[Music] hi I'm stepen Powers your host for Foresters podcast what it means where we explore the latest market dynamics impacting Executives and their customers today I'm joined by forester's vice president of emerging technology Brian Hopkins to discuss our annual top emerging Technologies list for 2024 welcome Brian hi Stephen how you doing I'm doing well thanks for joining today sure thanks for inviting me of course we've been doing this list for a number of years years I think he what 10 15 let's talk about how the methodology changed this year versus previous years absolutely as you know a couple of years ago we really got serious about how we use data to find the top 10 or top 20 emerging Technologies and we improveed that this year I mean generally the methodology has been we we look at what analysts are writing about because they have a pulse on the market we look at what clients are asking about in searching forest.com and the thing is we really added over the last few years are what is research Academia doing in terms of the things that they're uh investigating and what is industry doing in terms of the places they invest in so we continued that methodology this year one of the things we added in this year I thought was really interesting is we looked at the Venture investment arms of The Magnificent Seven firms figuring that those mostly high-tech firms all have Venture investment arms and they're investing in the technologies that they believe are going to drive the future and that we added that element into the analysis this year and because of our rigor in how we've begun to use data over the last few years even more deeply and then adding in some analysis through some of the Gen tools that that we're building internally we're able to actually come up with some pretty unique insights this year about some emerging technologies that perhaps have been ignored in the past or we have said aren't ready for prime time that we think are this year so I'm happy to talk about those so let's talk about some of those insights specifically what are some of the general Trends we saw from this year's research absolutely as you know one of the things we do every year when we do this is we we all Source a lot of data not only uh in the analysis I just talked about but we also spend a lot of time talking to our analysts and spend time talking to clients and kind of pulling all this information in and I spend when when I do this work uh some time thinking like what are the big things happening what are the trends going on here that should really shape what we think are the most important emerging technology and this year is three things really popped out and you'll see those as we talk through the Technologies we selected first thing is there's been an awful lot of emphasis specifically in many of the AI technologies that we've chosen to think that it's all about task Automation and efficiency at a task level and to some extent it is but one of the things that we found is that this task automation is not so easily translated into job replacement right which is a lot of things people are scared about in terms of you know where's automation is automation going to take my job what does the future of work look like what does the future of my business look like and while we see an awful lot of task automation potential it's uh sometimes problematic to translate that into actually replacing people or even measuring bottom line benefits at a job level so that's one of the things we were thinking about when we we get to look at some of the top emerging Technologies the second thing we saw is that Boomer on gen that we covered last year in our top 10 reports has continued but it's also everyone's realizing that it's a lot more difficult to master than perhaps we initially thought when we looked at the magic of chat GPT and now we realize that actually creating applications that we can trust and that deliver exactly what we need to drive the value we want we can actually only do that in a smaller set of use cases right now today and that with a lot of application infrastructure and data and governance built around it so it's proving to be a lot more difficult and that's kind of exacerbated by the fact that the technology every week or every month seems to produce a new generation of itself that we have to keep up with the third thing that you'll see in our emerging Technologies is that the future really belongs to those with a foresight to secure the present so there's a lot of security related Technologies in our list this year and that's on purpose because one of the things we see as we go deeper into this kind of edge enabled AI enabled future is these tools are enabling Bad actors and expanding the surface around which our our businesses can be attacked so all these tools go both ways you can use them for good or you can use them for bad and so we think that that firms that really understand this it's a two-sided coin or a double-edged sword and invest in securing the the present will benefit as we accelerate in the future so you'll see a lot of those themes in our emerging Technologies as well Brian before we dig into the list can you just give us a quick reminder of how the list is organized absolutely the way that we organize a list and we've done this now a number of years in a row is we organize the Technologies by benefit Horizon and we don't organize it by maturity because we think the word benefit Horizon has a lot more meaning for clients because the one thing clients want to know about emerging technology is when is it going to benefit me when is it ready for me and so we have this idea of short mid and long-term benefit Horizons that we assign to these Technologies based on our assessment of what the average client is going to see in terms of benefit so so for instance a short-term benefit arise in technology of which we have four means that the average Forester client should expect to see a tangible return on their investment in one to two years which is pretty short term the midterm benefit Horizon are things that are two to five years out and we think the average Forester client should see expected value in that time so we don't say tangible Roi because it's hard to measure Roi in kind of a fiveyear time frame for most technology but we do say that the thing that they expect to get from that technology they should get in that mid two to five year time frame that doesn't mean that more advanced clients can't get harder more tangible benefits in the short term it just becomes higher risk and harder and then the long-term benefits Horizon we assign a five plus year benefits time frame to and so therefore we really say that those Technologies are not going to deliver tangible value for most clients for five or more years and we use the word tangible value to really express the idea that it will be very hard for clients most average clients to measure and feel good that they've gotten the benefits any benefits from these Technologies and again that's the average client so of course there are clients who are accelerating through all of these Technologies but they're taking a lot of risk they're getting involved in developing the Technologies and they're seeing the benefits but it is a higher risk proposition so that's what we do with Benefits Horizon we've done it I think this is our third or fourth year in a row and it really does help clients answer the question they need to answer which is when is this going to benefit me I will say that even though we assign benefits Horizons to these Technologies based on the average Forester client I encourage every client I talk to to assign their own benefit Horizon based on their attitude towards and their ability to take good technology risk and that you know is causing a lot of clients to really think about what they need to do and take some additional risk on these Technologies where they wouldn't have in the past got it let's talk a little bit more about gen so when we recorded this episode last year geni was exploding and we've learned a bit more about that you know Genai in the past year so what geni learnings played into the top 10 list this year so that's a good question Stephen certainly we've learned a lot about generative AI as has everybody in the last year I think the main thing that we besides learning that it's harder than we might initially believed is that we needed to separate out generative AI for language from generative AI for visual content because the tools being built around these capabilities are moving in different speeds and different directions so while yes we we'll have multimodal models and we think that's the direction of the future so some of that visual content processing is brought into generative AI for language most of the multimodal models are still primarily language based B and we see a separate set of models based on different AI methodologies for Pure visual content creation and generation and there's different use cases for these so and different business users who are going to be interested in these so we've split those out this year and we think that really makes a lot of sense so that was one of the learnings this year I think the second learning for generative AI this year is just the simple speed at which it's iterating and that's important because it really requires our clients and us to kind of skate ahead of the puck as it were and kind of estimate what technologies you should bet on now in this space and where you need to invest in the future which is something you know we all struggle with our clients certainly struggle with and that is how to invest for the future when you need the benefits now and that's only been exacerbated by the speed that gen is moving because a lot of the things you have to do now aren't going to benefit you for a while but you have to kind of look forward and say am I going to need this capability to for instance generate different kinds of training data from a Knowledge Graph so that I can train new emerging models in the generative a or language or visual content space right since it's moving so quickly it's almost imperative that you begin to invest for the future which is coming much faster than we expect so that would be a second learning Brian is one of those type of gen moving faster than the other is language moving faster than visual content or vice versa no I don't think we can say that there's an awful lot going on in both spaces so of course there was a big announcement this last year with the Sora model from open AI being able to rapidly generate a substantial amount of very realistic content it's another one of those stories of nobody expects how quickly this is moving generative AI for language I would say there's more use cases around it specifically when you take a look at how generative AI for language is accelerating things like touring Bots for software coding or AI agents which we'll talk about in a moment so I think there's a bigger set of use cases around generative AI for language but generative AI for visual content and its ability to accelerate primarily marketing and all the content and all the customization of that content in the marketing space is that's moving very quickly got it so both of those are in the short-term Horizon list let's talk about some of the other emerging texts that are in that short-term benefits Horizon absolutely one of the things that as I talked about we generated some surprises this year and one of the surprises we generated this year was iot security which the need and technology for securing iot devices has been around for about as long as there's been iot which is you know 25 30 years now so it's hard to say that word is new but one of the things we looked at when we did our analysis year is the simple trending up of of interest and all the changes in the technology space in iot security so when we really looked at the data and looked at the importance of it we said we absolutely have to put this on the top 10 this year and so we did and spent a lot of time with our analysts covering iot security to think about actually what's different and so what's actually happening in the iot security space and I'm going to paraphrase this is there's been a lot of iot security type Solutions on the market for a while but what we see is other vendors let's say in the identity and access management space recognizing that they need to do identity and access management out closer to Edge environments where there are iot devices and so they're building iot security Technologies and capabilities into those platforms so and primarily these are focused on getting a handle and inventorying spefic specifically what devices are out there and what firmware or software is installed on those devices but still that's a huge step towards being able to secure them so we have this conflict going on between dedicated iot Security Solutions and iot security capabilities and other types of security tools and it's creating a big decision for clients in terms of how you get your iot security capability and all the while the expansion of the attack surface that is being driv driven by AI where new kind of automation Bots driven by AI can or hackers can use these Bots to scale the attacks they're making on devices is making it more critical than ever the clients make the decision around an iot security set of capabilities and the challenge they're having is where do you Source them from and so it's a really interesting environment right now and really important which is why we put it on our top 10 this year Brian before we move on to the next emerging Tech you hit on something that's interesting with all the issues around you know geni and security what do you think about the Trust In AI is it high is it low what's going on there what's the work that needs to be done there I think it's generally still fairly low when we talk to folks about AI specifically in generative AI the thing that everyone comes back to is it's I think you hear the words non-deterministic but what we really mean is that the results aren't repeatable that you never quite know what's going to come out of a generative AI model and that is quite rightfully in the near- term got a lot of folks who care about or are Professionals in the governance security and risk space concerned as well as as business Executives who would sponsor a lot of these initiatives so I think there's a number of kind of competing things going on today a you have to understand that there's a limited set of use cases in which this non-deterministic and sometimes erroneous set of outputs actually can produce applications that deliver business value and there's a lot of applications that use cases where we simply can't have that kind of error the other side of that coin is humans are very error prone as well and yet we have human employees of course that we just have to manage the fact that mistakes get made and so there's a fair bit of how do we deal with these AI software systems that make mistakes like humans do and how do we put governance processes and manage them in way similar that we manage human workers so that we can get advantages from them that software gives us the ability you know repeatability and scalability and and those things so there's just a fair bit of that going on so that organizations can learn where it's appropriate to trust the current generation of these tools and then there's an awful lot of work going on in the technology to create much much more trustworthy capabilities in the future really interesting stuff that we probably don't have time to get into but way down in the research of how do we get inside as it were the head of these neural network architectures understand the thoughts and what's happening in the layers of neurons and actually begin to develop methods to steer them and understand where they go off the rails and guide them back onto the rails so I really think we're going to see over the next few years three four five years massive leaps and capabilities around these models ability to be trustworthy while all at the same time learning where we can trust these models as businesses and those two will come together and I think we're going to see some big improvements over time but right now the trust is is fairly low got it before we move on to the medium- terms benefits Horizon let's talk about the last item in the short-terms benefits Horizon which is Turing Bots they moved from medium term to short term this year why well I think that's the big thing about generative Ai and everything is accelerating like touring Bots is even the best most experienced people in AI in the world are still being stunned by the speed at which these tools are advancing in their capabilities but when you really look back at it it's not all that surprising if you can track what's been happening with language models all the way back to 2013 and the kind of the word Tove initial release of of capability by a deep mind which is very stunning so tearing Bots is one of those things where we came up with this term I think back in 20120 to represent the idea that machine learning algorithms have given enough training data through deep learning neural networks can produce some fairly stunning abilities to reproduce things like code and I remember one of our analysts Mike gter myself and Diego Loi the three of us were talking about this actually Mike Gary's idea that we could have a deep learning algorithm given enough training data actually produce fairly usable code and that idea was a little controversial back in 2020 but we went with it because we did see the potential then of course generative AI took off and Turing Bots have accelerated so what we finally realized this year is that these Turing Bots would deliver benefits even faster than we thought last year and we're actually seeing that play out so that's why they're in our short term this year it's really tough because touring Bots can do very Advanced things right now but there's still a relatively small percentage of code in production but there's an awful lot of Open Source GitHub code being produced by touring Bots and the ability of touring Bots to do more things around the software development life cycle is only advancing I was talking to a vendor just yesterday one of the things we've thought about touring Bots is that not only can they help generate code and generate testing for code but if we could combine these two things with some ability to analyze existing code you could use cheering Bots to eliminate a lot of the technical debt you have by looking at the patterns of the code you write and understanding that in certain cases you don't have the right patterns of code deployed and that's causing performance issues and we could actually have a way for Turing Bots to go in and help you identify and even fix those so the ability for Turing Bots to eliminate technical debt optimize the code you have kind of really serving that architecture and Tech debt reduction function is something we thought about a few years ago and we're actually seeing it play out today so really excited about turn Bots and that plays into the point you made earlier which was that this isn't necessarily going to replace jobs it's going to help you do your job absolutely I mean realistically and of course this is a point of debate my opinion is that we're going to have far fewer actual true technical developers but those developers will be much more much much orders of magnitude more productive because they'll be able to spend their time actually orchestrating different kinds of touring Bots to help them build and deploy code at a much greater Pace I guess the debate is are we just going to have more code because of that and therefore we'll have the same number of humans doing it producing orders of magnitude more code are we going to have less humans producing that code so it's kind of like that's a debate up in the air right how much code are we gonna have and how many humans are going to be producing it my personal opinion is that we're going to have fewer higher Val more creative developers contributing much more value into the software generation process through touring Bots and a lot of the underlying plumbing and the stuff that we don't like doing will all be done for us that's my particular opinion on this but you know it's open for debate right now it's one of those things where we're kind of scratching our heads and let's see where this thing goes because it is very exciting so let's move on to the medium-term benefits Horizon Quantum security new this year let's talk about about why it made the list there's been a lot of talk as long as I've been covering Quantum Computing which is about 10 years of the quantum's ability to attack today's best pki type asymmetric key technically encryption and quite honestly that ability of quantum a quantum computer to break asymmetric key encryption is still something that is perhaps a decade away we did some detailed calculations and some report reports that we published that put it 50/50 chance around 10 years away so that seems like a long time unless you think about the amount of time that you want to have your information protected in case it's harvested today and decrypted later and the amount of time it takes to put in a set of security capabilities that Implement algorithms that can protect your data against the quantum attack that may happen at some point so what this really means is that if you want to be absolutely safe to ensure that the things that you're the the information that you're transmitting today over the Internet even though it's encrypted remains safe we think the time to begin investing in the infrastructure to protect that information is now and that's also timed with release of some of the postquantum um cryptography standards from organizations such as nist so all this is kind of coming together and we think that time that you have to really start investing in that infrastructure is actually now to ensure that you're completely safe because there's any number of breakthroughs in the quantum Computing space that could accelerate the timeline when our information actually becomes vulnerable to attack but the really cool thing about how we wrote this emerging technology this year is the side benefit you get from investing in what we call crypto agility is that you're going to get a much more robust and uh security architecture that enables you to understand uh what crypto you have out there and rapidly swap out algorithms for new better algorithms that protect you against things like future uh decryption by quantum computers so there's a side benefit to investing in preparing for Quantum Computing today you're going to get a much better security architecture and posture through that that's going to benefit you in many ways Beyond just Quantum being ready for Quantum attack so we really think that you know paying attention to that today is going to benefit you in so many ways in the future and that's why it made our top 10 list so Brian how about AI agents this year tell us more about what the research showed yeah agents are probably my favorite emerging technology this year and we've been writing about agents for at least three or four years now we used to call them intelligent agents which is like the technical academic term for AI system that could actually you go do stuff last year we called them autonomous workplace assistants based on a lot of the work that like Craig deair has done in implementation of AI capabilities and tools like RPA or digital process automation but was all focused on back office work automation which is where we were focused last year well one of the things we began to see this year is that through specifically through acceleration in the generative AI space an awful lot of you know front office or mid office applications customer facing applications of this including products built around it like you I think some people call them digital doubles or customer service agents so this year we decided to really expand what the emerging technology was and how it's playing out by rebranding it as AI agents and we're really seeing kind of like two things happen there number one an awful lot of products are being built bringing in new generative AI capabilities where they actually it isn't just about having a conversation like just a chat gbt interface or a gen interface but it's more about asking beneath that interface for a set of software components that may be AI models to go get some stuff done and then report back so that ability to provide a product or a service to a cust customer or an employee that goes beyond just answering questions but actually goes out and gets tasks done as an agent would on behalf of you know in kind of the principal agent kind of relationship so there's kind of a product facing capability that's happening so we're seeing a lot of kind of agent-based software tools pop up but the really interesting pieces below that software we're seeing a new agent-based architecture happen where we think that a lot of the things that we used to do in code different kinds of code and with lots of different business rules we're seeing kind of behind the scenes uh in the software a new kind of architecture emerge using large language models and other kinds of AI models as the elements of the code and how these models are working is they're taking tasks breaking those tasks down into plans and then coordinating the execution of those Plans by calling different tools code generation tools database lookups maybe some large language model type prompt manipulation there's all kinds of things going on and so this agent-based architecture as a new software Paradigm is underpinings that we see coming to Market so it's a kind of a really interesting time there's a lot of attention being paid to this in the market today at almost every Forum we go to about what this agent-based architecture is going to do for both software products and software architectures it's really quite fascinating it's probably one of my favorite emerging techs this year let's talk about autonomous Mobility is that new on the list this year it is we've been tracking it was in our top 20 list last year and this year our analysis showed that it it has come to a point where all of our clients need to pay attention to it no matter what industry they're in because to some extent it's going to impact them and so we went off and had a team that that are looking at this and we're going to write a lot more research around it this year but we call it autonomous Mobility technology because it's not just the self-driving cars which is what everybody thinks about but it's all the technology around self-driving or autonomous Mobility which includes things like drones and how these are revolutionizing certain sectors and industries uh we think the two places where we're going to see the most impact of this technology are in urban and public domain transportation and in kind of commercial transportation and Logistics the idea of self-driving cars moving around the populace where we all don't have to drive anymore we still think that's a fairly long way away simply because of the difficulties that AI systems have navigating the physical real world there's a lot of reasons why that's still really hard but in more constrained environments in manufacturing supply chain and Logistics certain use of drones certain use of self-driving cars for transportation or in urban scenarios where we have more control and it's a more it's a tighter environment we think that autonomous Mobility Technologies are going to have quite a big impact even in the short term in the next two to five years so we think it's time for clients to really dive into that technology and pay attention to it which is why we're going to spend some time on it for sure we had two technologies on the long-term Horizon XR and and zero trust Edge are they either of those any closer to reality this year than last year yeah those two were on our longterm list last year and we maintained them on our long-term list this year instead of advancing either one of them and you know frankly extended reality still because of the advancements that the foreign factors need to make and frankly consumer Readiness to adopt these Technologies over frankly what works pretty well for them today which are screens devices mice and now kind of natural language interaction through generative AI all that works pretty well for consumers today and so consumers aren't ready to make the jump to that next big interface of extended reality but we think someday that will happen and that's going to be kind of the best next big interface change but there's a consumer adoption and you know some progress that needs to be made and the convenience of these form factors it's going to take a while that said we still see a lot of Enterprise opportunities in some very narrow use cases training and onboarding being probably the number one extended fuel support being another one where Enterprises are beginning to adopt and make progress using new extended reality which includes virtual reality augmented reality and mixed reality for factors so we maintain our position is we think the big impact of those that kind of next big interface that changes the world is still a good distance away zero trust Edge again very interesting technology it's kind of down in the detail and the weeds we think whenever I talk to clients about it that they need to pay some attention to it and we still maintain that it is a number of years out for a lot of reasons number one the vendors have a lot still to figure out in terms of making tools that combine networking and security and Cloud enabling that all the vendors have some work to do to make that simple for clients and clients have a lot of work to do because zero trust Edge is essentially a cloud native approach to networking and security clients with a lot of Legacy devices that don't use modern communication protocols that can be easily made Cloud native or clients who aren't necessarily deeply into the ability to manage capabilities and software packages and security through Cloud native techniques and Technologies they're going to have a hard time with zero trust edge of getting the benefits so there's a Readiness in the vendor perspective and a client Readiness and a legacy device thing all kind of mixing around in zero trust Edge and we think it's still going to take a while before that really delivers for most clients the benefits that we know it can so we maintain those two as still in our long-term benefits arising this show Brian if there's one final thing we want our listeners to keep in mind as we wrap up our conversation what would that be I get to ask this question every year and I say the same thing and that's the pace of Technology advancement is accelerating and already seeing there things like touring Bots each year we've Advanced it to you know from long-term to midterm to shortterm and each of these Technologies are interplaying with each other a lot of these Technologies are accelerating the other Technologies so I think it becomes even more important each year that we do this that you have to begin to invest in technology in the foundations ahead of when you think you need them because the need is arriving in many cases faster than you think so that is always against that thing that we talked about earlier which is trust you know how do you trust in these Technologies and a lot of that trust is not only culture and process and just being able to feel good and safe about the way that you're applying these Technologies but it's also knowing you've done what you need in terms of the foundation the data management Foundation the security and risk Foundation all the things you'll need to be able to trust these so that's what I I say every year it's like clients have to recognize how these Technologies new generations are stacking on previous generations creating an acceleration of benefits which means you're going to need those foundations faster than you think to be ready for the future great well thanks for the conversation Brian really appreciate you taking the time to talk with us today you bet I enjoyed it every year if you want to learn more about our top 10 emerging Technologies for 2024 please visit for.com merging Tech that's F RR emerging Tech thanks for listening [Music]
2024-07-04