《德勤（Deloitte）：2025年技术趋势报告（英文版）（72页）.pdf》由会员分享，可在线阅读，更多相关《德勤（Deloitte）：2025年技术趋势报告（英文版）（72页）.pdf（72页珍藏版）》请在三个皮匠报告上搜索。

1、iTech Trends 2025Tech Trends 2025In Deloittes 16th annual Tech Trends report,AI is the common thread of nearly every trend.Moving forward,it will be part of the substructure of everything we do.02.Executive summary05.AI everywhere:Like magic,but with algorithms09.Spatial computing takes center stage

2、17.Whats next for AI?27.Hardware is eating the world37.IT,amplified:AI elevates the reach(and remit)of the tech function45.The new math:Solving cryptography in an age of quantum53.The intelligent core:AI changes everything for core modernization60.Breadth is the new depth:The power of intentional in

3、tersectionsINTRODUCTIONINTERACTIONINFORMATIONCOMPUTATIONBUSINESS OF TECHNOLOGYCYBER AND TRUSTCORE MODERNIZATIONCONCLUSIONTable of contents2INFORMATIONCOMPUTATIONBUSINESS OFTECHNOLOGYCYBER ANDTRUSTCOREMODERNIZATIONINTERACTIONWhats next for AI?Spatial computing takes center stageIT,amplified:AI elevat

4、es the reach(and remit)of the tech functionThe new math:Solving cryptography in an age of quantum The intelligent core:AI changes everything for core modernizationHardware is eating the worldFigure 1Six macro forces of information technologyTech Trends,Deloittes flagship technology report,explores t

5、he emergence of trends in three elevating forces(interaction,information,and computation)and three grounding forces(business of technology,cyber and trust,and core modernization)all part of our macro technology forces framework(figure 1).Tech Trends 2025,our 16th trip around the sun,previews a futur

6、e in which artificial intelligence will be as foundational as electricity to daily business and personal lives.As our team in Deloittes Office of the CTO put finishing touches on Tech Trends 2025,we realized that AI is a common thread in nearly every trend.We expect that going forward,AI will be so

7、ubiquitous that it will be a part of the unseen substructure of everything we do,and we eventually wont even know its there.Executive summary3Executive summaryIntroductionAI everywhere:Like magic,but with algorithmsGenerative AI continues to be the buzzword of the year,but Tech Trends 2025and in fac

8、t,the future of technologyis about much more than AI.This years report reveals the extent to which AI is being woven into the fabric of our lives.Well eventually take it for granted and think of it in the same way that we think of HTTP or electricity:Well just expect it to work.AI will perform quiet

9、ly in the background,optimizing traffic in our cities,personalizing our health care,or creating adaptative and accessible learning paths in education.We wont proactively use it;well simply experience a world in which it makes everything work smarter,faster,and more intuitivelylike magic,but grounded

10、 in algo-rithms.The six chapters of Tech Trends 2025 reflect this emerging reality.Interaction Spatial computing takes center stage Spatial computing continues to spark enterprise interest because of its ability to break down information silos and create more natural ways for workers and customers t

11、o interact with information.Were already seeing enter-prises find success with use cases like advanced simula-tions that allow organizations to test different scenarios to see how various conditions will impact their oper-ations.With a stronger focus on effectively managing spatial data,organization

12、s can drive more cutting-edge applications.In the coming years,advancements in AI could lead to seamless spatial computing experiences and improved interoperability,ultimately enabling AI agents to anticipate and proactively meet users needs.InformationWhats next for AI?To take advantage of the burg

13、eoning excitement around generative AI,many organizations have already adopted large language models(LLMs),the best option for many use cases.But some are already looking ahead.Despite their general applicability,LLMs may not be the most efficient choice for all organizational needs.Enterprises are

14、now considering small language models and open-source options for the ability to train LLMs on smaller,more accurate data sets.Together with multimodal models and AI-based simulations,these new types of AI are building a future where enterprises can find the right type of AI for each task.That inclu

15、des AI that not only answers questions but also completes tasks.In the coming years,a focus on execution may usher in a new era of agentic AI,arming consumers and organizations with co-pilots capable of transforming how we work and live.Computation Hardware is eating the worldAfter years of software

16、 dominance,hardware is reclaim-ing the spotlight.As AI demands specialized computing resources,companies are turning to advanced chips to power AI workloads.In addition,personal comput-ers embedded with AI chips are poised to supercharge knowledge workers by providing access to offline AI models whi

17、le“future-proofing”technology infrastruc-ture,reducing cloud computing costs,and enhancing data privacy.Although AIs increased energy demands pose sustainability challenges,advancements in energy sources and efficiency are making AI hardware more accessible.Looking forward,AIs continued integration

18、into devices could revolutionize the Internet of Things and robotics,transforming industries like health care through smarter,more autonomous devices.Business of technologyIT,amplified:AI elevates the reach(and remit)of tech talentAfter years of progressing toward lean IT and every-thing-as-a-servic

19、e offerings,AI is sparking a shift away from virtualization and austere budgets.Long viewed as the lighthouse of digital transformation through-out the enterprise,the IT function is now taking on AI transformation.Because of generative AIs applicability to writing code,testing software,and augmentin

20、g tech talent in general,forward-thinking technology leaders are using the current moment as a once-in-a-blue-moon 4opportunity to transform IT across five pillars:infra-structure,engineering,finance operations,talent,and innovation.As both traditional and generative AI capa-bilities grow,every phas

21、e of tech delivery could see a shift from human in charge to human in the loop.Such a move could eventually return IT to a new form of lean IT,leveraging citizen developers and AI-driven automation.Cyber and trustThe new math:Solving cryptography in an age of quantum In their response to Y2K,organiz

22、ations saw a loom-ing risk and addressed it promptly.Today,IT faces a new challenge,and it will have to respond in a similarly proactive manner.Experts predict that quantum comput-ers,which could mature within five to 20 years,will have significant implications for cybersecurity because of their abi

23、lity to break existing encryption methods and digital signatures.This poses a risk to the integrity and authenticity of data and communications.Despite the uncertainty of the quantum computer timeline,inaction on post-quantum encryption is not an option.Emerging encryption standards offer a path to

24、mitigation.Updating encryption practices is fairly straightforwardbut its a lengthy process,so organizations should act now to stay ahead of potential threats.And while theyre at it,they can consider tackling broader issues surrounding cyber hygiene and cryptographic agility.Core modernizationThe in

25、telligent core:AI changes everything for core modernization Core systems providers have invested heavily in AI,rebuilding their offerings and capabilities around an AI-fueled or AI-first model.The integration of AI into core enterprise systems represents a significant shift in how organizations oper

26、ate and leverage technology for competitive advantage.This transformation is about automating routine tasks and fundamentally rethinking and redesigning processes to be more intelligent,efficient,and predictive.It requires careful planning due to inte-gration complexity,strategic investment in techn

27、ology and skills,and a robust governance framework to ensure smooth operations.But beware of the automation para-dox:The more complexity is added to a system,the more vital human workers become.Adding AI to core systems may simplify the user experience,but it will make them more complex at an archit

28、ectural level.Deep technical skills are still critical for managing AI in core systems.ConclusionBreadth is the new depth:The power of intentional intersectionsOrganizations have long relied on innovation-driven new revenue streams,synergies created through mergers and acquisitions,and strategic par

29、tnerships.But increasingly,segmentation and specialization have given way to inten-tional intersections of technologies and industries.For example,when two technologies intersect,they are often complementary,but they can also augment each other so that both technologies ultimately accelerate their g

30、rowth potential.Similarly,new opportunities can emerge when companies aim to extend their market share by purpose-fully partnering across seemingly disparate industries.5AI everywhere:Like magic,but with algorithmsTwo years after generative artificial intelligence staked its claim as the free space

31、on everyones buzzword-bingo cards,youd be forgiven for imagining that the future of technology is simply more AI.Thats only part of the story,though.We propose that the future of technology isnt so much about more AI as it is about ubiquitous AI.We expect that,going forward,AI will become so fundame

32、ntally woven into the fabric of our lives that its everywhere,and so foundational that we stop noticing it.Take electricity,for example.When was the last time you actually thought about electrons?We no longer marvel that the lights turn onwe simply expect them to work.The same goes for HTTP,the unse

33、en thread that holds the internet together.We use it every day,but Id bet most of us havent thought about(let alone uttered)the word“hypertext”in quite some time.AI will eventually follow a similar path,becoming so ubiquitous that it will be a part of the unseen substruc-ture of everything we do,and

34、 we eventually wont even know its there.It will quietly hum along in the back-ground,optimizing traffic in our cities,personalizing our health care,and creating adaptative and accessible learn-ing paths in education.We wont“use”AI.Well just experience a world where things work smarter,faster,and mor

35、e intuitivelylike magic,but grounded in algo-rithms.We expect that it will provide a foundation for business and personal growth while also adapting and sustaining itself over time.Nowhere is this AI-infused future more evident than in this years Tech Trends report,which each year explores emerging

36、trends across the six macro forces of informa-tion technology(figure 1 in the executive summary).Half of the trends that weve chronicled are elevating forcesinteraction,information,and computationthat underpin innovation and growth.The other halfthe grounding forces of the business of technology,cyb

37、er and trust,and core modernizationhelp enterprises seamlessly operate while they grow.As our team put the finishing touches on this years report,we realized that this sublimation and diffu-sion of AI is already afoot.Not the“only trend”nor“every trend,”AI is the scaffolding and common thread buttre

38、ssing nearly every trend.(For those keeping a close eye at home,“The new math:Solving cryptography in an age of quantum”about the cybersecurity implica-tions of another game-changing technology,quantum computingis the only one in which AI does not have a foundational role.Yet behind the scenes,AI ad

39、vance-ments are accelerating advances in quantum.)Spatial computing takes center stage:Future AI advancements will enhance spatial-computing simu-lations,eventually leading to seamless spatial-com-puting experiences integrated with AI agents.Whats next for AI?:As AI evolves,the enterprise focus on l

40、arge language models is giving way to small language models,multimodal models,AI-based simulations,and agents that can execute discrete tasks.AI everywhere:Like magic,but with algorithmsTech Trends 2025 reveals how much artificial intelligence is being woven into the fabric of our livesmaking everyt

41、hing work smarter,faster,and more intuitively Kelly RaskovichINTRODUCTION6 Hardware is eating the world:After years of soft-ware dominance,hardware is reclaiming the spot-light,largely due to AIs impact on computing chips and its integration into end-user devices,the Internet of Things,and robotics.

42、IT,amplified:AI elevates the reach(and remit)of tech talent:AIs applicability to writing code,testing software,and augmenting tech talent is transform-ing IT and sparking a shift away from virtualization and austere budgets.The intelligent core:AI changes everything for core modernization:Core syste

43、ms providers have invested heavily in AI,which may simplify the user experience and data-sharing across applications but will make these systems more complex at an architectural level.Because we expect AI to become part of tomorrows foundational corelike electricity,HTTP,and so many other technologi

44、esits exciting to think about how AI might evolve in the next few years as it marches toward ubiquity,and how we as humans may benefit.We here at Tech Trends will be chronicling every step of the journey.Until next time,Kelly Raskovich Office of the CTOExecutive editor,Tech Trends7AI everywhere:Like

45、 magic,but with algorithmsTrending the trends2017201920202025201620182021202220232024Mixed realityBeyond marketingIntelligent interfacesHuman experienceplatformsAR and VRgo to workInternet of ThingsDigital realityBespoke for billionsRebooting the digital workplaceThrough the glassInterfaces in new p

46、lacesSpatial computing takes center stageINTERACTIONMachine intelligenceDark analyticsAI-fueledorganizationsDigital twinsIndustrializedanalyticsEnterprise datasovereigntyML Ops:Industrialized AIMachine datarevolutionData sharing made easyOpening up to AIGenie out of the botleWhats next for AI?INFORM

47、ATIONCloud goes verticalTrust economyEverything as-a-serviceNoOps in a serverless worldDemocratized trustBlockchain toblockchainsAPIimperativeBlockchain:Ready forbusinessAbove the cloudsSmarter,not harderHardware is eating the worldCOMPUTATIONInevitable architectureITunboundedConnectivity of tomorro

48、wArchitecture awakensFinance and the future of ITAutonomic platformsRight speed ITReengineeringtechnologyNo-collar workforceSupply unchainedStrategy,engineeredThe tech stack goes physicalDEI tech:Tools forequityFlexibility,the best abilityFrom DevOps to DevExIT,amplifiedBUSINESS OF TECHNOLOGYDevSecO

49、ps and the cyber imperativeEthical technology and trustZero trustCyber AIIn us we trustDefending realityThe new mathCYBER AND TRUSTReimagining core systemsThe new coreCore revivalIT,disrupt thyselfConnect and extendCore workoutThe intelligent coreCORE MODERNIZATIONNote:To learn more about past Tech

50、Trends,go to analysis.89Spatial computing takes center stageTodays ways of working demand deep expertise in narrow skill sets.Being informed about projects often requires significant specialized training and understanding of context,which can burden workers and keep information siloed.This has histo

51、rically been true especially for any workflow involving a physical component.Specialized tasks demanded narrow training in a variety of unique systems,which made it hard to work across disciplines.One example is computer-aided design(CAD)software.An experienced designer or engineer can view a CAD fi

52、le and glean much information about the project.But those outside of the design and engineering realmwhether theyre in marketing,finance,supply chain,project management,or any other role that needs to be up to speed on the details of the workwill likely struggle to understand the file,which keeps es

53、sential technical details buried.Spatial computing is one approach that can aid this type of collaboration.As discussed in Tech Trends 2024,spatial computing offers new ways to contextualize business data,engage customers and workers,and inter-act with digital systems.It more seamlessly blends the p

54、hysical and digital,creating an immersive technology ecosystem for humans to more naturally interact with the world.1 For example,a visual interaction layer that pulls together contextual data from business software can allow supply chain workers to identify parts that need to be ordered and enable

55、marketers to grasp a prod-ucts overall aesthetics to help them build campaigns.Employees across the organization can make meaning of and,in turn,make decisions with detailed information about a project in ways anyone can understand.If eye-catching virtual reality(VR)headsets are the first thing that

56、 come to mind when you think about spatial computing,youre not alone.But spatial computing is about more than providing a visual experience via a pair of goggles.It also involves blending standard busi-ness sensor data with the Internet of Things,drone,light detection and ranging(LIDAR),image,video,

57、and other three-dimensional data types to create digital representa-tions of business operations that mirror the real world.These models can be rendered across a range of inter-action media,whether a traditional two-dimensional screen,lightweight augmented reality glasses,or full-on immersive VR env

58、ironments.Spatial computing senses real-world,physical compo-nents;uses bridging technology to connect physical and digital inputs;and overlays digital outputs onto a blended interface(figure 1).2Spatial computings current applications are as diverse as they are transformative.Real-time simulations

59、have emerged as the technologys primary use case.Looking ahead,advancements will continue to drive new and exciting use cases,reshaping industries such as health care,manufacturing,logistics,and entertainmentwhich is why the market is projected to grow at a rate of 18.2%between 2022 and 2033.3 The j

60、ourney from the present to the future of human-computer interaction promises to fundamentally alter how we perceive and interact with the digital and physical worlds.Spatial computing takes center stageWhat is the future of spatial computing?With real-time simulations as just the start,new,exciting

61、use cases can reshape industries ranging from health care to entertainment.Kelly Raskovich,Bill Briggs,Mike Bechtel,and Ed BurnsINTERACTION10Now:Filled to the rim with simsAt its heart,spatial computing brings the digital world closer to lived reality.Many business processes have a physical componen

62、t,particularly in asset-heavy indus-tries,but,too often,information about those processes is abstracted,and the essence(and insight)is lost.Businesses can learn much about their operations from well-organized,structured business data,but adding physical data can help them understand those operations

63、 more deeply.Thats where spatial computing comes in.“This idea of being served the right information at the right time with the right view is the promise of spatial computing,”says David Randle,global head of go-to-market for spatial computing at Amazon Web Services(AWS).“We believe spatial computin

64、g enables more natural understanding and awareness of physical and virtual worlds.”4 One of the primary applications unlocked by spatial computing is advanced simulations.Think digital twins,but rather than virtual representations that monitor physical assets,these simulations allow organizations to

65、 test different scenarios to see how various conditions will impact their operations.Imagine a manufacturing company where designers,engineers,and supply chain teams can seamlessly work from a single 3D model to craft,build,and procure all the parts they need;doctors who can view true-to-life simula

66、tions of their patients bodies through augmented reality displays;or an oil and gas company that can layer detailed engineering models on top of 2D maps.The possibilities are as vast as our physical world is varied.The Portuguese soccer club Benficas sports data science team uses cameras and compute

67、r vision to track players Figure 1The possibilities of spatial operationsDigitalAugmented reality objectsInteractive digital objectsHolographic projectionsAudio outputsAvatarsGenerative AIPhysicalNext-gen displays Wearables(for example,headset,smart eyewear,and pins)Internet of Things devices(for ex

68、ample,biometric devices)Sensory tech(for example,haptic suits)Spatial audio devices Cameras Next-gen bateriesBridgingSensors(for example,LIDAR)and sensor fusionComputer visionGPS/spatial mapping sofware3D design and rendering toolsComprehensive next-gen network infrastructureData lakesSource:Abhijit

69、h Ravinutala et al.,“Dichotomies spatial computing:Navigating towards a beter future,”Deloite,April 22,2024.11Spatial computing takes center stagethroughout matches and develop full-scale 3D models of every move its players make.The cameras collect 2,000 data points from each player,and AI helps ide

70、ntify specific players,the direction they were facing,and criti-cal factors that fed into their decision-making.The data essentially creates a digital twin of each player,allowing the team to run simulations of how plays would have worked if a player was in a different position.Xs and Os on a chalkb

71、oard are now three-dimensional models that coaches can experiment with.5“Theres been a huge evolution in AI pushing these models forward,and now we can use them in deci-sion-making,”says Joao Copeto,chief information and technology officer at Sport Lisboa e Benfica.6 This isnt only about wins and lo

72、ssesits also about dollars and cents.Benfica has turned player development into a profitable business by leveraging data and AI.Over the past 10 years,the team has generated some of the highest player-transfer deals in Europe.Similar approaches could also pay dividends in warehouse oper-ations,suppl

73、y chain and logistics,or any other resource planning process.Advanced simulations are also showing up in medical settings.For instance,virtual patient scenarios can be simulated as a training supplement for nurses or doctors in a more dynamic,self-paced environment than text-books would allow.This m

74、ay come with several chal-lenges,such as patient data concerns,integration of AI into existing learning materials,and the question of realism.But AI-based simulations are poised to impact the way we learn.7 Simulations are also starting to impact health care delivery.Fraser Health Authority in Canad

75、a has been a pioneer in leveraging simulation models to improve care.8 By creating a first-of-its-kind system-wide digital twin,the public health authority in British Columbia generated powerful visualizations of patient movement through different care settings and simulations to determine the impac

76、t of deploying different care models on patient access.Although the work is ongoing,Fraser expects improvement in appropriate,need-based access to care through increased patient awareness of available services.New:Data is the differentiatorEnterprise IT teams will likely need to overcome signifi-can

77、t hurdles to develop altogether-new spatial comput-ing applications.They likely havent faced these hurdles when implementing more conventional software-based projects.While these projects have compelling busi-ness value,organizations will have to navigate some uncharted waters to achieve them.For on

78、e thing,data isnt always interoperable between systems,which limits the ability to blend data from different sources.Furthermore,the spaghetti diagrams mapping out the path that data travels in most organi-zations are circuitous at best,and building the data pipe-lines to get the correct spatial dat

79、a into visual systems is a thorny engineering challenge.Ensuring that data is of high quality and faithfully mirrors real-world condi-tions may be one of the most significant barriers to using spatial computing effectively.9 Randle of AWS says spatial data has not historically been well managed at m

80、ost organizations,even though it represents some of a businesss most valuable information.“This information,because its quite new and diverse,has few standards around it and much of it sits in silos,some of its in the cloud,most of its not,”says Randle.“This data landscape encompassing physical and

81、digital assets is extremely scattered and not well managed.Our customers first problem is managing their spatial data.”10Taking a more systematic approach to ingesting,orga-nizing,and storing this data,in turn,makes it more available to modern AI tools,and thats where the real learnings begin.Data p

82、ipelines deliver the fuel that drives business Weve often heard that data is the new oil,but for an American oil and gas company,the metaphor is becom-ing reality thanks to significant effort in replumbing some of its data pipelines.The energy company uses drones to conduct 3D scans of equipment in

83、the field and its facilities,and then applies 12computer vision to the data to ensure its assets operate within predefined tolerances.Its also creating high-fi-delity digital twins of assets based on data pulled from engineering,operational,and enterprise resource plan-ning systems.The critical piec

84、e in each example?Data integration.The energy giant built a spatial storage layer,using appli-cation program interfaces to connect to disparate data sources and file types,including machine,drone,busi-ness,and image and video data.11 Few organizations today have invested in this type of systematic a

85、pproach to ingesting and storing spatial data.Still,its a key factor driving spatial computing capabil-ities and an essential first step for delivering impactful use cases.Multimodal AI creates the context In the past,businesses couldnt merge spatial and busi-ness data into one visualization,but tha

86、t too is chang-ing.As discussed in“Whats next for AI?”multimodal AIAI tools that can process virtually any data type as a prompt and return outputs in multiple formatsis already adept at processing virtually any input,whether text,image,audio,spatial,or structured data types.12 This capability will

87、allow AI to serve as a bridge between different data sources,and interpret and add context between spatial and business data.AI can reach into disparate data systems and extract relevant insights.This isnt to say multimodal AI eliminates all barriers.Organizations still need to manage and govern the

88、ir data effectively.The old saying“garbage in,garbage out”has never been more prescient.Training AI tools on disorga-nized and unrepresentative data is a recipe for disaster,as AI has the power to scale errors far beyond what weve seen with other types of software.Enterprises should focus on impleme

89、nting open data standards and working with vendors to standardize data types.But once theyve addressed these concerns,IT teams can open new doors to exciting applications.“You can shape this technology in new and creative ways,”says Johan Eerenstein,executive vice president of workforce enablement a

90、t Paramount.13 Next:AI is the new UIMany of the aforementioned challenges in spatial computing are related to integration.Enterprises strug-gle to pull disparate data sources into a visualization platform and render that data in a way that provides value to the user in their day-to-day work.But soon

91、,AI stands to lower those hurdles.As mentioned above,multimodal AI can take a variety of inputs and make sense of them in one platform,but that could be only the beginning.As AI is integrated into more applications and interaction layers,it allows services to act in concert.As mentioned in“Whats nex

92、t for AI?”this is already giving way to agentic systems that are context-aware and capable of executing functions proactively based on user preferences.These autonomous agents could soon support the roles of supply chain manager,software developer,financial analyst,and more.What will separate tomorr

93、ows agents from todays bots will be their ability to plan ahead and anticipate what the user needs without even having to ask.Based on user preferences and historical actions,they will know how to serve the right content or take the right action at the right time.When AI agents and spatial computing

94、 converge,users wont have to think about whether their data comes from a spatial system,such as LIDAR or cameras(with the important caveat that AI systems are trained on high-quality,well-managed,interoperable data in the first place),or account for the capabilities of specific applications.With int

95、elligent agents,AI becomes the interface,and all thats necessary is to express a prefer-ence rather than explicitly program or prompt an appli-cation.Imagine a bot that automatically alerts financial analysts to changing market conditions,or one that crafts daily reports for the C-suite about change

96、s in the business environment or team morale.All the many devices we interact with today,be they phone,tablet,computer,or smart speaker,will feel downright cumbersome in a future where all we have to do is gesture toward a preference and let context-aware,AI-powered systems execute our command.Event

97、ually,once these systems have learned our preferences,we may not even need to gesture at all.13Spatial computing takes center stageThe full impact of agentic AI systems on spatial comput-ing may be many years out,but businesses can still work toward reaping the benefits of spatial comput-ing.Buildin

98、g the data pipelines may be one of the heaviest lifts,but once built,they open up myriad use cases.Autonomous asset inspection,smoother supply chains,true-to-life simulations,and immersive virtual environments are just a few ways leading enterprises are making their operations more spatially aware.A

99、s AI continues to intersect with spatial systems,well see the emergence of revolutionary new digital frontiers,the contours of which were only beginning to map out.141.Abhijith Ravinutala et al.,“Dichotomies Spatial Computing:Navigating Towards a Better Future,”Deloitte,April 22,2024.2.Ibid.3.Future

100、 Market Insights,Spatial Computing Market Outlook(2022 to 2032),October 2022.4.David Randle(global head of go-to-market,AWS),interview with the author,Sept.16,2024.5.Joao Copeto,chief information and technology officer,Sport Lisboa e Benfica,interview with the author,August 27,2024.6.Ibid.7.Isabelle

101、 Bousquette,“Companies finally find a use for virtual reality at work,”The Wall Street Journal,Sept.6,2024.8.Fraser Health,“Fraser Health Authority:System wide digital twin,”October 2023.9.Gokul Yenduri et al.,“Spatial computing:Concept,applications,challenges and future directions,”preprint,10.4855

102、0/arXiv.2402.07912(2024).10.Randle interview.11.Deloitte internal information.12.George Lawton,“Multimodal AI,”TechTarget,accessed Oct.29,2024.13.Johan Eerenstein(senior vice president of workforce enablement,Paramount),interview with the author,July 16,2024.Endnotes15Spatial computing takes center

103、stageIndustry leadershipFrances YuUnlimited Reality GM/Business lead|Principal|Deloitte Consulting LLP+1 312 486 2563|Frances Yu is a partner at Deloitte Consulting LLP,where she has served in a range of global practice leadership roles.She has helped Fortune 500 clients as well as Deloitte launch s

104、everal new ventures,evolved growth strategies,and transformed their demand value chain.Currently,she is the US and global business lead and general manager for Deloittes Unlimited Reality,a multinetwork inno-vation business for the industrial metaverse era,focusing on spatial computing,digital twin,

105、and multimodal AI and data.Nishanth Raj Unlimited Reality Spatial/Multimodal AI and data lead|Managing director|Deloitte Consulting LLP+1 832 970 7560|Nishanth Raj is a managing director and AI and data/Unlimited Reality leader at Deloitte Consulting,specializing in the Energy&Chemicals sector.With

106、over two decades of consulting experience,he helps clients leverage technology,AI,and data to drive business value,and transform them into insights-driven organizations.Stefan KircherUnlimited Reality CTO|Managing director|Deloitte Consulting LLP+1 404 631 2541|Stefan Kircher is a managing director

107、in the Products&Solutions practice of Deloitte Consulting LLP and CTO for Deloittes Unlimited Reality Business.He has over 25 years expertise in the industry,technology strategy,and solution-building across various industries,R&D,innovation,and partnerships with strategic tech partners like AWS.Robe

108、rt TrossUnlimited Reality GPS market offering leader|Principal|Deloitte Consulting LLP+1 703 251 1250|Robert Tross is a principal in Deloitte Consulting LLPs GPS Government Technology practice,leading the Unlimited Reality federal market offering.With over 25 years of experience,he specializes in om

109、ni-channel experiences across various platforms,including web,immersive/spatial,social media,mobile,wearables,and tablets,including others.AcknowledgmentsMuch gratitude goes to the many subject matter leaders across Deloitte that contributed to our research for the Interaction chapter:Lars Cromley,S

110、tefan Kircher,Kaitlyn Kuczer,Lena La,Tim Murphy,Ali Newman,Bob Tross,and Frances Yu.Continue the conversation1617Whats next for AI?Blink and youll miss it:The speed of artificial intelli-gences advancement is outpacing expectations.Last year,as organizations scrambled to understand how to adopt gene

111、rative AI,we cautioned Tech Trends 2024 readers to lead with need as they differentiate themselves from competitors and adopt a strategic approach to scaling their use of large language models(LLMs).Today,LLMs have taken root,with up to 70%of organizations,by some estimates,actively exploring or imp

112、lementing LLM use cases.But leading organizations are already considering AIs next chapter.Instead of relying on foundation models built by large players in AI,which may be more powerful and built on more data than needed,enterprises are now thinking about implementing multiple,smaller models that c

113、an be more efficient for business requirements.LLMs will continue to advance and be the best option for certain use cases,like general-purpose chatbots or simulations for scientific research,but the chatbot that peruses your financial data to think through missed reve-nue opportunities doesnt need t

114、o be the same model that replies to customer inquiries.Put simply,were likely to see a proliferation of different horses for different courses.A series of smaller models working in concert may end up serving different use cases than current LLM approaches.New open-source options and multimodal outpu

115、ts(as opposed to just text)are enabling organiza-tions to unlock entirely new offerings.In the years to come,the progress toward a growing number of smaller,more specialized models could once again move the goalposts of AI in the enterprise.Organizations may witness a fundamental shift in AI from au

116、gmenting knowledge to augmenting execution.Investments being made today in agentic AI,as this next era is termed,could upend the way we work and live by arming consumers and businesses with armies of sili-con-based assistants.Imagine AI agents that can carry out discrete tasks,like delivering a fina

117、ncial report in a board meeting or applying for a grant.“Theres an app for that”could well become“Theres an agent for that.”Now:Getting the fundamentals right LLMs are undoubtedly exciting but require a great deal of groundwork.Instead of building models themselves,many enterprises are partnering wi

118、th companies like Anthropic or OpenAI or accessing AI models through hyperscalers.4 According to Gartner,AI servers will account for close to 60%of hyperscalers total server spending.5 Some enterprises have found immediate busi-ness value in using LLMs,while others have remained wary about the accur

119、acy and applicability of LLMs trained on external data.6 On an enterprise time scale,AI advancements are still in a nascent phase(crawling or walking,as we noted last year).According to recent surveys by Deloitte and Fivetran and Vanson Bourne,in most organizations,fewer than a third of generative A

120、I experiments have moved into production,often because organizations struggle to access or cleanse all the data needed to run AI programs.7 To achieve scale,organiza-tions will likely need to further think through data and technology,as well as strategy,process,and talent,as outlined in a recent Del

121、oitte AI Institute report.Whats next for AI?While large language models continue to advance,new models and agents are proving to be more effective at discrete tasks.AI needs different horses for different courses.Kelly Raskovich,Bill Briggs,Mike Bechtel,and Abhijith RavinutalaINFORMATION18According

122、to Deloittes 2024 State of Generative AI in the Enterprise Q3 report,75%of surveyed organiza-tions have increased their investments in data-life-cycle management due to generative AI.8 Data is foundational to LLMs,because bad inputs lead to worse outputs(in other words,garbage in,garbage squared).Th

123、ats why data-labeling costs can be a big driver of AI invest-ment.9 While some AI companies scrape the internet to build the largest models possible,savvy enterprises create the smartest models possible,which requires better domain-specific“education”for their LLMs.For instance,LIFT Impact Partners,

124、a Vancouver-based orga-nization that provides resources to nonprofits,is fine-tun-ing its AI-enabled virtual assistants on appropriate data to help new Canadian immigrants process paperwork.“When you train it on your organizations unique persona,data,and culture,it becomes significantly more relevan

125、t and effective,”says Bruce Dewar,president and CEO of LIFT Impact Partners.“It brings authenticity and becomes a true extension of your organization.”10 Data enablement issues are dynamic.Organizations surveyed by Deloitte said new issues could be exposed by the scale-up of AI pilots,unclear regula

126、tions around sensitive data,and questions around usage of external data(for example,licensed third-party data).Thats why 55%of organizations surveyed avoided certain AI use cases due to data-related issues,and an equal proportion are working to enhance their data security.11 Organizations could work

127、 around these issues by using out-of-the-box models offered by vendors,but differen-tiated AI impact will likely require differentiated enter-prise data.Thankfully,once the groundwork is laid,the benefits are clear:Two-thirds of organizations surveyed say theyre increasing investments in generative

128、AI because theyve seen strong value to date.12 Initial examples of real-world value are also appearing across industries,from insurance claims review to telecom troubleshoot-ing and consumer segmentation tools.13 LLMs are also making waves in more specialized use cases,such as space repairs,nuclear

129、modeling,and material design.14 As underlying data inputs improve and become more sustainable,LLMs and other advanced models(like simulations)may become easier to spin up and scale.But size isnt everything.Over time,as methods for AI training and implementation proliferate,organizations are likely t

130、o pilot smaller models.Many may have data that can be more valuable than previously imagined,and putting it into action through smaller,task-oriented models can reduce time,effort,and hassle.Were poised to move from large-scale AI projects to AI everywhere,as discussed in this years introduction.New

131、:Different horses for different coursesWhile LLMs have a vast array of use cases,the library is not infinite(yet).LLMs require massive resources,deal primarily with text,and are meant to augment human intelligence rather than take on and execute discrete tasks.As a result,says Vivek Mohindra,senior

132、vice pres-ident of corporate strategy at Dell Technologies,“there is no one-size-fits-all approach to AI.There are going to be models of all sizes and purpose-built optionsthats one of our key beliefs in AI strategy.”15 Over the next 18 to 24 months,key AI vendors and enterprise users are likely to

133、have a toolkit of models comprising increasingly sophisticated,robust LLMs along with other models more applicable to day-to-day use cases.Indeed,where LLMs are not the optimal choice,three pillars of AI are opening new avenues of value:small language models,multimodal models,and agentic AI(figure 1

134、).Small language models LLM providers are racing to make AI models as effi-cient as possible.Instead of enabling new use cases,these efforts aim to rightsize or optimize models for existing use cases.For instance,massive models are not neces-sary for mundane tasks like summarizing an inspection repo

135、rta smaller model trained on similar documents would suffice and be more cost-efficient.Small language models(SLMs)can be trained by enter-prises on smaller,highly curated data sets to solve more specific problems,rather than general queries.For exam-ple,a company could train an SLM on its inventory

136、 information,enabling employees to quickly retrieve insights instead of manually parsing large data sets,a process that can sometimes take weeks.Insights from such an SLM could then be coupled with a user interface application for easy access.19Whats next for AI?Naveen Rao,vice president of AI at Da

137、tabricks,believes more organizations will take this systems approach with AI:“A magic computer that understands everything is a sci-fi fantasy.Rather,in the same way we organize humans in the workplace,we should break apart our problems.Domain-specific and customized models can then address specific

138、 tasks,tools can run deterministic calculations,and databases can pull in relevant data.These AI systems deliver the solution better than any one component could do alone.”16An added benefit of smaller models is that they can be run on-device and trained by enterprises on smaller,highly curated data

139、 sets to solve more specific problems,rather than general queries,as discussed in“Hardware is eating the world.”Companies like Microsoft and Mistral are currently working to distill such SLMs,built on fewer parameters,from their larger AI offerings,and Meta offers multiple options across smaller mod

140、els and fron-tier models.17 Finally,much of the progress happening in SLMs is through open-source models offered by companies like Hugging Face or Arcee.AI.18 Such models are ripe for enterprise use since they can be customized for any number of needs,as long as IT teams have the internal AI talent

141、to fine-tune them.In fact,a recent Databricks report indicates that over 75%of organizations are choosing smaller open-source models and customizing them for specific use cases.19 Since open-source models are constantly improving thanks to the contributions of a diverse programming community,the siz

142、e and efficiency of these models are likely to improve at a rapid clip.Figure 1Different AI for different needsSmall language modelsMultimodalAgenticInputTextTextMore than textDataTo be determinedLessSignificantCustomizationVendors provide out-of-the-box capabilities,but works best when tailoredNeed

143、 to be customized and trained on data they would work withLess customization possible due to the volume of data requiredOutputMostSomeMoreFocusText,customizable,applied to diferent use cases(trainable)Can take concrete actionsCant train on smaller data sets;needs greater input and has widervariety o

144、f outputSource:Deloite research.20Multimodal models Humans interact through a variety of mediums:text,body language,voice,videos,among others.Machines are now hoping to catch up.20 Given that business needs are not contained to text,its no surprise that companies are looking forward to AI that can t

145、ake in and produce multiple mediums.In some ways,were already accus-tomed to multimodal AI,such as when we speak to digital assistants and receive text or images in return,or when we ride in cars that use a mix of computer vision and audio cues to provide driver assistance.21 Multimodal generative A

146、I,on the other hand,is in its early stages.The first major models,Googles Project Astra and OpenAIs GPT-4 Omni,were showcased in May 2024,and Amazon Web Services Titan offering has similar capabilities.22 Progress in multimodal gener-ative AI may be slow because it requires significantly higher amou

147、nts of data,resources,and hardware.23 In addition,the existing issues of hallucination and bias that plague text-based models may be exacerbated by multimodal generation.Still,the enterprise use cases are promising.The notion of“train once,run anywhere(or any way)”promises a model that could be trai

148、ned on text,but deliver answers in pictures,video,or sound,depending on the use case and the users preference,which improves digital inclu-sion.Companies like AMD aim to use the fledgling tech-nology to quickly translate marketing materials from English to other languages or to generate content.24 F

149、or supply chain optimization,multimodal generative AI can be trained on sensor data,maintenance logs,and warehouse images to recommend ideal stock quantities.25 This also leads to new opportunities with spatial comput-ing,which we write about in“Spatial computing takes center stage.”As the technolog

150、y progresses and model architecture becomes more efficient,we can expect to see even more use cases in the next 18 to 24 months.Agentic AI The third new pillar of AI may pave the way for changes to our ways of working over the next decade.Large(or small)action models go beyond the question-and-an-sw

151、er capabilities of LLMs and complete discrete tasks in the real world.Examples range from booking a flight based on your travel preferences to providing auto-mated customer support that can access databases and execute needed taskslikely without the need for highly specialized prompts.26 The prolife

152、ration of such action models,working as autonomous digital agents,heralds the beginnings of agentic AI,and enterprise software vendors like Salesforce and ServiceNow are already tout-ing these possibilities.27Chris Bedi,chief customer officer at ServiceNow,believes that domain-or industry-specific a

153、gentic AI can change the game for humans and machine interaction in enter-prises.28 For instance,in the companys Xanadu platform,one AI agent can scan incoming customer issues against a history of incidents to come up with a recommenda-tion for next steps.It then communicates to another autonomous a

154、gent thats able to execute on those recommendations,and a human in the loop reviews those agent-to-agent communications to approve the hypotheses.In the same vein,one agent might be adept at managing workloads in the cloud,while another provi-sions orders for customers.As Bedi says,“Agentic AI canno

155、t completely take the place of a human,but what it can do is work alongside your teams,handling repetitive tasks,seeking out information and resources,doing work in the background 24/7,365 days a year.”29 Finally,aside from the different categories of AI models noted above,advancements in AI design

156、and execution can also impact enterprise adoptionnamely,the advent of liquid neural networks.“Liquid”refers to the flexi-bility in this new form of training AI through a neural network,a machine learning algorithm that mimics the human brains structure.Similar to how quantum computers are freed from

157、 the binary nature of classical computing,liquid neural networks can do more with less:A couple dozen nodes in the network might suffice,versus 100,000 nodes in a more traditional network.The cutting-edge technology aims to run on less computing power,with more transparency,opening up possibili-ties

158、 for embedding AI into edge devices,robotics,and safety-critical systems.30 In other words,its not just the applications of AI but also its underlying mechanisms that are ripe for improvement and disruption in the coming years.21Whats next for AI?Next:Theres an agent for thatIn the next decade,AI co

159、uld be wholly focused on execu-tion instead of human augmentation.A future employee could make a plain-language request to an AI agent,for example,“close the books for Q2 and generate a report on EBITDA.”Like in an enterprise hierarchy,the primary agent would then delegate the needed tasks to agents

160、 with discrete roles that cascade across differ-ent productivity suites to take action.As with humans,teamwork could be the missing ingredient that enables the machines to improve their capabilities.31 This leads to a few key considerations for the years to come(figure 2):AI-to-AI communication.Agen

161、ts will likely have a more efficient way of communicating with each other than human language,as we dont need human-imitating chatbots talking to each other.32 Better AI-to-AI communication can enhance outcomes,as fewer people will need to become experts to benefit from AI.Rather,AI can adapt to eac

162、h persons communication style.33 Job displacement and creation.Some claim that roles such as prompt engineer could become obso-lete.34 However,the AI expertise of those employees will remain pertinent as they focus on managing,training,and collaborating with AI agents as they do with LLMs today.For

163、example,a lean IT team with AI experts might build the agents it needs in a sort of“AI factory”for the enterprise.The signif-icant shift in the remaining workforces skills and education may ultimately reward more human skills like creativity and design,as mentioned in previous Tech Trends.Privacy an

164、d security.The proliferation of agents with system access is likely to raise broad concerns about cybersecurity,which will only become more important as time progresses and more of our data is accessed by AI systems.New paradigms for risk and trust will be required to make the most out of applying A

165、I agents.Figure 2Compound AI journeyRetrieve dataSmall language model1Apply tools to analyze data and create insightsHuman2Create customer-facing social media content based on insightsSmall language model3Generate marketing images based on output from step 3Multimodal4Review for accuracy and appropr

166、iatenessHuman5Schedule the marketing post for the most opportune time,based on content and target audience.Repeat process as needed.Agentic6Source:Deloite research.22 Energy and resources.AIs energy consumption is a growing concern.35 To mitigate environmental impacts,future AI development will need

167、 to balance performance with sustainability.It will need to take advantage of improvements in liquid neural networks or other efficient forms of training AI,not to mention the hardware needed to make all of this work,as we discuss in“Hardware is eating the world.”Leadership for the future.AI has tra

168、nsformative potential,as everyone has heard plenty over the last year,but only insofar as leadership allows.Applying AI as a faster way of doing things the way theyve always been done will result in,at best,missed potential,and,at worst,amplified biases.36 Imaginative,courageous leaders should dare

169、to take AI from calcified best practices to the creation of“next practices,”where we find new ways of organizing ourselves and our data toward an AI-enabled world.When it comes to AI,enterprises will likely have the same considerations in the future that they do today:data,data,and data.Until AI sys

170、tems can reach arti-ficial general intelligence or learn as efficiently as the human brain,37 they will be hungry for more data and inputs to help them be more powerful and accurate.Steps taken today to organize,streamline,and protect enterprise data could pay dividends for years to come,as data deb

171、t could one day become the biggest portion of technical debt.Such groundwork should also help enterprises prepare for the litany of regulatory challenges and ethical uncertainties(such as data collection and use limitations,fairness concerns,lack of transparency)that come with shepherding this new,p

172、owerful technology into the future.38 The stakes of garbage in,garbage out are only going to grow:It would be much better to opt for genius in,genius squared.39 23Whats next for AI?1.Carl Franzen,“More than 70%of companies are experimenting with generative AI,but few are willing to commit more spend

173、ing,”VentureBeat,July 25,2023.2.Tom Dotan and Deepa Seetharaman,“For AI giants,smaller is sometimes better,”The Wall Street Journal,July 6,2024.3.Google Cloud,“Multimodal AI,”accessed October 2024.4.Silvia Pellegrino,“Which companies have partnered with OpenAI?,”Tech Monitor,May 15,2023;Maxwell Zeff

174、,“Anthropic launches Claude Enterprise plan to compete with OpenAI,”TechCrunch,September 4,2024;Jean Atelsek and William Fellows,“Hyperscalers stress AI credentials,optimization and developer empowerment,”S&P Global Market Intelligence,accessed October 2024.5.Gartner,“Gartner forecasts worldwide IT

175、spending to grow 8%in 2024,”press release,April 17,2024.GARTNER is a registered trademark and service mark of Gartner,Inc.and/or its affiliates in the U.S.and internationally and is used herein with permission.All rights reserved.6.Patricia Licatta,“Between sustainability and risk:Why CIOs are consi

176、dering small language models,”CIO,August 1,2024.7.Jim Rowan et al.,“Now decides next:Moving from potential to performance,”Deloittes State of Generative AI in the Enterprise Q3 report,August 2024;Mark Van de Wiel,“New AI survey:Poor data quality leads to$406 million in losses,”Fivetran,March 20,2024

177、.8.Rowan et al.,“Now decides next:Moving from potential to performance.”9.Sharon Goldman,“The hidden reason AI costs are soaringand its not because Nvidia chips are more expensive,”Fortune,August 23,2024.10.Deloitte Insights,“Lifting up the nonprofit sector through generative AI,”September 23,2024.1

178、1.Jim Rowan et al.,“Now decides next:Moving from potential to performance.”12.Ibid.13.Ibid.14.Sandra Erwin,“Booz Allen deploys advanced language model in space,”SpaceNews,August 1,2024;Argonne National Laboratory,“Smart diagnostics:How Argonne could use Generative AI to empower nuclear plant operato

179、rs,”press release,July 26,2024;Kevin Maik Jablonka et al.,“14 examples of how LLMs can transform materials science and chemistry:A reflection on a large language model hackathon,”Digital Discovery 5(2023).15.Phone interview with Vivek Mohindra,senior vice president of corporate strategy,Dell Technol

180、ogies,October 11,2024.16.Phone interview with Naveen Rao,vice president of AI at Databricks,October 2,2024.17.YouTube,“Introducing the next evolution of generative AI:Small language models,”Microsoft Dynamics 365,video,May 9,2024;Llama team,“The Llama 3 herd of models,”Meta,July 23,2024.18.Rachel Me

181、tz,“In AI,smaller,cheaper models are getting big attention,”Bloomberg,August 8,2024.19.Databricks,“AI is in production,”accessed October 2024.20.MIT Technology Review Insights,“Multimodal:AIs new frontier,”May 8,2024.21.Akesh Takyar,“Multimodal models:Architecture,workflow,use cases and development,

182、”LeewayHertz,accessed October 2024.22.NeuronsLab,“Multimodal AI use cases:The next opportunity in enterprise AI,”May 30,2024.23.Ellen Glover,“Multimodal AI:What it is and how it works,”Built In,July 1,2024.24.Mary E.Morrison,“At AMD,opportunities,challenges of using AI in marketing,”Deloittes CIO Jo

183、urnal for The Wall Street Journal,July 2,2024.25.NeuronsLab,“Multimodal AI use cases:The next opportunity in enterprise AI.”26.Oguz A.Acar,“AI prompt engineering isnt the future,”Harvard Business Review,June 6,2023.27.Salesforce,“Agentforce,”accessed October 2024;ServiceNow,“Our biggest AI release i

184、s here,”accessed October 2024.28.Phone interview with Chris Bedi,chief customer officer at ServiceNow,September 30,2024.29.Ibid.30.Brian Heater,“What is a liquid neural network,really?,”TechCrunch,August 17,2023.31.Edd Gent,“How teams of AI agents working together could unlock the techs true power,”

185、Singularity Hub,June 28,2024.32.Will Knight,“The chatbots are now talking to each other,”WIRED,October 12,2023.33.David Ellis,“The power of AI in modeling healthy communications,”Forbes,August 17,2023.34.Acar,“AI prompt engineering isnt the future.”35.James Vincent,“How much electricity does AI cons

186、ume?,”The Verge,February 16,2024.36.IBM,“Shedding light on AI bias with real world examples,”October 16,2023.37.University of Oxford,“Study shows that the way the brain learns is different from the way that artificial intelligence systems learn,”January 3,2024.38.Nestor Maslej et al.,The AI Index 20

187、24 annual report,AI Index Steering Committee,Institute for Human-Centered AI,Stanford University,Stanford,CA,April 2024.39.Deloitte,Work Re-Architected video series,accessed October 2024.Endnotes24Industry leadershipJim RowanHead of AI|Principal|Deloitte Consulting LLPJ|+1 617 437 3470Jim Rowan is a

188、 principal at Deloitte and is currently the Head of AI for Deloitte.He helps clients transform their businesses using data powered analytical and AI solutions that enable better decision making.Over the course of his career,Rowan has served clients across the life sciences,health care,and telecommun

189、ications indus-tries.He also has deep knowledge of the finance function in these organizations,having led analytics,planning and forecasting,and close projects that enable the finance function to embrace digital transformations.Rowan formerly led AI&Data Operations within Deloitte Consultings Strate

190、gy&Analytics practice.Nitin Mittal Global AI leader|Principal|Deloitte Consulting LLPNitin Mittal is a principal with Deloitte Consulting LLP.He currently serves as the US Artificial Intelligence(AI)Strategic Growth Offering Consulting leader and the Global Strategy,Analytics and M&A leader.He is th

191、e 2019 recipient of the AI Innovator of the Year award at the AI Summit New York.He specializes in advising clients to achieve competitive advantage through data and cognitive powered transformations that promote amplified intelligence and enable our clients to make strategic choices and transform a

192、head of disruption.Throughout his career,Mittal has served as a trusted advisor to global clients and has worked across a number of industry sectors.His primary focus has been working with life sciences and health care clients,implementing large scale data programs that promote organizational intell

193、igence,and the use of advanced analytics and AI to drive insights and business strategy.Lou DiLorenzo JrPrincipal|AI&Data Strategy Practice leader|US CIO&CDAO Programs,national leader|Deloitte Consulting LLP+1 612 397 4000|Lou DiLorenzo serves as the national leader of Deloitte Consultings AI&Data S

194、trategy practice and the Deloitte US CIO and CDAO Executive Accelerator programs.He is a member of Deloittes Generative AI practice leadership team and heads the Generative AI Incubator.With over 20 years of cross-sector operating,entre-preneurial,and consulting experience,he has a successful record

195、 of bringing key stakeholders together to help lead change,develop new capabilities,and deliver positive financial results.Previously,DiLorenzo served as COO of a consumer health insurance startup and as Global CIO for the Food Ingredients&Bio Industrial divi-sion at Cargill.He is a frequent technol

196、ogy contributor to leading publications and hosts the podcast,Techfluential.Continue the conversation25Whats next for AI?AcknowledgmentsMuch gratitude goes to the many subject matter leaders across Deloitte that contributed to our research for the information chapter:Lou DiLorenzo,Lena La,Nitin Mitt

197、al,Sanghamitra Pati,Jim Rowan,and Baris Sarer.2627Hardware is eating the worldAfter years of“software eating the world,”its hard-wares turn to feast.We previewed in the computation chapter of Tech Trends 2024 that as Moores Law comes to its supposed end,the promise of the AI revolution increasingly

198、depends on access to the appropriate hard-ware.Case in point:NVIDIA is now one of the worlds most valuable(and watched)companies,as specialized chips become an invaluable resource for AI computation workloads.According to Deloitte research based on a World Semiconductor Trade Statistics forecast,the

199、 market for chips used only for generative AI is projected to reach over US$50 billion this year.A critical hardware use case for enterprises may lie in AI-embedded end-user and edge devices.Take personal computers(PCs),for instance.For years,enterprise laptops have been commodified.But now,we may b

200、e on the cusp of a significant shift in computing,thanks to AI-embedded PCs.Companies like AMD,Dell,and HP are already touting the potential for AI PCs to“future-proof”technology infrastructure,reduce cloud comput-ing costs,and enhance data privacy.With access to offline AI models for image generati

201、on,text analysis,and speedy data retrieval,knowledge workers could be supercharged by faster,more accurate AI.That being said,enterprises should be strategic about refreshing end-user computation on a large scaletheres no use wasting AI resources that are limited in supply.Of course,all of these adv

202、ancements come at a cost.Data centers are a new focus of sustainability as the energy demands of large AI models continue to grow.4 The International Energy Agency has suggested that the demands of AI will significantly increase electric-ity in data centers by 2026,equivalent to Swedens or Germanys

203、annual energy demands.5 A recent Deloitte study on powering AI estimates that global data center electricity consumption may triple in the coming decade,largely due to AI demand.6 Innovations in energy sources and efficiency are needed to make AI hardware more accessible and sustainable,even as it p

204、roliferates and finds its way into everyday consumer and enterprise devices.Consider that Unit 1 of the nuclear plant Three Mile Island,which was shut down five years ago due to economic reasons,will reopen by 2028 to power data centers with carbon-free electricity.7Looking forward,AI hardware is po

205、ised to step beyond IT and into the Internet of Things.An increasing number of smart devices could become even more intelligent as AI enables them to analyze their usage and take on new tasks(as agentic AI,mentioned in“Whats next for AI?”advances).Todays benign use cases(like AI in toothbrushes)are

206、not indicative of tomorrows robust potential(like AI in lifesaving medical devices).8 The true power of hardware could be unlocked when smarter devices bring about a step change in our relationship with robotics.Now:Chips ahoy!A generation of technologists has been taught to believe software is the

207、key to return on investment,given its scalability,ease of updates,and intellectual property protections.9 But now,hardware investment is surging as computers evolve from calculators to cogitators.10 We wrote last year that specialized chips like graphics-pro-cessing units(GPUs)were becoming the go-t

208、o resources for training AI models.In its 2024 TMT Predictions report,Deloitte estimated that total AI chip sales in 2024 would be 11%of the predicted global chip market of Hardware is eating the worldThe AI revolution will demand heavy energy and hardware resourcesmaking enterprise infrastructure a

209、 strategic differentiator once againKelly Raskovich,Bill Briggs,Mike Bechtel,and Abhijith RavinutalaCOMPUTATION28US$576 billion.11 Growing from roughly$US50 billion today,the AI chip market is forecasted to reach up to US$400 billion by 2027,though a more conservative estimate is US$110 billion(figu

210、re 1).12Large tech companies are driving a portion of this demand,as they may build their own AI models and deploy specialized chips on-premises.13 However,enter-prises across industries are seeking compute power to meet their IT goals.For instance,according to a Databricks report,the financial serv

211、ices industry has had the highest growth in GPU usage,at 88%over the past six months,in running large language models(LLMs)that tackle fraud detection and wealth management.14 All of this demand for GPUs has outpaced capacity.In todays iteration of the Gold Rush,the companies provid-ing“picks and sh

212、ovels,”or the tools for todays tech transformation,are winning big.15 NVIDIAs CEO Jensen Huang has noted that cloud GPU capacity is mostly Figure 1The surge in AI hardware investmentUS$50 billion2024 projectionUS$400 bilion2027 optimistic forecastUS$110 billion2027 conservative forecastAI chip marke

213、t forecastsSource:Duncan Stewart et al.,“Gen AI chip demand fans a semi tailwind for now,”Deloite Insights,November 29,2023.29Hardware is eating the worldfilled,but the company is also rolling out new chips that are significantly more energy-efficient than previous iterations.16 Hyperscalers are buy

214、ing up GPUs as they roll off the production line,spending almost$US1 trillion on data center infrastructure to accommodate the demand from clients who rent GPU usage.17 All the while,the energy consumption of existing data centers is pushing aging power grids to the brink globally.18Understandably,e

215、nterprises are looking for new solu-tions.While GPUs are crucial for handling the high workloads of LLMs or content generation,and central processing units are still table stakes,neural processing units(NPUs)are now in vogue.NPUs,which mimic the brains neural network,can accelerate smaller AI work-l

216、oads with greater efficiency and lower power demands,19 enabling enterprises to shift AI applications away from the cloud and apply AI locally to sensitive data that cant be hosted externally.20 This new breed of chip is a crucial part of the future of embedded AI.Vivek Mohindra,senior vice presiden

217、t of corporate strat-egy at Dell Technologies,says,“Of the 1.5 billion PCs in use today,30%are four years old or more.None of these older PCs have NPUs to take advantage of the latest AI PC advancements.”21 A great refresh of enterprise hard-ware may be on the horizon.As NPUs enable end-user devices

218、 to run AI offline and allow models to become smaller to target specific use cases,hardware may once again be a differentiator for enterprise performance.In a recent Deloitte study,72%of respondents believe generative AIs impact on their industry will be“high to transformative.”22 Once AI is at our

219、fingertips thanks to mainstream hardware advancements,that number may edge closer to 100%.New:Infrastructure is strategic againThe heady cloud-computing highs of assumed unlim-ited access are giving way to a resource-constrained era.After being relegated to a utility for years,enterprise infrastruct

220、ure(for example,PCs)is once again strategic.Specifically,specialized hardware will likely be crucial to three significant areas of AI growth:AI-embedded devices and the Internet of Things,data centers,and advanced physical robotics.While the impact on robot-ics may occur over the next few years,as w

221、e discuss in the next section,we anticipate that enterprises will be grappling with decisions about the first two areas over the next 18 to 24 months.While AI scarcity and demand persist,the following areas may differentiate leaders from laggards.Edge footprint By 2025,more than 50%of data could be

222、generated by edge devices.23 As NPUs proliferate,more and more devices could be equipped to run AI models without rely-ing on the cloud.This is especially true as generative AI model providers opt for creating smaller,more efficient models for specific tasks,as discussed in“Whats next for AI?”With q

223、uicker response times,decreased costs,and greater privacy controls,hybrid computing(that is,a mix of cloud and on-device AI workloads)could be a must-have for many enterprises,and hardware manu-facturers are betting on it.24 According to Dell Technologies Mohindra,processing AI at the edge is one of

224、 the best ways to handle the vast amounts of data required.“When you consider latency,network resources,and just sheer volume,moving data to a centralized compute location is inefficient,ineffec-tive,and not secure,”he says.“Its better to bring AI to the data,rather than bring the data to AI.”25 One

225、 major bank predicts that AI PCs will account for more than 40%of PC shipments in 2026.26 Similarly,nearly 15%of 2024 smartphone shipments are predicted to be capable of running LLMs or image-generation models.27 Alex Thatcher,senior director of AI PC expe-riences and cloud clients at HP,believes th

226、at the refresh in devices will be akin to the major transition from command-line inputs to graphical user interfaces that changed PCs in the 1990s.“The software has funda-mentally changed,replete with different tools and ways of collaborating,”he says.“You need hardware that can accelerate that chan

227、ge and make it easier for enterprises to create and deliver AI solutions.”28 Finally,Apple and Microsoft have also fueled the impending hardware refresh by embedding AI into their devices this year.29 As choices proliferate,good governance will be crucial,and enterprises have to ask the question:How

228、 many of our people need to be armed with next-generation devices?Chip manufacturers are in a race to improve AI horsepower,30 but enterprise customers cant afford to refresh their entire edge footprint with each new 30advancement.Instead,they should develop a strategy for tiered adoption where thes

229、e devices can have the most impact.Build versus buy For buying or renting specialized hardware,organi-zations may typically consider their cost model over time,the expected time frame of use,and the necessity for progress.However,AI is applying another level of competitive pressure to this decision.

230、With hardware like GPUs still scarce and the market clamoring for AI updates from all organizations,many companies have been tempted to rent as much computing power as possible.Organizations may struggle to take advantage of AI if they dont have their data enablement in order.Rather than scrambling

231、for GPUs,it may be more efficient to understand where the organization is ready for AI.Some areas may concern private or sensitive data;investing in NPUs can keep those workloads offline,while others may be fine for the cloud.Thanks to the lessons of cloud in the past decade,enterprises know that th

232、e cost of runaway models operating on runaway hardware can quickly balloon.31 Pushing these costs to operating expenditure may not be the best answer.Some estimates even say that GPUs are underutilized.32 Thatcher believes enterprise GPU utilization is only 15%to 20%,a problem that HP is addressing

233、through new,efficient methods:“Weve enabled every HP workstation to share its AI resources across our enterprise.Imagine the ability to search for idle GPUs and use them to run your workloads.Were seeing up to a sevenfold improve-ment in on-demand computing acceleration,and this could soon be indust

234、ry standard.”33In addition,the market for AI resources on the cloud is ever-changing.For instance,concerns around AI sover-eignty are increasing globally.34 While companies around the world approved running their e-commerce platforms or websites on American cloud servers,the applicabil-ity of AI to

235、national intelligence and data management makes some hesitant to place AI workloads overseas.This opens up a market for new national AI cloud providers or private cloud players.35 GPU-as-a-service computing startups are an alternative to hyperscalers.36 This means that the market for renting compute

236、 power may soon be more fragmented,which could give enter-prise customers more options.Finally,AI may be top of mind for the next two years,but todays build versus buy decisions could have impacts beyond AI considerations.Enterprises may soon consider using quantum computing for the next generation

237、of cryptography(especially as AI ingests and transmits more sensitive data),optimization,and simulation,as we discuss in“The new math:Solving cryptography in an age of quantum.”Data center sustainabilityMuch has been said about the energy use of data centers running large AI models.Major bank report

238、s have ques-tioned whether we have the infrastructure to meet AI demand.37 The daily power usage of major chatbots has been equated to the daily consumption of nearly 180,000 US households.38 In short,AI requires unprecedented resources from data centers,and aging power grids are likely not up to th

239、e task.While many companies may be worried about getting their hands on AI chips like GPUs to run workloads,sustainability may well be a bigger issue.Currently,multiple advancements that aim to make AI more sustainable are underway.Enterprises should take note of advancements in these areas over the

240、 next two years when considering data centers for AI(figure 2):Renewable sources:Pressure is mounting on the providers of data centers and AI-over-the-cloud to find sustainable energy sourcesand the rapidly growing focus on AI may help transition the overall economy to renewables.39 Major tech compa

241、nies are already exploring partnerships with nuclear energy providers.40 Online translation service DeepL hosts a data center in Iceland thats cooled by the naturally frigid air and is fully powered by geothermal and hydroelectric power.41 And in El Salvador,compa-nies are even exploring how they co

242、uld power data centers with volcanos.42 Sustainability applications:While building AI consumes a lot of energy,applying AI can,in many cases,offset some of these carbon costs.AI is already 31Hardware is eating the worldbeing used to map and track deforestation,melting icebergs,and severe weather pat

243、terns.It can also help companies track their emissions and be more efficient in using data centers.43 Hardware improvements:New GPUs and NPUs have already saved energy and cost for enter-prises.Innovation is not stalling.Intel and Global Foundries recently unveiled new chips that can use light,rathe

244、r than electricity,to transmit data.44 This could revolutionize data centers,enabling reduced latency,more distributed construction,and improved reliability.While this fiber optic approach is expensive now,costs may come down over the next couple of years,enabling this type of chip to become mainstr

245、eam.Finally,an infrastructure resurgence wouldnt be complete without a nod to connectivity.As edge devices proliferate and companies rely on renting GPU usage from data centers,the complexities of interconnectivity could multiply.High-performance interconnect tech-nologies like NVIDIAs NVLink are al

246、ready primed for communications between advanced GPUs and other chips.45 Advancements in 6G can integrate global terrestrial and non-terrestrial networks(like satellites)for ubiquitous connectivity,such that a company in Cape Town relying on a data center in Reykjavik has minimal lag.46 As The Wall

247、Street Journal has noted,the AI transfor-mation for enterprises is akin to the transition to elec-tric that many car manufacturers are experiencing.47 Technology infrastructure needs to be rethought on a component-by-component basis,and the decisions made today around edge footprint,investment in sp

248、ecialized hardware,and sustainability can have lasting impacts.Next:We were promised robotsIf todays hardware requires a strategic refresh,enter-prises may have much more on their plates in the next decade when robotics become mainstream and smart devices become worthy of their label.Consider the ex

249、am-ple of the latest smart factories,which use a cascade of computer vision,ubiquitous sensors,and data to build machines that can learn and improve as they manufac-ture products.48 Instead of simply providing readings or adjusting on one parameter,like a thermostat,mesh networks of multiple AI-embe

250、dded devices can create collaborative compute environments and orchestrate diverse resources.49Another form of smart factory is being developed by Mytra,a San Franciscobased company that simplifies the manual process of moving and storing warehouse materials.The company has developed a fully modular

251、 storage system composed of steel“cubes,”which can Figure 2Advancements in areas related to AI requirements ConsiderImplementRenewable sourcesTracking the energy costs of AI on cloudSeek out innovative sustainability solutionsEnergy-saving applicationsApplying AI to discover potential energy savings

252、Optimize emissions tracking and data usageHardware improvementsMonitoring technological advancements in AIInvest in new energy-efcient chipsSource:Deloite research.32be assembled together in any shape that supports 3D movement and storage of material within,manipulated by robots and optimized throug

253、h software.50 Chris Walti,chief executive officer of Mytra,believes this modu-lar approach unlocks automation for any number of unpredictable future applications:“Its one of the first general-purpose computers for moving matter around in 3D space.”51 Walti believes there is immense potential to appl

254、y robot-ics to relatively constrained problems,such as moving material in a grid or driving a vehicle in straight lines.52 Until now,in many cases,a good robot has been hard to find.Sustainability,security,and geopolitics are all salient concerns for such a technology.And thats after we even muster

255、the infrastructure noted earlier,including data,network architecture,and chip availability,to make such a leap forward possible.As the saying goes,“hard-ware is hard.”53 Over the next decade,advancements in robotics applied to more and more complex situa-tions could revolutionize the nature of manuf

256、acturing and other physical labor.The potential leads directly to humanoid roboticsbots that are dynamic,constantly learning,and capable of doing what we do.Economists and businesses alike have argued that aging populations and labor shortages necessitate greater investment in robotics and automatio

257、n.54 In many cases,this entails large industrial robots completing relatively simple tasks,as noted above,but more complex tasks require“smarter”mechanical muscle that can move around as humans do.Take the example of Figure AIs humanoid robots tested at the BMW plant in Spartanburg,South Carolina.55

258、 The autonomous robot,through a combination of computer vision,neural networks,and trial and error,successfully assembled parts of a car chassis.56As the furthest star of progress in this realm,we might anticipate humanoid robots performing a broad vari-ety of tasks,from cleaning sewers to ferrying

259、materials between hospital rooms or even performing surgeries.57 Just as AI is currently transforming knowledge work,the increased presence of robots could greatly affect physical work and processes in manufacturing and beyond.In both cases,companies should be sure to find ways for humans and machin

260、es to work together more efficiently than either could do alone.Labor shortages addressed by robotics should then free up human time for more of the uniquely creative and complex tasks where we thrive.As the author Joanna Maciejewska has said astutely,“I want AI to do my laundry and dishes so that I

261、 can do art and writing,not for AI to do my art and writing so that I can do my laundry and dishes.”58 33Hardware is eating the world1.Jon Quast,“Artificial intelligence(AI)juggernaut Nvidia is one of the worlds most valuable companies.Heres what investors should know,”The Motley Fool,June 22,2024.2

262、.Duncan Stewart et al.,“Gen AI chip demand fans a semi tailwind for now,”Deloitte Insights,November 29,2023;World Semiconductor Trade Statistics(WSTS),“Semiconductor market forecast spring 2023,”June 6,2023.3.Rob Enderle,“AMD enters AI PC race,closes Microsoft Copilot+launch gaps,”TechNewsWorld,July

263、 15,2024;Saba Prasla,“Meet the future of computing with AI PCs,”Dell Blog,May 31,2024;HP,“HP unveils industrys largest portfolio of AI PCs,”press release,March 7,2024.4.Taiba Jafari et al.,“Projecting the electricity demand growth of generative AI large language models in the US,”Center on Global En

264、ergy Policy,July 17,2024.5.International Energy Agency,Electricity 2024:Analysis and forecast to 2026,revised May 2024.6.Deloitte,“Powering artificial intelligence,”accessed November 18,2024.7.Constellation,“Constellation to launch Crane Clean Energy Center,restoring jobs and carbon-free power to th

265、e grid,”press release,September 20,2024.8.Shira Ovide,“This$400 toothbrush is peak AI mania,”The Washington Post,April 5,2024;David Niewolny,“Boom in AI-enabled medical devices transforms healthcare,”NVIDIA Blog,March 26,2024.9.Marc Andreessen,“Why software is eating the world,”Andreessen Horowitz,A

266、ugust 20,2011.10.John Thornhill,“How hardware is(still)eating the world,”The Financial Times,February 15,2024.11.Stewart et al.,“Gen AI chip demand fans a semi tailwind for now.”12.Ibid.13.NVIDIA,“NVIDIA hopper GPUs expand reach as demand for AI grows,”press release,March 21,2023.14.Databricks,State

267、 of data+AI,accessed October 2024.15.John Thornhill,“The likely winners of the generative AI gold rush,”The Financial Times,May 11,2023.16.Matt Ashare,“Nvidia sustains triple-digit revenue growth amid AI building boom,”CIO Dive,August 29,2024;NVIDIA,“Nvidia(NVDA)Q2 2025 earnings call transcript,”The

268、 Motley Fool,August 28,2024;Dean Takahashi,“Nvidia unveils next-gen Blackwell GPUs with 25X lower costs and energy consumption,”VentureBeat,March 18,2024.17.Matt Ashare,“Big tech banks on AI boom as infrastructure spending heads for trillion-dollar mark,”CIO Dive,August 5,2024;DellOro Group,“Worldwi

269、de data center capex to grow at a 24 percent CAGR by 2028,”press release,August 1,2024.18.Evan Halper,“Amid explosive demand,America is running out of power,”The Washington Post,March 7,2024.19.Chris Hoffman,“What the heck is an NPU,anyway?Heres an explainer on AI chips,”PCWorld,September 18,2024.20

270、.Anshel Sag,“At the heart of the AI PC battle lies the NPU,”Forbes,April 29,2024.21.Phone interview with Vivek Mohindra,senior vice president of corporate strategy,Dell Technologies,October 11,2024.22.Christie Simons et al.,2024 global semiconductor industry outlook,Deloitte,2024.23.Aditya Agrawal,“

271、The convergence of edge computing and 5G,”Control Engineering,August 7,2023;Baris Sarer et al.,“AI and the evolving consumer device ecosystem,”Deloittes CIO Journal for The Wall Street Journal,April 24,2024.24.Matthew S.Smith,“When AI unplugs,all bets are off,”IEEE Spectrum,December 1,2023.25.Phone

272、interview with Vivek Mohindra,senior vice president of corporate strategy,Dell Technologies,October 11,2024.26.Patrick Seitz,“AI PCs are here.Let the upgrades begin,computer makers say,”Investors Business Daily,July 5,2024;Sam Reynolds,“AI-enabled PCs will drive PC sales growth in 2024,say research

273、firms,”Computerworld,January 11,2024.27.Phil Solis et al.,“The future of next-gen AI smartphones,”IDC,February 19,2024.28.Phone interview with Alex Thatcher,senior director of AI PC experiences and cloud clients at HP,October 4,2024.29.Rob Waugh,“Assessing Apple Intelligence:Is new on-device AI smar

274、t enough for the enterprise?,”The Stack,September 12,2024;Matt OBrien,“Microsofts new AI-enabled laptops will have a photographic memory of your virtual activity,”Fortune,May 20,2024.Tech Trends is an independent publication and has not been authorized,sponsored,or otherwise approved by Apple Inc.30

275、.Luke Larsen,“AMD just won the AI arms race,”Digital Trends,June 3,2024.31.David Linthicum,“Learning cloud cost management the hard way,”InfoWorld,July 16,2024.32.Tobias Mann,“Big Cloud deploys thousands of GPUs for AI yet most appear under-utilized,”The Register,January 15,2024.33.Phone interview w

276、ith Alex Thatcher,senior director of AI PC experiences and cloud clients at HP,October 4,2024.34.Christine Mui,“Welcome to the global AI sovereignty race,”Politico,September 18,2024.35.Ibid.36.Bobby Clay,“Graphics processing service providers step up to meet demand for cloud resources,”S&P Global Ma

277、rket Intelligence,July 19,2024.37.Goldman Sachs,Top of Mind 129,June 25,2024.38.Cindy Gordon,“ChatGPT and generative AI innovations are creating sustainability havoc,”Forbes,March 12,2024.39.Molly Flanagan,“AI and environmental challenges,”Environmental Innovations Initiative,accessed October 2024;D

278、eloitte,“Powering artificial intelligence.”40.Jennifer Hiller and Sebastian Herrera,“Tech industry wants to lock up nuclear power for AI,”The Wall Street Journal,July 1,2024.41.Robert Scheier,“4 paths to sustainable AI,”CIO,January 31,2024.42.Tom Dotan and Asa Fitch,“Why the AI industrys thirst for

279、new data centers cant be satisfied,”The Wall Street Journal,April 24,2024.Endnotes3443.Victoria Masterson,“9 ways AI is helping tackle climate change,”World Economic Forum,February 12,2024.44.Kirk Ogunrinde,“Intel is using lasers to help meet AI demands on data centers,”Forbes,June 26,2024.45.Rick M

280、erritt,“What is NVLink?,”NVIDIA,March 6,2023.46.Garry Kranz,“What is 6G?Overview of 6G networks&technology,”TechTarget,last updated November 2023.47.Steven Rosenbush,“AI will force a transformation of tech infrastructure,”The Wall Street Journal,September 11,2024.48.Majeed Ahmad,“Sensor fusion with

281、AI transforms the smart manufacturing era,”EE Times,July 26,2023.49.Melissa Malec,“AI orchestration explained:The what,why&how for 2024,”HatchWorks AI,last updated June 6,2024.50.Phone interview with Chris Walti,chief executive officer of Mytra,October 11,2024.51.Ibid.52.Ibid.53.Sara Holoubek and Je

282、ssica Hibbard,“Why hardware is hard,”Luminary Labs,accessed October 2024.54.Peter Dizikes,“Study:As a population gets older,automation accelerates,”MIT News,September 15,2021;Hans Peter Bronomo,“Inside Googles 7-year mission to give AI a robot body,”WIRED,September 10,2024.55.BMW Group,“Successful t

283、est of humanoid robots at BMW Group Plant Spartanburg,”press release,August 6,2024.56.Ibid.57.Viktor Doychinov,“An army of sewer robots could keep our pipes clean,but theyll need to learn to communicate,”The Conversation,January 26,2021;Case Western Reserve University,“5 medical robots making a diff

284、erence in healthcare,”Online Engineering Blog,accessed October 2024;National Institute of Biomedical Imaging and Bioengineering(NIBIB),“Robot performs soft tissue surgery with minimal human help,”press release,April 20,2022.58.Joanna Maciejewskas post on X,March 29,2024.35Hardware is eating the worl

285、dIndustry leadershipNitin Mittal Global AI leader|Principal|Deloitte Consulting LLPNitin Mittal is a principal with Deloitte Consulting LLP.He currently serves as the US Artificial Intelligence(AI)Strategic Growth Offering Consulting leader and the Global Strategy,Analytics and M&A leader.He is the

286、2019 recipient of the AI Innovator of the Year award at the AI Summit New York.He specializes in advising clients to achieve competitive advantage through data and cognitive powered transformations that promote amplified intelligence and enable our clients to make strategic choices and transform ahe

287、ad of disruption.Throughout his career,Mittal has served as a trusted advisor to global clients and has worked across a number of industry sectors.His primary focus has been working with life sciences and health care clients,implementing large scale data programs that promote organizational intellig

288、ence,and the use of advanced analytics and AI to drive insights and business strategy.Abdi GoodarziUS Enterprise Performance Portfolio leader|Principal+1 714 913 1091|Abdi Goodarzi is a principal with Deloitte Consulting LLP,leading Deloittes Enterprise Performance(EP)Offerings Portfolio.This portfo

289、lio of six offerings provide strategy,implement and operate services for variety of enterprise functions,from end-to-end busi-ness and IT transformation,to digital supply chain optimization,manufacturing and product strategies,and procurement as-a-ser-vice,to global finance,shared services,planning,

290、ITSM,and full scale AMS and BPO.This portfolio offers competency in many ERP platforms such as SAP,Oracle,Workday Financials and Infor,in addition to ServiceNow,Anaplan,Ariba,and Coupa,as well as real estate solutions such as Nuvolo,as well as PLM,planning and fulfillment,and engineering solutions s

291、uch as Siemens,PTC,O9,OMP and IBP.AcknowledgmentsMuch gratitude goes to the many subject matter leaders across Deloitte that contributed to our research for the Computation chapter:Lou DiLorenzo,Abdi Goodarzi,Lena La,Nitin Mittal,Manish Rajendran,Jim Rowan,and Baris Sarer.Continue the conversation36

292、37IT,amplified:AI elevates the reach(and remit)of the tech functionMuch has been said,including within the pages of Tech Trends,about the potential for artificial intelligence to revolutionize business use cases and outcomes.Nowhere is this more true than in the end-to-end life cycle of soft-ware en

293、gineering and the broader business of informa-tion technology,given generative AIs ability to write code,test software,and augment tech talent in general.Deloitte research has shown that tech companies at the forefront of this organizational change are ready to real-ize the benefits:They are twice a

294、s likely as their more conservative peers to say generative AI is transforming their organization now or will within the next year.1 We wrote in a Tech Trends 2024 article that enterprises need to reorganize their developer experiences to help IT teams achieve the best results.Now,the AI hype cycle

295、has placed an even greater focus on the tech functions ways of working.IT has long been the lighthouse of digital transformation in the enterprise,but it must now take on AI transformation.Forward-thinking IT leaders are using the current moment as a once-in-a-generation opportunity to redefine role

296、s and responsibilities,set investment priorities,and communicate value expecta-tions.More importantly,by playing this pioneering role,chief information officers can help inspire other tech-nology leaders to put AI transformation into practice.After years of enterprises pursuing lean IT and every-thi

297、ng-as-a-service offerings,AI is sparking a shift away from virtualization and austere budgets.Gartner predicts that“worldwide IT spending is expected to total$5.26 trillion in 2024,an increase of 7.5%from 2023.”2 As we discuss in“Hardware is eating the world,”hard-ware and infrastructure are having

298、a moment,and enter-prise IT spending and operations may shift accordingly.As both traditional AI and generative AI become more capable and ubiquitous,each of the phases of tech deliv-ery may see a shift from human in charge to human in the loop.Organizations need a clear strategy in place before tha

299、t occurs.Based on Deloitte analysis,over the next 18 to 24 months,IT leaders should plan for AI transformation across five key pillars:engineering,talent,cloud financial operations(FinOps),infrastructure,and cyber risk.This trend may usher in a new type of lean IT over the next decade.If commercial

300、functions see an increased number of citizen developers or digital agents that can spin up applications on a whim,the role of the IT func-tion may shift from building and maintaining to orches-trating and innovating.In that case,AI may not only be undercover,as we indicate in the introduction to thi

301、s years report,but may also be overtly in the boardroom,overseeing tech operations in line with human needs.Now:Spotlightand higher spendingon IT For years,IT has been under pressure to streamline sprawling cloud spend and curb costs.Since 2020,however,investments in tech have been on the rise thank

302、s to pent-up demand for collaboration tools and the pandemic-era emphasis on digitalization.3 According IT,amplified:AI elevates the reach(and remit)of the tech functionAs the tech function shifts from leading digital transformation to leading AI transformation,forward-thinking leaders are using thi

303、s as an opportunity to redefine the future of ITKelly Raskovich,Bill Briggs,Mike Bechtel,and Abhijith RavinutalaBUSINESS OF TECHNOLOGY38to Deloitte research,from 2020 to 2022,the global average technology budget as a percentage of revenue jumped from 4.25%to 5.49%,an increase that approx-imately dou

304、bled the previous revenue change from 2018 to 2020.4 And in 2024,US companies average budget for digital transformation as a percentage of revenue is 7.5%,with 5.4%coming from the IT budget.5As demand for AI sparks another increase in spending,the finding from Deloittes 2023 Global Technology Leader

305、ship Study continues to ring true:Technology is the business,and tech spend is increasing as a result.Today,enterprises are grappling with the new rele-vance of hardware,data management,and digitization in ramping up their usage of AI and realizing its value potential.In Deloittes Q2 State of Genera

306、tive AI in the Enterprise report,businesses that rated themselves as having“very high”levels of expertise in generative AI were increasing their investment in hardware and cloud consumption much more than the average enterprise.6 Overall,75%of organizations surveyed have increased their investments

307、around data-life-cycle management due to generative AI.7 These figures point to a common theme:To realize the highest impact from gen AI,enterprises likely need to accelerate their cloud and data modernization efforts.AI has the potential to deliver efficiencies in cost,inno-vation,and a host of oth

308、er areas,but the first step to accruing these benefits is for businesses to focus on making the right tech investments.8 Because of these crucial investment strategies,the spotlight is on tech leaders who are paving the way.According to Deloitte research,over 60%of US-based technology leaders now re

309、port directly to their chief exec-utives,an increase of more than 10 percentage points since 2020.9 This is a testament to the tech leaders increased importance in setting the AI strategy rather than simply enabling it.Far from a cost center,IT is increasingly being seen as a differentiator in the A

310、I age,as CEOs,following market trends,are keen on staying abreast of AIs adoption in their enterprise.10 John Marcante,former global CIO of Vanguard and US CIO-in-residence at Deloitte,believes AI will fundamen-tally change the role of IT.He says,“The technology organization will be leaner,but have

311、a wider purview.It will be more integrated with the business than ever.AI is moving fast,and centralization is a good way to ensure organizational speed and focus.”11As IT gears up for the opportunity presented by AIperhaps the opportunity that many tech leaders and employees have waited forchanges

312、are already under-way in how the technology function organizes itself and executes work.The stakes are high,and IT is due for a makeover.New:An AI boost for ITOver the next 18 to 24 months,the nature of the IT func-tion is likely to change as enterprises increasingly employ generative AI.Deloittes f

313、oresight analysis suggests that,by 2027,even in the most conservative scenario,gen AI will be embedded into every companys digital product or software footprint(figure 1),as we discuss across five key pillars.12Engineering In the traditional software development life cycle,manual testing,inexperienc

314、ed developers,and disparate tool environments can lead to inefficiencies,as weve discussed in prior Tech Trends.Fortunately,AI is already having an impact on these areas.AI-assisted code gener-ation,automated testing,and rapid data analytics all save developers more time for innovation and feature d

315、evelopment.The productivity gain from coding alone is estimated to be worth US$12 billion in the United States alone.13 At Google,AI tools are being rolled out internally to developers.In a recent earnings call,CEO Sundar Pichai said that around 25 percent of the new code at the tech-nology giant is

316、 developed using AI.Shivani Govil,senior director of product management for developer products,believes that“AI can transform how engineering teams work,leading to more capacity to innovate,less toil,and higher developer satisfaction.Googles approach is to bring AI to our users and meet them where t

317、hey areby bringing the technology into products and tools that developers use every day to support them in their work.Over time,we can create even tighter alignment between the code and business requirements,allowing faster feedback loops,improved product market fit,and 39IT,amplified:AI elevates th

318、e reach(and remit)of the tech functionbetter alignment to the business outcomes.”14 In another example,a health care company used COBOL code assist to enable a junior developer with no experience in the programming language to generate an explanation file with 95%accuracy.15 As Deloitte recently sta

319、ted in a piece on engineering in the age of gen AI,the developer role is likely to shift from writing code to defining the architecture,reviewing code,and orchestrating functionality through contextu-alized prompt engineering.Tech leaders should anticipate human-in-the-loop code generation and revie

320、w to be the standard over the next few years of AI adoption.16 Figure 1How generative AI might transform IT ways of workingOver the next 18 to 24 months,enterprises may experience vast improvement in their technology teams as generative AI is increasingly embedded into ways of working.Deloittes fore

321、sight analysis suggests that by 2027,even in the most conservative scenario,gen AI will be embedded into every companys digital product/software footprint.Manual and time-consuming processes like code reviews,infrastructure configuration,and budget management can be automated and improved,as we move

322、 from current to target state of AI in IT.Source:Deloite research and analysis.The problem Necessary changesRecommended actionsEngineeringTalentCloud financial operationsInfrastructureCyberManual,inefcient aspects of thetraditional sofware developmentlife cycle Shif from writing code to defining the

323、architecture,reviewing code,andorchestrating functionality Tech leaders should expect human-in-the-loop code generation and review to becomethe standard Executives struggle to hire workers withthe right backgrounds and are forced todelay projects AI can generate rich learning anddevelopment media as

324、 well asdocumentation to upskill talent Tech leaders should implement regular AI-powered learning recommendations andpersonalization as a new way of working Runaway spend is common in the cloud,since resources can be provisioned with a click AI-powered cost analysis,paterndetection,and resource allo

325、cation can optimize IT spend at new speedsLeaders should consistently apply AI to help it earn its keep and optimize costsNearly half of enterprises are handling tasks like security,compliance,and service management on a manual basisAutomated resource allocation,predictivemaintenance,and anomaly det

326、ectioncould revolutionize IT systems Leaders should work toward an ITinfrastructure that can heal itself as needed through AI Generative AI and digital agents open upmore atack surfaces than ever for bad actors Automated data masking,incidentresponse,and policy generation can optimize cybersecurity

327、responses Enterprises should take steps to further authenticate data and digital media throughnew tech or processes 40Talent Technology executives surveyed by Deloitte last year noted that they struggle to hire workers with critical IT backgrounds in security,machine learning,and software architectu

328、re,and are forced to delay projects with finan-cial backing due to a shortage of appropriately skilled talent.17 As AI becomes the newest skill in demand,many companies may not even be able to find all the talent they need,leading to a hiring gap wherein nearly 50%of AI-related positions cannot be f

329、illed.18As a result,tech leaders should focus on upskilling their own talent,another area where AI can help.Consider the potential benefits of AI-powered skills gap analyses and recommendations,personalized learning paths,and virtual tutors for on-demand learning.Bayer,the life sciences company,has

330、used generative AI to summarize procedural documents and generate rich media such as animation for e-learning.19 Along the same lines,AI could generate documentation to help a new developer under-stand a legacy technology,and then create an associated learning podcast and exam for that same develope

331、r.At Google,developers thrive on hands-on experience and problem-solving,so leaders are keen to provide AI learning and tools(like coding assistants)that meet developers where they are on their learning journey.“We can use AI to enhance learning,in context with emerging technologies,in ways that ant

332、icipate and support the rapidly changing skills and knowledge required to adapt to them,”says Sara Ortloff,senior director of developer experience at Google.20 As automation increases,tech talent would take an over-sight role and enjoy more capacity to focus on innovation that can improve the bottom

333、 line(as we wrote about last year).This could help attract talent since,according to Deloitte research,the biggest incentive that attracts tech talent to new opportunities is the work they would do in the role.21Cloud financial operationsRunaway spending became a common problem in the cloud era when resources could be provisioned with a click.Hyperscalers have offered data and tooling for finance