:14:43] the query poisoning how Intel is working with the industry and the ecosystem to help mitigate

[00:14:49] these risks it comes down to providing the hardware security and software security capabilities

[00:14:55] to protect the the whole life cycle so it's when you're running your AI you'll be able to do

[00:15:00] secure AI or confidential AI by protecting the model and the inference it's doing whether that's

[00:15:06] deployed in the cloud or at the edge and using confidential computing is one of the key mechanisms

[00:15:11] that organizations and cloud providers are doing to protect that that last stage of the AI process

[00:15:16] to even earlier stages by being able to do attestation of the system make sure I don't I have a

[00:15:21] clean environment when I'm performing the labeling and the tuning and the model development and

[00:15:27] and giving you that life cycle approach so that can get a visibility from an enterprise perspective

[00:15:32] of the security and the trust in the AI systems and that's ultimately a question that we're seeing

[00:15:37] asked at the strategic level in organizations and by governments you know sometimes it's called

[00:15:42] ethical AI but really ethics course how do I trust this AI how do I make sure it's going to do

[00:15:47] the right things how do I make sure this this AI is going to be used in the right way and building

[00:15:54] the underlying infrastructure that can provide both the security confidentiality and integrity for

[00:15:59] that AI in those AI tools but also being able to attest what went into the box what were the data sets

[00:16:06] what were the where was this algorithm tuned what tools that were used and so that work holistic

[00:16:12] approach is what's going to be needed and you know at Intel we're providing those key foundational

[00:16:17] building blocks to help organizations build more trust worthy AIs.

[00:16:22] And when we've been scrolling down our new speeds over the last 18 months when we're hearing about

[00:16:28] AI it's all about the private sector and how they're leveraging AI and new software new tools etc

[00:16:34] but I think very often we missed the fact that AI's impact on the public sector is equally as

[00:16:39] profound if not more so so can you shine a light on this maybe share a few examples of how AI

[00:16:45] is revolutionising enterprise operations within government and public sector environments because

[00:16:50] there's a bit of a stereotype that this is this area is slow to adopt to the pace of change they

[00:16:57] haven't got the same financial backing etc but that is changing isn't it absolutely and we're seeing

[00:17:03] AI permeate all aspects of the public sector one thing to keep in mind is the public sector is actually

[00:17:09] not different from the private sector in many ways if you look at the broader government

[00:17:14] the largest healthcare provider is the veterans administration a government organization the

[00:17:20] largest one of the largest financial institutions is the IRS and the Medicaid Medicare systems

[00:17:26] that are processing payments for healthcare and so when you look at it they have a lot of

[00:17:31] the same challenges the same applications or same capabilities that you find in the private sector

[00:17:36] logistics you know UPS or FedEx knows how to get your package there on time the Finsletch

[00:17:41] just sticks out to get jeeps and supplies to wherever the army or the military needs it so they're

[00:17:46] dealing with a lot of the same problems and so the innovations we're seeing AI have an impact on

[00:17:51] in the private sector are absolutely being leveraged in the public sector just often not as publicly

[00:17:59] exposed or publicly talked about but we've seen you know things being applied whether it be object

[00:18:04] tracking and object recognition a lot of the facial recognition or not facial but you know human

[00:18:08] detection person detection object detection and tracking use cases across the government

[00:18:14] one example that forestry did a while ago was being able to detect blight in the national forces

[00:18:20] using an AI model and being able to put that on to a drone a set of drones that can fly through

[00:18:25] the forest at a much better efficiency than sending a lone forestry ranger to walk the path

[00:18:32] and so they were able to get really good information about the state of in this case blight

[00:18:37] or disease on the trees and in many cases be able to catch it early enough to be able to mitigate

[00:18:42] or be able to recognize where where it's been affecting so they can apply their resources more

[00:18:48] appropriately so we're seeing a lot of examples I think the other thing that people forget you know

[00:18:53] everyone looks at the sort of the cool mission side of you know whether it's the military with all

[00:18:56] the things that they do or other citizen services but on the back end the public sector is very

[00:19:02] much like the private sector you have HR you have finance you have capital budgets you have ERP

[00:19:07] applications you have a lot of documents and compliance and processes that all could benefit

[00:19:12] from large language models and the application of AI and we're seeing innovations and deployments

[00:19:17] of these capabilities into I call the non-sexy part of the business which is automating manual processes

[00:19:25] improving the user experience on enterprise resource funding applications providing

[00:19:30] analytic tools to make better decisions to reduce redundancy and and find cost efficiencies

[00:19:37] and so those things are actually having some of the bigger impacts one example was being able to

[00:19:41] take a manual document based process that used to take about six months of manual human people

[00:19:48] to go through and read documents and make sure they were in the right format and apply them to

[00:19:52] the right datasets and be able to automate that down to something like 18 minutes and so

[00:19:58] that that transformation had a significant impact to the bottom line of reduction in time to getting

[00:20:05] that process done but also saving time and energy on errors by having an automated tool be able to

[00:20:11] speed up that process and so we're seeing across the public sector in all aspects both the

[00:20:16] mission side as well as the back end enterprise AI is being transformed. I think it was marked

[00:20:21] way no one said that history never repeats itself but it does often rely I mean this is not the first

[00:20:27] time that we've seen tech disruptive businesses and indeed traditional business models and

[00:20:32] I suspect that you've learned a thing or two from the cycles to out your career too so with

[00:20:37] your extensive background in tech leadership and innovation what strategies do you think

[00:20:42] or will be crucial for driving successful product development and market introduction

[00:20:47] in today's tech landscape because you can't feel daunting challenging and overwhelming for

[00:20:51] business leaders but the reality is we've been here before haven't we absolutely and you'll

[00:20:57] at it's to that same arc twain quote it's not that we have to do something different it's applying the

[00:21:03] right processes and to use a quote from one of my favorite books good to grade is get the right

[00:21:09] people on the bus surround yourself with a diverse team that understands the business applications

[00:21:14] these systems and infrastructure that's getting to point on represents the customer represents

[00:21:20] the the compliance framework and get them all together as we start to adopt or look at these

[00:21:25] technology innovations you know from a product development perspective is if you understand the

[00:21:30] customer use case whether customers external or internal internal and give them the ability

[00:21:38] to drive your requirements it will make your the successful deployment that much easier because

[00:21:43] you'll already have it anticipated many of the concerns or many the needs of the very constituency

[00:21:49] that you're trying to address additionally planning for what is that environment you're going to

[00:21:53] have to live in so a lot of development happens you know in a lab or in a small set of of capabilities

[00:21:59] understanding what is needed for the scale what is needing to secure that capability at scale when

[00:22:05] you get outside the four walls of the research and development team those are some of the key

[00:22:10] recommendations in there nothing new I mean this is something we've been doing for years about how do

[00:22:15] you best adopt innovative technology into existing environments is bring the right people in the room

[00:22:22] early in the process and understand the deployment requirements at the beginning as opposed to let's

[00:22:28] build a really cool widget then figure out how do we force it down stream and the last thing is

[00:22:34] that is the flip side is that a lot of times business leaders want to analyze everything to death they

[00:22:38] want to understand what's the impact of this innovation and one of my key recommendations is find

[00:22:44] an application that is a good target and actually go implement and along the same lines don't take

[00:22:52] the most important application to your business don't create your business if something goes wrong

[00:22:56] but also at the same time don't take an application or environment that is meaningless it needs to be

[00:23:02] like I call that Goldilocks zone something that is impactful but not mission critical for that first

[00:23:07] innovation so that a when you're if it does have a problem or fails or there's issues it doesn't hurt

[00:23:13] the business or take things off line that are critical to your operations the same time if you

[00:23:19] are successful and actually has meaning to the organization they can see how it is scale it

[00:23:23] and so you need to find that sort of sweet spot of where to apply these innovative technologies in

[00:23:28] development and before we started recording the podcast today we were both geeking out a little bit

[00:23:33] on the concept of AI computing the speed of new processes and chips etc so looking ahead especially

[00:23:41] around some of the trends that we're seeing at the moment are there any other emerging technologies

[00:23:47] that you say that you see playing a pivotal role in federal IT solutions and is there any

[00:23:53] if you can share on how Intel is preparing to meet these future demands because the speed of

[00:23:58] change is just phenomenal and it is a lot of exciting things happening absolutely I think one of

[00:24:02] the key things and this came out a few months ago as we're seeing you know as AI is really becoming

[00:24:08] pervasive our new concept and technology space is the AI PC and it's really about extending AI

[00:24:16] all the way to where clients and users are operating and you know as we were talking about the

[00:24:22] application of AI note takers and AI summary tools in this environments where you're doing your work

[00:24:30] teams zoom what have you it's just one example of where AI is per you know at the at the very edge

[00:24:36] where we're operating you know we're using AI all the time and so the idea of having an AI PC

[00:24:41] a PC that has built-in acceleration for AI so that I can operate on AI at the edge and not be

[00:24:48] completely reliant on the cloud and the cloud it plays a key role as well but understanding that

[00:24:53] it's an end to end story and so we're seeing AI really become the foundation for a lot of the way

[00:24:58] that businesses and transactions are transforming and I think one of the really cool innovations

[00:25:03] of last year is this idea of having native acceleration and native AI capabilities all the way out

[00:25:10] at the client PC wow incredibly exciting stuff there and I've got a small on behalf of any

[00:25:18] technology strategists especially in the public sector that might be listening to our conversation

[00:25:23] today any advice that you would offer to them to maybe effectively leverage AI and cybersecurity

[00:25:30] solutions to better enhance their operations and security posture because there's a lot of good

[00:25:35] stuff coming along but there's also a fair amount of protection that needs to be put in place to

[00:25:39] protect that good stuff really absolutely and from a strategic perspective at the key is to make

[00:25:44] sure to build in those plans and requirements from the get go so as you're looking to deploy

[00:25:49] whether it be AI or other innovative technologies understanding both your requirements but also

[00:25:55] what are the tools you need to deploy with those technologies to properly secure them to meet your

[00:26:00] risk posture your risk needs and how do you deploy them at scale and I think you know from a

[00:26:05] strategic perspective it's looking holistically across the organization you know at the tactical

[00:26:10] level of people looking well how do I protect my application how do I protect the ins and outs of

[00:26:14] my of the data into my application but an enterprise level our strategic level it's understanding

[00:26:20] what is that full life cycle what you know follow the bit where does the data flow and so that's

[00:26:25] it's really thinking strategically on these tactical implementations and how do I build the

[00:26:30] right infrastructure how do I get the right resources to deploy AI at scale and so when we look

[00:26:38] strategically across organizations it's understanding how do you transition out of the research

[00:26:44] out of the development phase and transition it into scale and building the right infrastructures

[00:26:49] to meet the security posture of the organization is really going to be the challenge for technology

[00:26:54] strategies to think more holistically and to pull from different domains in order to achieve your

[00:26:59] mission and enterprise goals and as I said at the very beginning of our conversation today when

[00:27:04] we first spoke back in September 22 need of us could have predicted the scale of the AI explosion

[00:27:10] but now of course that genie is out of the bottle it ain't going back in so what's your vision for

[00:27:15] the future of Intel's initiatives in the federal space particularly regarding cyber security and AI

[00:27:21] and how do you see these technologies evolving to address both government and national security needs

[00:27:27] again he's a huge topic right now especially I think I was reading before we started the conversation

[00:27:32] that four billion people across the planet are we going we'll be going to the holes this year and

[00:27:38] technology is playing a big part in this isn't oh absolutely um and Neil at at its course how

[00:27:44] are we going to help the federal government scale their operations with AI and cyber security and

[00:27:51] really be able to protect against them the the kinds of threats and adversaries we're dealing with

[00:27:56] and at its core I think the key is that word scale it's you know when we think about four billion

[00:28:02] people going to the polls or the millions of people that are leveraging technology in their everyday

[00:28:07] lives um at its core is how do you do these things at the scale necessary to service those large

[00:28:14] constituencies and so I think that's where a lot of Intel's initiatives around around being able to

[00:28:19] do it at scale but also to do it at scale in a low cost or at least a TCO-improved environment

[00:28:26] and whether that be things like you know sort of doing better at data ingestion and data processing

[00:28:31] be able to move the data more efficiently be able to deal with the scale computing the hyper

[00:28:37] scalers all the way down to the edge and at its core that's where Intel is providing innovations

[00:28:42] up and down the stack from you know like we talked about briefly a moment ago the AI PC

[00:28:48] to building capabilities into the network infrastructure with new architectures for improving the

[00:28:53] way that data flows across the edge in cloud networks um with our infrastructure processing units

[00:29:00] that pushes a lot of that compute right up to the network edge all the way to how the hyper scalers are

[00:29:06] deploying quicker the capabilities to scale these applications and things like large language models

[00:29:11] that are huge and massive trillions of parameters need are consumed data and they need the processing

[00:29:18] in order to be able to do it and so where AI's you know I'm sorry Intel vision and the initiative

[00:29:24] around the future is providing the right technology to scale the out your compute at the right place

[00:29:29] and understanding that it is a heterogeneous computing architecture that will best benefit

[00:29:35] the enterprise and the global economy understand you need a particular kind of computing at the edge

[00:29:41] a particular kind of computing in the network and a different kind of computing in the cloud

[00:29:46] and having the right compute for the right workload and that's really at the core of Intel's vision

[00:29:50] of why we have CPUs and and accelerators across the spectrum from low power to hyper scaler

[00:29:58] and then looking out towards the future it's then looking at what custom capabilities are necessary

[00:30:04] to increase that scale increase that performance or increase that capability whether it be around

[00:30:09] security or data met data flow or AI and you know as Intel announced with our IDM 2.0 strategy the

[00:30:15] Intel foundry is the opportunity for customers and partners to come in and do custom hardware to

[00:30:22] meet their specific application needs when you start talking about those large scale environments

[00:30:27] and so being able to not just take a general sort of off the shelf but when you need it having

[00:30:32] the ability to customize to to meet your mission or meet your enterprise needs and I'm conscious

[00:30:38] as a couple of technologists we are fascinated and passionate about the future and where we're heading but

[00:30:43] if we look back from our one at none of us are able to achieve any degree of success without a

[00:30:48] little help along the way so to finish the podcast today I'd love to find out more if there was a

[00:30:52] particular person that you might be grateful towards helped you get you where you are in your career

[00:30:57] or maybe just saw something in you back in the day and invested a little time into you who would

[00:31:02] that person be and why be great to give them a little shout out to them sure so when I was very

[00:31:07] starting off in my career in doing cybersecurity startups I reached out to Bruce Schneier to get

[00:31:13] some advice on an algorithm that we were implementing and he really took me under his wing and was

[00:31:19] an absolute great mentor tell me understand not just you know how do we need to fix our

[00:31:24] algorithm in that particular product but understanding secure you know the the cryptography and

[00:31:29] security ecosystem at that time and was an absolute key mentor for both my first company and

[00:31:34] my second company when we were doing the startups in the 90s around data encryption on the edge

[00:31:41] and the client as well as looking at enterprise systems and security for mainframe systems so I'd

[00:31:46] like to get a shout out to Bruce Schneier those haven't read he's got a variety of books he puts out

[00:31:51] regular newsletters but he was an absolute crucial person in my career to help me better understand

[00:31:57] the environment better understand the techniques and the technologies that were being used at the time

[00:32:03] and really just took me under his wing and helped guide me to understand some of the nuance

[00:32:07] of the proper ways of implementing cryptography a proper waste implement uh public key and asymmetric

[00:32:14] techniques in order to protect data and protect users absolutely love that quick shout out to Bruce

[00:32:20] and finally looking forward of course you're right in the eye of the storm of all this

[00:32:25] cutting edge technology emerging technologies I've got to ask I'll be off of everybody else listing

[00:32:30] around the world how old were you so fragile how do you keep up to spade with all this well it is

[00:32:36] a fire hose there are some key things I look at uh tell me keep abreast of what's going on one of

[00:32:42] the best news feeds that I like is it's called the hacker news and what I really like about it

[00:32:47] is it's produced out of India and what's really cool about that is that oftentimes the news articles

[00:32:52] that they see just because there's so much further ahead in the time time zones I will get their

[00:32:57] feed in the morning and be three or four hours ahead of the US uh newswags as far as what are

[00:33:02] the new attacks what are the new vulnerabilities uh where the new research so I from a security

[00:33:07] perspective I love the hacker news uh like I said because it just gives me early insights into what's

[00:33:13] going to happen and then from there it's it's finding the key people who are publishing information

[00:33:18] and putting things out and putting you know getting those feeds um so people like uh Robert Hansen

[00:33:23] has a podcast as well as uh puts out regular articles uh that are just fascinating to read so

[00:33:28] finding what we call you know those key delcalse were the ones they're putting out really good content

[00:33:34] and make sure that you're on their feeds listening to podcasts like yours meals and listening

[00:33:37] from people in the industry is really how we uh keep up to speed with just the sheer volume of things

[00:33:43] that are going on um and then you know read uh you know I I've got a stack of books behind me here

[00:33:50] um and so I'm always looking for the next book on whether be AI or technology

[00:33:55] and then lastly it goes back to that quote from uh Jim Collins is good to great surround your people

[00:34:01] yourself with people smarter than you and listen to them and I have architects on my team

[00:34:06] with domains of experience and keep and and skills that I don't have and so I am leveraging their

[00:34:12] smart brains in order to keep abreast of everything from AI to networking protocols um and what's

[00:34:17] going to be happening in the next several months in those domains wow absolutely incredible so much

[00:34:23] food for thought and our conversation today and for anyone listening wanting to just dig a

[00:34:27] little bit deeper on anything we talked about obviously Intel is a huge website anyway in particular

[00:34:32] you'd like to point people so if you go to intel.com slash public sector uh you'll see all the

[00:34:38] content that we're putting out focused on the US government um and from there you can obviously

[00:34:43] link onto the broader intel and the capabilities we're doing around confidential computing in AI

[00:34:47] can all be linked off that page and if you want to follow my feed um I'm SO RIN on LinkedIn so

[00:34:53] so I'll have a link in as where you'll find me. Well so well I'll get make sure links are there for

[00:34:58] both of those places and so much I loved about our conversation today because we covered everything

[00:35:03] from the new background of AI threats and risks start in your journey towards a zero trust

[00:35:09] architecture and how AI is transforming the public sector enterprise and I think it's so important

[00:35:15] to shine a light on that but it's not all about AI and technology or AI PCs and all that shiny

[00:35:22] stuff that we get excited about it's also about getting the right people on the bus getting those

[00:35:26] diverse voices and the role of humans and collaborating with technology that's where the magic

[00:35:31] happens but thank you so much for a timely reminder of that and showing your insights today.

[00:35:36] I'm thinking you'll as a pleasure to come back and uh and be with you on the podcast today.

[00:35:40] One of the standout lines from Stephen our conversation today for me was it's about getting the right

[00:35:46] people on the bus. It's about surrounding yourself with a diverse team that understands the business

[00:35:52] applications, the systems, the infrastructure that he's getting deployed on that represents the

[00:35:56] customer, represents the compliance member and get them all together so we can start to adopt

[00:36:03] or look at these technology innovations. I think that answer there encapsulates everything that

[00:36:09] I love about Steve and our conversation today but as we draw our conversation to a close I think

[00:36:14] it's clear that the path of mastering AI cybersecurity and the latest technology trends is intricate

[00:36:21] and thrilling but what as we all embark on this zero trust journey harness the power of AI

[00:36:27] in government and steer through that sea of emerging technologies it does require a keen

[00:36:33] understanding of our needs a strategic approach to innovation and a unyielding commitment to

[00:36:39] learning and adaptation and adaptation so I can't thank Steve enough for sharing his wisdom

[00:36:44] today the value of collaboration diversity and expertise in navigating the complexities of

[00:36:50] today's tech environment each not all about the technology it's people right at the heart of this

[00:36:56] but what strategies will you be adopting to stay ahead in the ever evolving tech landscape

[00:37:00] how will you build your team of experts to navigate these challenges as always share your thoughts

[00:37:05] with me email me techblowgrytoroutlook.com twitter linked in instagram just at Neil C Hughes

[00:37:13] and remember the journey of innovation is continuous so journey not a destination

[00:37:19] and it's filled with learning adaptation and collaboration so let's keep this conversation going

[00:37:25] join me again tomorrow we'll do it all again with a different guest and a different topic

[00:37:30] thank you Felicity and until next time don't be a stranger