《Thoughtworks：2025科技棱鏡年度報告：五大框架解鎖技術戰略新范式（英文版）（50頁）.pdf》由會員分享，可在線閱讀，更多相關《Thoughtworks：2025科技棱鏡年度報告：五大框架解鎖技術戰略新范式（英文版）（50頁）.pdf（50頁珍藏版）》請在三個皮匠報告上搜索。

1、Looking GlassBringing tech-led business changes into focus2025Looking Glass Thoughtworks,Inc.All Rights Reserved.2Introduction 3Operationalizing AI for business impact 4Strengthening the data value chain 11Reimagining responsible tech for the era of generative AI 18Enabling richer experiences throug

2、h multimodal interactions 26Unlocking greater value from physical-digital convergence 34Glossary 42Looking GlassIntroductionWelcome to the Looking Glass 2025.Unlike many tech trend reports,Thoughtworks Looking Glass is not intended to shine a light on the latest buzzwords.Instead,we take a long term

3、 look at the technology horizons and explore what that means for businesses.What are the things you need to know about now?And whats likely to be important in the longer term?The Looking Glass enables you to understand and interpret emerging technologies so you can make sound,strategic choices for y

4、our organization.The relentless speed of technological advancement makes it harder to predict whats coming and where your investments will pay off the most.Breakthroughs in areas such as agentic AI promise to upend how we think about technology.But how quickly should you prepare to adapt?Heres where

5、 Thoughtworks Looking Glass comes in.In this edition,we explore more than 90 trends through five distinct perspectives that define the evolving tech landscape in business.Some of these trends are already transforming operations,while others remain just over the horizon,sparking interest and debate b

6、ut still unfolding.For business leaders,keeping a broad,strategic perspective on these developments both current and future is essential.Looking Glass offers exactly that:a framework to gain a comprehensive understanding of key trends.The five lenses provide clarity and focus,helping ensure your org

7、anization remains adaptable,resilient and ready to harness or respond to the inevitable shifts in technology that shape our modern world.Rachel Laycock Chief Technology Officer,ThoughtworksLooking Glass Thoughtworks,Inc.All Rights Reserved.4Operationalizing AI for business impactThe mainstreaming of

8、 AI and generative AI in particular is continuing apace.But as AI proliferates,its more evident that successfully operationalizing AI models and bringing them to production remains a challenge.From questionable output to unintended consequences,there are a host of real and projected scenarios that p

9、revent organizations from leveraging AI to its full potential.Enterprises continue to struggle with data quality,data accessibility and the challenges of data at scale,all of which remain foundational to robust,effective AI.As our data platform lens explores,careful data curation,and effective data

10、engineering and architecture are essential.The importance of synthetic data,particularly in research contexts,as a tool to avoid privacy and data integrity issues is also becoming more and more apparent.Organizations also need to develop better approaches to the evaluation and control of AI systems.

11、Forward-looking enterprises are adopting evals tests of AI output to determine reliability,accuracy and relevance and guardrails,programmed policy layers that mitigate the inherent unpredictability of generative systems.As adoption increases,improving the mechanisms through which AI systems are conn

12、ected with enterprise applications grows more important.Proxy services are emerging to help developers link AI models with the applications they build.Looking Glass Thoughtworks,Inc.All Rights Reserved.5AI agents are sometimes positioned as the next step in the evolution of AI,due to their capacity

13、to mimic human reasoning.However,the technology remains relatively new,and finding applications for agents requires domain expertise,as well as the ability to precisely map and model complex processes and interactions.To build a sustainable and productive AI practice,its vital that the organization

14、doesnt resort to shortcuts,acquires the requisite skills and keeps innovation rooted in business realities.Operationalizing AI for business impact“The lessons from automation endeavors in the 80s could help to build the right level of human-AI agent handovers.We must focus on augmenting humans rathe

15、r than trying to substitute their current tasks completely.”Srinivasan RaguramanTechnical Principal,ThoughtworksSignals The emergence of small language models,such as Microsofts phi-3,and AMDs AMD 135.These make it possible to run AI models at the edge of networks on devices like mobile phones,and b

16、ecause they are relatively lightweight,focused and efficient,have a range of positive business,security and sustainability implications.LLMs also continue to evolve,with Anthropics Claude 3.5 Sonnet LLM,which has set industry benchmarks in terms of performance,recently upgraded to include computer u

17、se capabilities.Research showing that for many organizations,AI investments and adoption arent necessarily translating into deployment or business impact.While interest in(and spending on)AI solutions remains high,businesses are beginning to pay more attention to the cost of AI projects,and stepping

18、 up efforts to ensure they deliver value.The coming into force of the European Unions AI Act,which sets an international benchmark by laying out obligations around data governance,documentation,human oversight and security for businesses adopting AI systems.Sustained,massive investment in data cente

19、rs,with Google even turning to nuclear power to generate the vast amounts of power its AI offerings are likely to require.This indicates AI is a long-term bet that will continue to gain momentum in the business context,and in society as a whole.The growth of tools simplifying how engineers and other

20、s interface with AI models,such as LiteLLM and Langchain.Renewed focus on tackling LLM hallucinations and fabrications,with novel techniques like semantic entropy being applied to root out errors,and LLMs policing the output of other LLMs.Rising awareness of shadow AI,or the use of unsanctioned AI t

21、ools in the enterprise context,which could pose significant problems for companies if sensitive information is leaked to LLMs by employees.In one recent survey a third of organizations admitted to finding it hard to monitor the illicit use of AI among their teams.Looking Glass Thoughtworks,Inc.All R

22、ights Reserved.6Operationalizing AI for business impactTrends to watchSeeing nowBeginning to seeOn the horizonAdoptAnalyzeAnticipateStrategic recommendation138474948404243444546413926101526273334353628293031322122232425201914954711171281813163375150Seeing nowAdopt1.Accessibility in multimodal experi

23、ences2.Agent-based simulation3.AI agents4.AI as a service5.AI in security6.AI-assisted software development7.Automated compliance8.Collaboration ecosystems9.Data mesh10.Edge computing11.Ethical frameworks12.Evaluating and managing AI outputs13.Evolutionary architectures14.Explainable AI15.Generative

24、 AI16.Integrated data and AI platforms17.Interfacing with AI18.LLMOps19.MLOps20.Model training optimization21.Online machine learning22.Platforms as products23.Privacy first24.Software-defined vehicles25.Vector databasesAnalyze26.AI marketplaces 27.AI safety and regulation28.AI-generated media29.Aut

25、omated workforce30.Autonomous robots31.Changing perceptions of AI32.Easing access to generative AI33.Federated learning34.Multimodal AI35.Personalized healthcare36.Synthetic dataAnticipate37.Understandable consentBeginning to seeAdopt38.AI-ready data39.Fine grained data access controlsAnalyze40.AI O

26、bservability41.Data lineage42.GenAI computer control43.Intelligent machine to machine collaboration44.Production immune systems45.Small language models46.Talk to dataAnticipate47.Adversarial machine learning48.Affective(emotional)computing49.AI in roboticsOn the horizonAdoptAnalyze50.AI avatarsAntic

27、ipate51.AGI research Thoughtworks,Inc.All Rights Reserved.7The opportunitiesBy getting ahead of the curve on this lens,organizations can:Enhance knowledge management and transfer by adopting GenAI to help employees sift through,summarize and analyze stores of enterprise data,whether structured or un

28、structured.A wide range of products are emerging to facilitate the retrieval and dissemination of important information in industries like property.Harness AI to accelerate processes like legacy modernization and coding.Thoughtworks is already successfully applying GenAI to assist teams with one of

29、the most difficult aspects of modernization:understanding and unpacking the intricate web of connections that typically underpin legacy systems and codebases.AI assistants can also significantly boost the productivity of software development and other teams by taking over frequent,repetitive tasks.E

30、xplore AI agents to elevate automation,potentially transforming how employees perform tasks like scheduling and customer support,and raising the bar for engagement and personalization in customer interactions.Boost the speed at which LLMs are brought into production,and their effectiveness when depl

31、oyed through emerging practices and tools like LLMOps,which accelerate model development;retrieval-augmented generation(RAG),which can enhance models reliability;and AI gateways or smart endpoints to connect AI systems to applications.Develop and communicate a joined-up AI strategy that empowers emp

32、loyees to experiment with AI in a structured way,while preventing the emergence of shadow AI that could pose a threat to the organizations intellectual property or reputation.Leverage small language models to bring AI innovations to edge devices,offering opportunities for everything from operational

33、 analytics to personalization without compromising privacy,since data doesnt have to be moved to the center of a network.Lead the way in terms of compliance and ethical AI practices.We urge our clients not just to follow but embrace regulations like the EU AI Act,as such legislation often reflects w

34、ider societal sentiment and concerns and potential customers take notice of businesses that are responding.Operationalizing AI for business impactLooking Glass Thoughtworks,Inc.All Rights Reserved.8Operationalizing AI for business impactWhat weve donePEXAThoughtworks partnered with digital property

35、technology company PEXA,AWS and Redactive to develop an innovative and versatile AI assistant that has boosted the productivity of PEXAs employees by providing personalized answers to queries and augmenting tasks like information discovery.Seamlessly integrated with PEXAs internal systems,the soluti

36、on also met robust requirements for data security and privacy by equipping the assistant with permissions awareness,ensuring employees are only able to access information cleared for sharing.Looking Glass Thoughtworks,Inc.All Rights Reserved.9Operationalizing AI for business impactActionable adviceT

37、hings to do(Adopt)Identify AI champions who can help guide and teach your organization about the potential use cases for emerging solutions but understand that AI can and will be applied in different ways in almost every part of the enterprise,which means these champions need to keep an open mind.Ha

38、ving people with a clear idea of what good looks like can reduce risks and ensure AI initiatives focus on meaningful business results.Implement a holistic and comprehensive AI strategy for your organization that includes guidelines on permitted tools and the contexts in which AI can be used,to minim

39、ize the risks of shadow AI.Adopt retrieval-augmented generation(RAG)when developing AI systems,to give reliability an uplift and position models to create more specific outputs.Integrating evals and observability can further enhance the resilience of systems over the long term.Embed AI throughout th

40、e software development lifecycle.Maximum results are achieved when the role of AI isnt just limited to coding,but assists with processes like testing and documentation.Apply data mesh and data product thinking to ensure AI applications are built on the robust data foundation needed to ensure they de

41、liver business or customer value.Disciplines like data curation,which creates,organizes and manages data sets so theyre transparent and easily accessible,also contribute to the success of AI.Use proxies to simplify the way teams interact and leverage AI models,paving the way for the enhancement of a

42、pplications they develop with AI features and capabilities.Looking Glass Thoughtworks,Inc.All Rights Reserved.10Operationalizing AI for business impactThings to consider(Analyze)Avoid whats known as the substitution myth the idea that AI can simply directly replace a human.Instead,build and implemen

43、t systems that augment roles to make teams more productive and engaged,while acknowledging the continued importance of human judgement and oversight.Be cognizant of varied expectations around AI.Research suggests people may approach AI differently depending on cultural background,with some wanting a

44、 high degree of control and others prioritizing a sense of connection.These differences,as well as variances in context or situation,need to be understood and acknowledged when planning and implementing AI.Pay close attention to costs,and try to identify the approaches most likely to meet your needs

45、 while generating return on investment.Running AI models can be expensive,especially if expenses like employee compensation are factored in.Keeping spending in check requires active financial monitoring(i.e.FinOps)and consideration of things like small language models.Monitor AI regulation and futur

46、e policy developments,particularly how these intersect with privacy laws,which could have a massive impact on the data resources available for AI projects.Multiple US states,and countries from Canada to India and Japan,are planning to enhance or roll out legislation that will set guardrails around A

47、I use and development.Things to watch for(Anticipate)Questions around legal liability and accountability for the negative consequences of AI use.As issues such as AI misleading customers and the associated legal challenges emerge,authorities like the EU are moving to make organizations more culpable

48、.The potential growth of AI companions,designed to provide emotional support,friendship or even intimacy.While these could help combat loneliness and isolation,they may also have troubling implications for human interaction,requiring businesses to think carefully about the introduction of AI with co

49、mpanion-like features.Looking Glass Thoughtworks,Inc.All Rights Reserved.11Strengthening the data value chainLeveraging data platforms and AIAs enterprise adoption of AI gains pace,theres rising awareness of datas role as a differentiator,and a source of competitive edge.Developing the capabilities

50、to leverage data at speed and scale,and become truly data-driven,has become an emerging priority.Treating data as a product represents one of the most effective means to achieve this goal,and the best way to build and distribute data products is through data platforms.The principles that underpin hi

51、gh-performance data platforms remain the same decentralization and federated data ownership but new trends and opportunities in the space are presenting challenges that organizations need to be prepared for.In particular,the rise of generative AI(GenAI),and the importance of unstructured data in it,

52、requires teams to think differently about how data is managed and processed.Its becoming critical to treat unstructured data as a first class citizen,not as structured datas poorer cousin.Its also important to note the rising need for better and ideally automated governance of data products.Data pro

53、ducts reusable data assets engineered to deliver trusted datasets for specific purposes exist in dynamic environments where the needs of teams and the wider organization are constantly evolving,and its important that they also develop in a way that delivers value.Maintaining the capacity for competi

54、tive and sustainable change requires intentional design of cohesive centralized and decentralized capabilities.Some organizations are navigating away from creating consensus-based single sources of truth to forming integrated contextual truths.Looking Glass Thoughtworks,Inc.All Rights Reserved.12Equ

55、ally essential is ensuring data products are built with a clear line to business adoption.Platform and product thinking can help,but theres a need to move beyond existing paradigms and tooling,and consider applying human-centered design for more effective ways for data to be consumed and leveraged b

56、y business users.GenAI and trends like talk to data and graph-based discovery are creating promising opportunities in this space,transforming the way teams interact with and consume data.Strengthening the data value chain“An open and evolving data and AI platform allows organizations to embrace unce

57、rtainty in rhythm with changing demands,fostering a culture of continuous learning.”Nimisha AsthagiriTechnical Principal and Data Mesh Leader,ThoughtworksSignals Unstructured data moving from a supporting to a starring role.Theres growing focus on the use of unstructured data(such as text,video,imag

58、es and audio)to build better AI training models,which requires integrating and working across different types of data in as frictionless a way as possible.Startups in this space are gaining significant investment and the likes of IBM are unveiling new products specifically designed to help enterpris

59、es unleash the potential of unstructured data in analytics and AI.Enterprises applying GenAI to better leverage unstructured data.GenAIs ability to parse and summarize vast quantities of the information contained in everything from meeting recordings to PowerPoint presentations,and to support natura

60、l language interactions,is transforming the way teams access and use data and enhancing knowledge management.However,this trend is also raising questions as to whether AI and GenAI platforms should be integrated with other data platforms or kept distinct,which,in some cases,is leading to platform pr

61、oliferation.More organizations grappling with the challenges of treating data as a product,as it becomes a business imperative.Research shows the vast majority of businesses see clear benefits from such an approach,including improved data sharing and strengthening the connection between data and bus

62、iness goals.However,they are confronting multiple barriers along the way,from fragmented systems to uncertainty about data provenance.The rising importance of data discoverability.By empowering users to better discover,understand and use data assets,data catalogs can play an important role in data p

63、latforms and a data product approach.But they can also cause more issues than they solve if their user experiences or capabilities are limited,impeding the discovery process.The recent introduction of knowledge graphs to data platforms is addressing these risks,making it possible to draw out relatio

64、nships and nuances in data that are typically lost in the process of abstraction.More pressure being put on data teams to demonstrate ROI and manage costs more effectively.The increasingly established link between data strategy and enterprise performance also means these teams can no longer work in

65、isolation;instead strategies should be co-developed with,and create platforms that deliver results for,the business.Looking Glass Thoughtworks,Inc.All Rights Reserved.13Strengthening the data value chainTrends to watchSeeing nowBeginning to seeOn the horizonAdoptAnalyzeAnticipateStrategic recommenda

66、tion134423536372610152132263133272829302223242520191495471117128181316339404138Seeing nowAdopt1.AI as a service2.Automated compliance3.Collaboration ecosystems4.Data catalog5.Data fitness functions6.Data mesh7.Data product specification8.Developer experience platforms9.Digital twin10.Edge computing1

67、1.Ethical frameworks12.Explainable AI13.FinOps14.Green computing15.Integrated data and AI platforms16.Knowledge graphs17.MLOps18.Model training optimization19.Online machine learning20.Platforms as products21.Privacy first22.Privacy-enhancing technologies(PETs)23.Secure software delivery24.Smart sys

68、tems and ecosystems25.Vector databasesAnalyze26.Autonomous robots27.Autonomous vehicles28.Data clean room29.Data marketplaces30.Decentralized data architectures31.Federated learning32.Semantic representational technologies33.Synthetic dataAnticipate34.Understandable consentBeginning to seeAdopt35.AI

69、-ready data36.Data contract37.Fine grained data access controlsAnalyze38.Data lineage39.Integrating unstructured data40.Intelligent machine to machine collaboration41.Talk to dataAnticipate42.Decentralized personal data storesOn the horizonAdoptAnalyzeAnticipate Thoughtworks,Inc.All Rights Reserved.

70、14The opportunitiesBy getting ahead of the curve on this lens,organizations can:Consolidate data and AI platform capabilities,enabling AI as a service to embed this new technology and empower users to leverage it successfully throughout the organization.Surveys have shown that despite concerns about

71、 the wider impacts of AI,adoption has positive implications for teams collaboration,efficiency and performance.Use AI(and GenAI)to build and maintain data products more effectively.Emerging AI tools have the potential to contribute to data products in a number of ways,from synthesizing and analyzing

72、 information garnered in end-user research or testing,to accelerating coding and creating documentation that can smooth the path to effective adoption.Enhance control over costs.With data management often dominating enterprise technology spending,introducing new tooling to track data lineage and ana

73、lyze the impact of complex data initiatives can help teams determine and demonstrate ROI with greater precision.FinOps thinking can contribute significantly to this process by strengthening the links between tech and business teams and ensuring investments come with financial accountability.Strength

74、en data governance by introducing emerging best practices and structures.These include data clean rooms,secure,self-contained environments where enterprises can blend proprietary and third-party data to improve analytics and personalization while protecting customer privacy;and data contracts,which

75、by setting ground rules for data users and consumers,can improve transparency and trust when sharing data across an organization.Combine knowledge graphs and GenAI,which can enhance understanding of large,complex data sets by mapping the relationships among entities within them.This opens the possib

76、ility of more semantic approaches to integration,which in turn can help create a better user experience for data consumers.In addition,combining knowledge graphs and GenAI can also deliver better LLM responses because were taking explicit knowledge from knowledge graphs and combining it with implici

77、t statistical knowledge from LLMs.Strengthening the data value chainLooking Glass Thoughtworks,Inc.All Rights Reserved.15Strengthening the data value chainWhat weve donePfizerThoughtworks is working actively with these leading pharmaceutical companies to create data mesh platforms that enhance their

78、 ability to create and deliver transformative data products.With Pfizer,we helped develop cutting-edge layered platforms serving AI-powered data products,graph-based semantic interoperability,and LLM-based agents that drive the firms oncology research,supporting early drug discovery.GileadFor Gilead

79、,we supported the design and implementation of Gilead DnA,a scalable enterprise-wide data platform that provides data engineers and researchers with a secure self-service environment for data processing,complete with talk to data functionality.Looking Glass Thoughtworks,Inc.All Rights Reserved.16Str

80、engthening the data value chainActionable adviceThings to do(Adopt)Lay the right foundations for creating effective data products by implementing a data mesh,which places data within the reach of teams that need it most and reduces friction between data producers and consumers.Automate data governan

81、ce as much as possible to ensure policies are implemented consistently and with minimal impact on data usage and consumer experience.Fitness functions and more rigorous monitoring of service level indicators(SLIs)can be good places to start.Start treating unstructured data as a first class citizen t

82、hat is given the same attention and prominence as structured data in your data platform,and draw on its potential to improve analytics and AI models.Invest in a superior data product development experience to accelerate adoption.Mapping decision journeys can help the organization better understand a

83、nd trace how to move from use cases to data,and particularly AI data,products.Looking Glass Thoughtworks,Inc.All Rights Reserved.17Things to consider(Analyze)Extend user experience and human-centered design to data and AI.This includes thinking carefully about how to build the best possible interfac

84、e and experience for discovering and accessing data,out of an expanding range of GenAI-enabled options.Examine ways to track and document data lineage and improve metadata for data products for data consumers.Doing so can also enhance data governance and data engineering by highlighting opportunitie

85、s to smooth the flow of data throughout the organization.AI tools can play a valuable role in this process by providing a quick and precise snapshot of datas history and transformations.Adopt mechanisms to minimize the risk of creeping centralization.Encourage teams to think less about creating a si

86、ngle source of truth and more about adopting federated data management that efficiently delivers what the use case or context demands.Track ROI for data and AI transformations.Its important to be able to demonstrate the value and impact being driven by data and AI initiatives.Theres no single way of

87、 doing this,but its a valuable step in ensuring teams remain value-focused and that projects in this area have organizational buy-in.Things to watch for(Anticipate)Strengthening the data value chain Next-gen user experiences like voice and VR impacting data discovery.By allowing users to query data

88、naturally and moving data visualization into a three-dimensional space,new tools promise to transform the way teams perceive,interact with and understand information,paving the way for deeper analysis and collaboration.Propagating more granular access controls as data platforms and products scale to

89、 more users and data product development accelerates.Studies show data professionals are already walking a fine line between prioritizing security and not impeding the efficiency and flexibility data platforms are designed to provide.Adopting GenAI and knowledge graphs to improve data discovery and

90、better describe and document entities in large data sets.Looking Glass Thoughtworks,Inc.All Rights Reserved.18Reimagining responsible tech for the era of generative AIThe importance of responsible tech has been widely discussed for a number of years now.However,across the industry it has struggled t

91、o properly cut through,remaining a somewhat marginal concern.While this might not be unsurprising during a period of economic uncertainty and tighter budgets,the rise of generative AI has made the topic more urgent than ever.This is because the ethical,legal and even philosophical questions raised b

92、y the technology can,at least in part,be addressed through responsible technology principles and practices.This means 2025 is the year businesses need to properly embrace responsible technology.Without it,attempts to experiment and innovate with generative AI and associated technologies may contain

93、risks that businesses simply do not need during a challenging period,ranging from the financial consequences of compliance failures to damaged consumer trust.But what does embracing responsible tech in the generative AI era actually involve?We see it as beginning with a recognition that has long bee

94、n absent that responsible tech isnt something you bolt on to existing activities and projects:its something that needs to be embedded in organizational values,team practices and cultures.Looking Glass Thoughtworks,Inc.All Rights Reserved.19Reimagining responsible tech for the era of generative AIIt

95、means considering all the potentially negative consequences of,for instance,a new generative AI chatbot,whether in terms of data leaks and privacy breaches,all the way to disturbing and harmful content being served to end users.It also means anticipating regulatory and legislative demands rather tha

96、n simply reacting to new legislation;given legislation is often driven by the concerns and interests of the wider public ie.your consumers its invariably a useful heuristic for building trust with those that matter most to your business.You may need to fight for responsible tech on two fronts:in how

97、 these technologies are built and deployed;and in how these technologies are shaping what we do.Generative AI capabilities have been added to all manner of products,which may catch some consumers unaware for instance,they might not realize the service representative theyre chatting to isnt human.Thi

98、s means part of responsible tech isnt just about being decisive its also about being sensitive to the unknowns that are inherent in an environment where generative AI is everywhere.This mindset should be extended across the breadth of what your organization does,regardless of whether AI is a priorit

99、y for your business or completely outside your operational scope.“Responsible technology isnt a checklist its a mindset.In the age of generative AI,embedding ethics at the core of innovation isnt just about avoiding risks;its about building trust,anticipating change and leading with purpose in a fas

100、t-changing world.”Ken MugragePrincipal Technologist,ThoughtworksSignals Questions about accountability and legal liability for harmful technology consequences.The lawsuit of a mother of a teenager who died after interacting with an AI chatbot could be a significant moment in AI legislation and shape

101、 how we think about liability and responsibility.Increasing awareness or perhaps chaos around what data has and has not been used to train major AI models.For instance,there was a remarkable moment when an OpenAI leader didnt know whether Sora was trained on YouTube videos,while abusive material was

102、 found in the LAION 5-b dataset.There was also confusion when LinkedIn suspended data processing in the UK in September 2024 following concern from the Information Commissioners Office about the way UK users data would be used to train generative AI.Water consumption of data centers is causing signi

103、ficant concern.At a local level in areas affected by drought there is even greater political friction,highlighting the ongoing environmental questions raised by AI usage.Looking Glass Thoughtworks,Inc.All Rights Reserved.20Reimagining responsible tech for the era of generative AI The consequences of

104、 corporate greenwashing are becoming tangible as organizations are held to account for false claims about their green credentials.The interdependence of the industry underlining the importance of trust and transparency at a technical level demonstrated by the CrowdStrike outage and other supply chai

105、n vulnerabilities like the XZ Utils backdoor.Increasing shadow AI in organizations.As the generative AI market has grown,it is incredibly easy for employees to experiment with AI without oversight.This can create significant privacy risks.The fragmentation of the social media landscape.The lack of s

106、tability in this space best demonstrated in the mass exodus of X(formerly Twitter)users underlines significant content and safety problems with platforms as well as growing consumer fears about online safety,privacy,mis/disinformation and even their digital consumption habits more broadly.The growth

107、 of impact investing.This is where investors target big social problems like health or the environment with a view to capturing value.It has been called out as something that rather than tackling social issues can actually exacerbate them.Legislation taking on manipulative design.The EU,for instance

108、,has introduced a law aimed at tackling dark patterns.Thoughtworks,Inc.All Rights Reserved.21Reimagining responsible tech for the era of generative AITrends to watchSeeing nowBeginning to seeOn the horizonAdoptAnalyzeAnticipateStrategic recommendation3931322216232124171819202512610151495471112813326

109、27283435363733293038Seeing nowAdopt1.Accessibility in multimodal experiences2.AI in security3.AI-assisted software development4.Automated compliance5.Biometric authentication6.Decentralized security7.DevSecOps8.Digital carbon management9.Edge computing10.Ethical frameworks11.Green computing12.Model

110、training optimization13.Privacy first14.Privacy-enhancing technologies(PETs)15.Secure software deliveryAnalyze16.AI safety and regulation17.AI-generated media18.Alternative currencies19.Automated workforce20.Changing perceptions of AI21.Data clean room22.Data marketplaces23.Impact funds24.Internet r

111、egulation25.Synthetic dataAnticipate26.Addictive tech27.Next-generation cryptography28.Understandable consentBeginning to seeAdopt29.Data contract30.Fine grained data access controlsAnalyze31.Data lineage32.Production immune systemsAnticipate33.Adversarial machine learning34.Affective(emotional)comp

112、uting35.Decentralized personal data stores36.Quantum computing37.Responsible tech facilitationOn the horizonAdoptAnalyze38.AI avatarsAnticipate39.AGI researchLooking Glass Thoughtworks,Inc.All Rights Reserved.22The opportunitiesBy getting ahead of the curve on this lens,organizations can:Strengthen

113、organizational culture by working to ensure there is genuine and authentic alignment between corporate rhetoric and values and the perspectives and values of employees.This is certainly not easy at a commercially challenging time,but mistrust and cynicism will have long-term consequences that may pr

114、ove even more difficult to repair.Revisit and strengthen those values.Values need to be maintained and evolved if left without stewardship they will prove useless.Organizations,and leaders in particular,should spend time considering existing values and whether theyre relevant and,most importantly,ac

115、tionable.They need to be things that can guide behaviors and decision-making at every level;they should be things that can be put into practice.Leverage AI thoughtfully.AI can offer a competitive advantage:but simply rushing to integrate it is risky.The real opportunity is to be thoughtful about why

116、,where and how AI capabilities are used.This will not only minimize potential risks,it will also strengthen your relationship with customers and ensure you are delivering even more value.Build trust with consumers.Concerns about privacy and the way data is used and managed continue to grow.Businesse

117、s that seek to buck the perceived trend for ever-increasing extraction may be able to gain an advantage in the market.Transparency and trust can be a differentiator.Be more intentional and considered about what youre doing to meet user needs.Foregrounding accessibility practices in your organization

118、 can help ensure you are building products,services and experiences that provide value for even more people.Reduce waste and improve efficiency.While businesses should,of course,be focusing on their environmental impact,in challenging times the bottom line takes absolute priority above everything el

119、se.However,its possible to do both:in fact,framing environmental action in terms of efficiency can be an effective way to ensure it is taken seriously across the business.In other words,business leaders need to present responsibility and commercial impact as things that are closely intertwined,not m

120、utually exclusive.Focus on skill development.Avoid the temptation to automate everything and rely on AI tools to do more with less.What happens when you dont have knowledge or skills needed to solve problems down the line?Considering how human skills and AI capabilities can complement one another wi

121、ll ensure you have a team that is able to help the organization reach its future objectives.Reimagining responsible tech for the era of generative AILooking Glass Thoughtworks,Inc.All Rights Reserved.23Reimagining responsible tech for the era of generative AIWhat weve doneSwann SecuritySwann Securit

122、y partnered with Thoughtworks to develop the worlds first AI security concierge,a groundbreaking solution designed to enhance home protection while safeguarding privacy.Leveraging generative AI,the system engages visitors naturally,whether homeowners are present or not.Prioritizing privacy,Thoughtwo

123、rks crafted a prompt engineering strategy to ensure the AIs responses respect security boundaries and defend against intrusive or adversarial interactions.The AI concierge can manage deliveries,greet guests and deflect inappropriate requests,all while maintaining a courteous and secure demeanor.Rigo

124、rous testing ensured the systems resilience against privacy threats and unanticipated scenarios.This customizable framework allows Swann to tailor settings to household-specific needs,enhancing user control.Showcased at CES and named a Smart Home Honoree,this innovation sets the stage for Swanns fut

125、ure AI-powered products,demonstrating that cutting-edge security solutions can protect both homes and privacy in an increasingly connected world.Looking Glass Thoughtworks,Inc.All Rights Reserved.24Reimagining responsible tech for the era of generative AIActionable adviceThings to do(Adopt)Use the R

126、esponsible Tech Playbook for practical activities software delivery teams can actually do on projects.Implement holistic and consistent policies around AI use.Shadow AI without oversight can lead to a diverse range of issues,risking everything from privacy breaches to reputational problems.Leverage

127、new techniques to make generative AI more reliable.This includes things like evals(a form of testing whereby outputs are assessed according to the context in which they will be received)and guardrails,a programmed set of policies that permits and prevents certain kinds of outputs.These need to be un

128、derpinned by your values and to do that you need to be able to define and articulate them.Ensure your technology strategy is collaborative,not top down.Although senior decision makers have an important role to play in setting a vision and an agenda for how an organization will use technology,involvi

129、ng other parties in that process not only builds organizational trust but also helps increase confidence in the decisions that are made.One effective way of doing this is to create a technology radar like the one we create twice a year at Thoughtworks.It allows people to question,voice concerns and

130、propose alternatives in a way that is safe and supportive.Treat responsible tech practices as a capability and skill issue.Identify relevant training opportunities and resources for technology teams and other parts of the organization and ensure that teams see this as a valued area in which to devel

131、op new skills.Be intentional about the social platforms youre using.Are they spaces in which you want to play?Do you want to be associated with the type of content that is shared there?Looking Glass Thoughtworks,Inc.All Rights Reserved.25Reimagining responsible tech for the era of generative AIActio

132、nable adviceThings to consider(Analyze)Dedicate time to revisiting and reconsidering your values.Are they meaningful?Can they be practiced?Do they actually guide action?Consider whether your organization is putting its professed values into action.If its not,why not?What are the business risks holdi

133、ng you back?Think about how you might measure responsibility.This could include anything from environmental measures through to employee perception and morale.Employee attitudes to your values.Do people feel they are being embraced?Are they something they themselves can enact?Legislation.Monitor new

134、 regulations and analyze how they may impact your organization.Keeping a close eye on broader conversations about future changes to law can also help you be prepared for the future and avoid being reactive.Ownership and accountability inside the organization.Many issues around responsibility quite l

135、iterally require oversight from a responsible person.Expecting it will take care of itself is likely to be ineffective.Think seriously about who should be responsible or accountable and who should own or measure your performance in this area.Pay attention to your software supply chain.Do you underst

136、and whats in your stack?Consider using a software bill of materials(SBOM)to track dependencies and provide technological transparency.Things to watch for(Anticipate)How the law evolves on legal culpability for negative impacts of tech.This is a question largely pertinent to AI,but it is also importa

137、nt to monitor legislation around content,data privacy and accessibility.Consumer attitudes to AI.AI is currently extremely hyped.However,this doesnt mean that consumer sentiment will follow the industry in the medium or long-term.Poor products,negative and dangerous effects and even pure fatigue may

138、 cause many people to see AI as either problematic or overhyped.The future of ESG investing.Just because ESG falls out of favor does not mean responsibility whether thats social,environmental or otherwise no longer matters.However,it may make it harder to make the case for it.Leaders need to do the

139、right thing and ensure that there is a commitment to behaving with integrity and transparency this will strengthen consumer confidence in your brand in the long-term.Looking Glass Thoughtworks,Inc.All Rights Reserved.26Enabling richer experiences through multimodal interactions Modes of interaction

140、between people and machines have proliferated rapidly in recent years,to encompass text,voice,images,video,gestures and with affective computing,even emotional cues that may be unconscious.This presents new possibilities,but also a challenge in requiring organizations to think through how different

141、forms of interaction can be combined and used across the entire customer experience.Doing that can raise multiple questions.When is gesture or image recognition more helpful than a standard text interaction?When are physical buttons important,and in what situations could they constitute an obstacle?

142、Carefully considering context and cognitive load can help ensure a positive experience that does not jar,irritate or unintentionally make an interaction more difficult.Innovation should never degrade a product or service.While enterprises focus on enhancing their customers online experiences they ca

143、n gather insights that are applicable in other areas of the customer experience.For instance,fashion retailers are increasingly using virtual fitting rooms to give shoppers a better sense of how their clothes will look when worn by capturing their mix and match choices online,in-store staff can read

144、ily locate those items when the consumers come into the shop.Looking Glass Thoughtworks,Inc.All Rights Reserved.27Enabling richer experiences through multimodal interactions AI is playing a significant and expanding role not only in how we interact with services and products most often through chatb

145、ots but also how we model interactions.Agentic AI,for instance,can simulate human behavior to help sales teams identify prospects,or anticipate the customer response to a product or campaign.Businesses should also remember not every new interaction is embraced by the customer.The backlash provoked b

146、y trends like gamification and addictive tech mean organizations need to be more considered in who theyre building for,and why.This involves paying close attention to the friction in the customer experience and the accessibility implications of different channels for interaction,but also rethinking

147、the foundations of what it means to interact with a product or brand.The best outcomes stem from avoiding the temptation to prioritize only optimization or stickiness,and considering more holistically how innovation in interaction can add value,or expand the effective range of a service.“Human compu

148、ter interaction is becoming more diverse than ever,which means organizations in a huge range of industries have the opportunity to create exceptionally rich experiences for everyone from their customers to their own employees.However,what is critical is ensuring experiences are always context specif

149、ic,so thinking about what mode is most relevant when is vital.”Chloe BlanchardPrincipal Design Researcher,ThoughtworksSignals AI-driven chatbots becoming increasingly commonplace in domains such as customer service.Research suggests the global chatbot market,valued at just over US$5 Billion in 2023,

150、will surge to over US$36 Billion by 2032.Chatbots also continue to evolve,with techniques like retrieval-augmented generation(RAG)improving the accuracy and reliability of interactions.Steady growth in the voice interaction space,with more companies adopting voice assistants,and voice searches and r

151、equests accounting for a growing proportion of consumer activity.Legislation taking aim at dark patterns,aspects of the user experience that are subtly designed to nudge users into performing certain actions that are typically not to their benefit.Article 25 of the European Unions Digital Services A

152、ct is one example.Similarly,accelerating and intensified gamification,as seen in the prevalence of controversial loot boxes,is coming under more government scrutiny.Consumers growing more conscious of screen time and actively seeking to avoid addictive tech.Theres been a noticeable rise in apps to h

153、elp users manage how they interact with their devices,or cut down interactions altogether,with Google recently rolling out screentime reminders as part of the Digital Wellbeing suite for the Android mobile operating system.Australia banning under 16s from social media is one high-profile example of

154、the rising fear around excessive screen exposure.Looking Glass Thoughtworks,Inc.All Rights Reserved.28Enabling richer experiences through multimodal interactions Agentic AI is being presented as the next frontier of generative AI.Able to handle more complex tasks,accept directions in natural languag

155、e and work with other software tools,agents could change how we collaborate,complete tasks and interact with other technologies.Continued innovation in AR and VR.Despite the lukewarm response to some of the high-profile headsets that have launched,we still see companies making big bets in this space

156、.AI integration is coming to Ray-Ban Meta smart glasses and Meta has also unveiled plans to release a neural wristband in the future,notably as part of its Orion project.Even if consumers have yet to embrace this technology fully,theres no shortage of companies seeking to find a killer app.A neurali

157、nk chip being inserted into a person for the first time,marking a major step forward in the development of direct brain-computer interfaces.The chip was also recently greenlit for clinical trials outside the US,as a potential means to enable paralyzed individuals to use the mind to engage with digit

158、al devices.Consumer appetite for AR/VR retail experiences,which iconic brands like Louis Vuitton have leveraged to engage customers and enhance their brands.One recent survey showed nearly a third of US consumers are keen for VR to recreate brick-and-mortar shopping routines.Thoughtworks,Inc.All Rig

159、hts Reserved.29Enabling richer experiences through multimodal interactions Trends to watchSeeing nowBeginning to seeOn the horizonAdoptAnalyzeAnticipateStrategic recommendation1262728265438971011181917202122231213141516242529303132Seeing nowAdopt1.Accessibility in multimodal experiences2.AI agents3.

160、AI-assisted software development4.Biometric authentication5.Developer experience platforms6.Digital twin7.Green computing8.Model training optimization9.Privacy firstAnalyze10.Augmented reality11.Automated workforce12.Autonomous robots13.Autonomous vehicles14.Changing perceptions of AI15.Consumer XR1

161、6.Internet regulation17.Mindful screen interaction18.Multimodal interactions19.Personalized healthcare20.Satellite networks21.Semantic representational technologies22.Tactile interaction23.Touchless interactionsAnticipate24.Addictive tech25.Understandable consentBeginning to seeAdopt26.Industrial XR

162、Analyze27.GenAI computer control28.Talk to dataAnticipate29.Affective(emotional)computing30.Brain-computer interfaces31.Next-generation wearablesOn the horizonAdoptAnalyzeAnticipate32.AGI researchLooking Glass Thoughtworks,Inc.All Rights Reserved.30Enabling richer experiences through multimodal inte

163、ractions The opportunitiesBy getting ahead of the curve on this lens,organizations can:Explore multimodal forms of interaction to create customer engagement breakthroughs.Choices on which channels of interaction to adopt or offer need to be guided by context,situation,geography and user preferences.

164、But examples like Canons World Unseen exhibition,which featured images accessible to people with visual impairments,show how new technologies can expand what it means to interact with a product,service,or company,and deepen relationships with users or consumers.Expand their potential market by bring

165、ing accessibility to bear.Innovations like Signapses automated sign language translation,and AI models that allow visually impaired people to interact with applications through natural language,show the potential for companies to open access to their brand or services for new customer groups.Leverag

166、e AI agents to offer new conveniences to consumers.Though agentic AI remains a nascent category,the technology is developing quickly,and can mimic human responses and behavior with uncanny levels of accuracy.This opens the possibility of automating more complex tasks and processes to help consumers

167、accomplish their goals,whether by retrieving information,providing tailored advice or offering multilingual support.Adopt extended reality(XR)to improve training and onboarding.Organizations such as Bostons Mass General Brigham hospital are applying AR,VR and mixed reality to replicate real-life env

168、ironments and situations,helping employees grow familiar with their roles and master new skills in a more immersive way.XR can be particularly valuable for training workers in high-risk industries,by exposing them to emergencies without any real-world danger.Reimagine how teams,and customers,interac

169、t with data.By applying AR and VR,organizations can move data visualization into new dimensions,providing users richer,more in-depth experiences of information,lightening the cognitive load and even potentially accelerating time to insights.Looking Glass Thoughtworks,Inc.All Rights Reserved.31Enabli

170、ng richer experiences through multimodal interactions What weve doneAbraham Lincoln Presidential Library and MuseumThoughtworks partnered with Google Cloud to transform the visitor experience at this important institution focusing on the life and legacy of Abraham Lincoln.AI,XR and VR technologies a

171、re being applied to make exhibits more immersive,by providing photorealistic experiences and access to additional layers of information.The project is also developing assisted visual guides to enhance accessibility for people with disabilities,and multilingual content for non-English speaking visito

172、rs.Looking Glass Thoughtworks,Inc.All Rights Reserved.32Enabling richer experiences through multimodal interactions Actionable adviceThings to do(Adopt)Develop interactions with accessibility in mind,and consider how new forms of interaction can support the creation of experiences that cater to all

173、kinds of users.This may require educating and upskilling design and development teams,so theyre aware of the expanding range of technologies that can be leveraged or in some cases,rethinking how to approach the interaction.Examine the impact of voice search on the way you manage and index products a

174、nd information.As technologies like GenAI-enabled agents make more queries and interactions verbal,the way organizations identify and categorize information will need to change.Optimization of data for voice search,and natural language metadata and tagging are likely to become increasingly important

175、 features of major platforms for interaction such as websites.Prioritize quality,reliability and stability when developing new interactions,and subject them to a comprehensive testing strategy.More sophisticated interactions can throw up new challenges,and even degrade a service or brand.Chatbots ha

176、ve been known to go off script and should be subjected to regular evaluations,especially when techniques like RAG are involved.Looking Glass Thoughtworks,Inc.All Rights Reserved.33Enabling richer experiences through multimodal interactions Things to consider(Analyze)Choose the right interaction for

177、the context.Its easy to get excited about new technologies,but you should always be questioning whether theyre the best fit for a particular use case,or how theyll improve the experience of end users.Sometimes gesture recognition isnt needed,and a simple button will do.As companies like carmaker Hyu

178、ndai have found out,consumers often prefer old-fashioned means of interaction as their inherent frictions can impart a sense of confidence,even joy.Think through user mental models when using technologies like AI.When rolling out something like a chatbot,its vital to map out what people are likely t

179、o be expecting from an interaction,and what assumptions theyll bring to the table.Steps to help end-users write better prompts,or to explain outputs,can improve the overall user experience with new technologies.Dont pursue stickiness at all costs.The drive to optimize all experiences or create somet

180、hing addictive isnt just dangerous for users;its often unsustainable and can provoke backlash.Rather than contributing to dark patterns,consider the flipside:are there any opportunities to nudge users towards better,more cost effective,environmentally friendly or healthy decisions?Things to watch fo

181、r(Anticipate)Evolving consumer tech opening new interaction platforms.As products like Metas Orion exit the experimental phase and enter the mainstream,companies need to be prepared to build out new forms of interaction that cater to their expanding user bases.Tighter regulation taking aim at dark p

182、atterns and other forms of interaction perceived as manipulative.Rather than simply rules to follow,such developments should be seen as an opportunity to build better interactions that are sustainable over the long term,foster trust with user bases and create commercial advantage over optimization-o

183、bsessed competitors.Looking Glass Thoughtworks,Inc.All Rights Reserved.34Unlocking greater value from physical-digital convergenceThe convergence of the physical and the digital is advancing in both industry and consumer tech even if progress isnt always smooth,or readily evident.One of the bigger r

184、ecent developments in the space was the launch of Apples Vision Pro in early 2024.While it garnered much fanfare initially,the excitement quickly dwindled,and adoption has proven disappointing.But that shouldnt be viewed as emblematic of the trend as a whole,which continues to develop in lower-profi

185、le ways that will present opportunities and gain traction beyond the current core base of power users.One case in point is consumer health technologies,where a recent deal has valued rising star Oura,a maker of smart rings,at over US$5 Billion.The popularity of Ouras elegant body-monitoring devices

186、demonstrates theres significant potential for convergence in specific domains like personalized healthcare.In some cases,hybrid experiences are almost business-as-usual.Take examples such as video doorbells,digital menus or ride pooling services.They might not seem earth shattering,but they point to

187、 how easily consumers can adapt to the convergence of physical and virtual services and products.Looking Glass Thoughtworks,Inc.All Rights Reserved.35Unlocking greater value from physical-digital convergenceTaking advantage of these opportunities will depend on a high level of domain understanding,a

188、nd selectively implanting cross-reality innovations in ways that deliver tangible benefits rather than building products or systems without clearly defined business or consumer use cases.Enterprises will also need to be sensitive to changing consumer expectations and emerging regulations such as the

189、 EU Data Privacy Act and Texas Data Privacy and Security Act,which restrict what companies can do with the consumer data garnered by connected devices.Rather than a barrier,regulatory developments should be seen as a chance for proactive businesses to think through how they leverage more prevalent h

190、ybrid technologies safely and effectively.The EU Act may enable more players to participate,potentially growing the ecosystem.And by ensuring data is accurate,reliable and carefully managed,and conducting robust testing to ensure system reliability and resilience,enterprises can demonstrate transpar

191、ency and build trust with their customer base.Meanwhile,the physical-digital trend will get another boost as GenAI systems turn from consuming the available written data on the internet to integrating data from smart devices and wearables.Expect significant innovation in this space.“Its important th

192、at we are thoughtful about when,where and how we use technology be it a new product,feature,leveraging GenAI or introducing IoT.More and more Im seeing users and legislators holding businesses accountable especially when it comes to the collection,use and management of data,something at the very cor

193、e of physical-digital convergence.”Mackenzie DysartDelivery Principal,ThoughtworksSignals The mixed track record of consumer VR/AR tech,which despite multiple high-profile launches in 2024,from the aforementioned Vision Pro to the Meta Quest 3S,has yet to truly take off.This is due in part to doubts

194、 about their interfaces,and the purpose such devices serve.The return of tactile buttons is another sign some consumer VR/AR devices may be racing ahead of what the market wants,or is ready for.Success stories arising from domain-specific,business use cases.The expansion of Saga Robotics,which devel

195、ops multi-functional robots capable of treating the diseases that strike some crops,and BMWs promising testing of humanoid robots to reduce employees strain in its assembly lines,show how leveraging convergence in a targeted way can deliver positive outcomes.Looking Glass Thoughtworks,Inc.All Rights

196、 Reserved.36Unlocking greater value from physical-digital convergence The rise of generative AI enhancing supply chain management,by automating processes and expanding the types and quantity of data used to predict product demand and disruptions.That said,this is a relatively new field,and enterpris

197、es should monitor the extent to which such initiatives can deliver ROI.The continued emergence of high-fidelity digital twins.Their full impact may have yet to be realized,but developments like NVIDIAs powerful Earth 2 platform,which breaks new ground for climate and weather forecasting,show how dig

198、ital twins are solving problems by allowing data to be visualized and manipulated in a more immersive way.Industrial automation and IoT introducing new potential vectors for cyberattacks,as they are integrated into more essential business or production processes.New legislation around data collectio

199、n and privacy,often specifically targeting wearables.Jurisdictions like California and Colorado are leading the charge with laws that extend privacy protection to data gathered by health implants and wearable devices.More such moves are likely to follow,urging enterprises active in this space to sol

200、idify their governance strategies and adopt policies on ethical data use.Looking Glass Thoughtworks,Inc.All Rights Reserved.37Unlocking greater value from physical-digital convergenceTrends to watchSeeing nowBeginning to seeOn the horizonAdoptAnalyzeAnticipateStrategic recommendation1461112137891012

201、543161718191520Seeing nowAdopt1.Biometric authentication2.Digital carbon management3.Digital twin4.Privacy first5.Software-defined vehiclesAnalyze6.Augmented reality7.Autonomous robots8.Autonomous vehicles9.Hardware security10.Personalized healthcare11.Satellite networks12.Tactile interaction13.Touc

202、hless interactionsAnticipateBeginning to seeAdopt14.Industrial XRAnalyzeAnticipate15.Affective(emotional)computing16.AI in robotics17.Brain-computer interfaces18.Next-generation robotics19.Next-generation wearablesOn the horizonAdoptAnalyze20.AI avatarsAnticipate Thoughtworks,Inc.All Rights Reserved

203、.38Unlocking greater value from physical-digital convergenceThe opportunitiesBy getting ahead of the curve on this lens,organizations can:Improve interactions for customers and employees by thinking through the accessibility implications of hybrid technologies and embedded systems.In addition to mak

204、ing some experiences more intuitive and impactful for example,by creating training simulations that allow employees to get hands-on with processes in a risk-free way physical-digital convergence is opening new possibilities for the differently abled.Financial services is one area in particular where

205、 digital onboarding and biometric authentication are improving access to services for people who face mobility issues.Embrace emerging regulation to build trust with consumers.Research suggests consumers are likely to abandon companies they dont trust with their data,and gravitate towards those with

206、 transparent data practices.Leverage automation,embedded systems and other cross-reality technologies to improve operational resilience and efficiency,and deliver value for the business or end-users.The application of digital twins in supply chain management demonstrates how cutting-edge simulations

207、 and real-time data can predict risks and identify areas of a network or system that are ripe for optimization.Empower employees by adopting automation and smart devices,to take over repetitive or even dangerous tasks and enable teams to concentrate on more strategic or creative work.Companies like

208、Microsoft are also adopting a hybrid approach to bridge the engagement and productivity gaps between remote and on-site teams.Develop a deeper understanding of your operations and your customers.The data collected and shared by smart devices and embedded systems can unlock new levels of information

209、and insight,to guide both tactical and strategic decision making.New data sources can also enable leaps forward in the organizations use of AI and analytics provided data is harnessed and used in a secure and sensitive way.Thoughtworks,Inc.All Rights Reserved.39Unlocking greater value from physical-

210、digital convergenceWhat weve doneReeceWho wouldnt want their dream bathroom?But where to start?Research by Reece,Australias leading supplier of bathroom products,revealed that over a third of renovators struggled to visualize the end result.This insight sparked innovation.Reece partnered with Though

211、tworks to create a 3D bathroom planner,Imagin3D,empowering customers to select products,visualize their new space and confidently make decisions.Showroom consultants and trades also use the tool to finalize contracts and proposals.In the first three months post-launch,over 30,000 customers turned th

212、eir dream bathrooms into reality with Imagin3D.Looking Glass Thoughtworks,Inc.All Rights Reserved.40Unlocking greater value from physical-digital convergenceActionable adviceThings to do(Adopt)Treat data as the key enabler of physical-digital convergence.Smarter devices adopted by consumers or in th

213、e work environment can vastly increase the amount of data available to the enterprise;however,its only when this data is accurate,reliable and carefully governed that it can be used to inform decisions or credibly reconstruct real-world conditions.Embrace data engineering principles to ensure data t

214、ravels where it is needed and can be leveraged with confidence.Make domain knowledge the starting point for hybrid innovations and experiments.Rather than attempting to apply new physical-digital solutions enterprise-wide,explore how they can improve or supplement parts of a specific business proces

215、s.Applying these technologies in a limited way is more likely to yield results,and can also build the knowledge and momentum needed to support broader adoption.Embed security and resilience in systems that span the physical and digital by adhering to proven engineering practices,such as robust testi

216、ng and continuous delivery.Having team structures that encompass embedded and backend development can smooth the path to production.Looking Glass Thoughtworks,Inc.All Rights Reserved.41Unlocking greater value from physical-digital convergenceThings to consider(Analyze)Consider how existing and new l

217、egislation might impact how your business uses data,particularly data connected to individual consumers.Conduct an honest assessment of not just existing risks,but the risks that might emerge if regulations were to tighten further.Look for opportunities to improve trust and transparency.Understand t

218、hat automation,robotics and IoT are long-term infrastructure projects.Making the most of convergence in complex environments such as manufacturing facilities will often require significant investment,the retooling of multiple roles,and skills that the organization may lack.Its important to consider

219、these barriers and whether the organization has the capacity to deliver.The decisions leaders make now could have implications for the next five to 10 years.Examine opportunities to improve experience through convergence of the physical and digital not just for customers or employees facing physical

220、 or location-related barriers,but for everyone.The use of biometrics to facilitate payments,and the adoption of contactless technologies to create a safer,less strain-inducing environment for workers,are good examples of convergence that can benefit a wide range of end users.Things to watch for(Anti

221、cipate)Tighter legislation around data privacy,and a hardening of consumer attitudes,which may limit what the organization can do with data or its ability to pursue convergence-related opportunities.This is especially when it comes to sensitive areas like health,with research from Deloitte showing c

222、onsumers are increasingly concerned about privacy breaches or excessive tracking linked to their wearable devices.The evolution of the AR/VR market.While these technologies remain outside the mainstream,it may be only a matter of time before a device emerges that kicks adoption into high gear,and cr

223、eates viable commercial applications or use cases in the process.GlossaryAAccessibility in multimodal experiences:The expansion of interactions beyond traditional interfaces to include XR,voice,image and gesture recognition,among others,brings new challenges in accessibility.Ensuring inclusivity in

224、these contexts requires innovative design and testing to accommodate diverse user needs.Addictive tech:Some applications are specifically designed to be addictive through the use of techniques like gamification and dark patterns.This is driven by fierce competition for eyeballs and engagement and wh

225、ile there may be commercial reasons to adopt such an approach,an increasing awareness of the societal and environmental harms of addictive tech makes addictive tech a key issue organizations need to think seriously about.Adversarial machine learning:These are attacks on(or using)machine learning sys

226、tems.Attackers may tamper with training data or identify specific inputs that a model classifies poorly to deliberately create undesired outcomes.Affective(emotional)computing:A collective term for systems and devices that can recognize,interpret,process,simulate and respond to human emotions.Agent-

227、based simulation:The use of simulated independent agents,each working towards their own goals,to model a real world situation.Such simulations can help us understand complex phenomena such as the spread of diseases or protein folding.AGI research:The concept of artificial general intelligence(AGI)re

228、fers to an AI system that possesses a broad range of capabilities across a range of intellectual tasks its often compared to human-level intelligence.Debates about the threshold for AGI remain,and research into ways of achieving it continues and will play a part in wider discussions about AI and hum

229、anity.AI agents:Functionality built into applications which combines the functionality of publicly available generative AI models with specific knowledge from outside the model,such as product information.One of the most interesting manifestations of this trend are agentic assistants in which AI age

230、nts are used to accomplish particular tasks in certain domains,like HR or CRM.AI-as-a-service:“Ready-to-go”AI solutions offered as a service on cloud platforms.They often dont require specialized AI or ML skills to be used.AI-assisted software development:The use of AI to speed up or improve softwar

231、e development.Examples include code completion in IDEs,AI-created automated tests,AI that can detect bugs or even AI code generation tools.42 Thoughtworks,Inc.All Rights Reserved.Looking Glass Thoughtworks,Inc.All Rights Reserved.43GlossaryAI avatars:A digital representation of a person.The use of a

232、rtificial intelligence allows the avatar to mimic the person it represents,thus making it ostensibly more convincing and lifelike.AI-generated media:Images,audio or video that have been created or manipulated by AI.Also known as synthetic media.AI in robotics:Bringing todays AI capabilities into rob

233、otics is creating new levels of intelligence.It can help robots better respond to situations and external stimuli and ostensibly make decisions about what actions to take in relation to its environment.AI in security:AI is today often deployed both defensively,to respond to threats more dynamically,

234、and offensively,to probe for weaknesses in a system.AI marketplaces:Marketplaces such as AWS Marketplace,Google TensorFlow Hub and MS Azure Marketplace enable independent developers and companies to sell their models to a global market.They also allow consumers to quickly leverage those models to cr

235、eate value quickly.AI observability:AI systems are notoriously opaque.Their complexity can make it very difficult to determine the relationship between inputs and outputs.AI observability is the broad practice of monitoring and analyzing an AI systems behaviors and performance to increase understand

236、ing and confidence that it is working as intended.AI-ready data:AI-ready data is data that has been structured and organized in a way that makes it easy for it to be integrated with AI systems.It has a number of specific qualities:high-quality(auditable and verifiable),consistent across different pl

237、atforms and robust,comprehensive metadata.AI safety and regulation:Government regulation and guidance on the use of AI,intended to ensure responsible use and consequences of AI systems.This includes monitoring,compliance and good practice and is beginning to be extended to consumer interactions with

238、 AI.Alternative currencies:Currencies other than money,such as cryptocurrencies or reputation-based currency.Increasingly,this includes vendor-specific reward-based currencies such as Starbucks Stars or Amazon Coins.Augmented reality(AR):Where the physical world is combined with the digital.A limite

239、d form of AR is now ubiquitous,delivered via Apple and Android mobile devices,capable of overlaying virtual objects to a camera view of the world.More advanced AR is delivered via a dedicated headset such as Apple Vision Pro,Microsofts Hololens or Metas Quest 3.Automated compliance:The use of techno

240、logy to make all the data required to satisfy compliance reports,checks and balances readily available.In many cases,automation simplifies reporting by sifting through data.Increasingly,though,AI is beginning to replace manual decision-making.Automated workforce:The use of technology to perform repe

241、atable or predictable workflows.Automated workforce doesnt mean completely replacing humans;in some cases human-machine“teaming”may produce better results than either working alone.Autonomous robots:Smaller and cheaper than their industrial counterparts,robots with on-board AI are able to sense thei

242、r environment,navigate,learn to complete tasks and even fix themselves and other things.Autonomous vehicles:Self-driving cars,trucks and public transport.While the headline focus may be on self-driving cars,autonomous vehicles also have high potential for specialized industrial and business applicat

243、ions such as mining and factory floors.Looking Glass Thoughtworks,Inc.All Rights Reserved.44BBiometric authentication:A way of verifying an individuals identity that uses fingerprint,facial recognition or other similar technologies.It is today a valuable cybersecurity tool in many different domains

244、and industries.Brain-computer interfaces:A device that reads and analyzes signals from the brain and turns them into an input mechanism for a computer.The human and the device,after a period of training,work together to encode and decode human intentions.CChanging perceptions of AI:AI technologies h

245、ave been widely hyped and are therefore extremely visible not just in the industry but in wider society and culture.This means that attitudes and understandings of it whether thats enthusiasm and excitement or distrust are necessarily important to organizations that decide to use it.The pace of tech

246、nological change,moreover,means that attitudes could also change quickly.Collaboration ecosystems:When individuals or organizations share common goals,they will likely want to work together.To do so,though,they need a set of tools and resources they can use to unlock value effectively a good example

247、 is a remote environment for development teams.This is what a collaboration ecosystem is:it allows people to solve problems together.Consumer XR:Consumer XR refers to products and services that give users extended reality experiences.High-profile devices like the Apple Vision Pro are shaping consume

248、r XR,but the field is highly dependent on innovations in retinal resolution to ensure properly immersive experiences.Context-aware systems:Systems that dynamically adapt their behavior using real-time contextual information,such as user location,activity,or preferences.While the concept has existed

249、since the early days of ubiquitous computing,advancements in AI,IoT,and edge computing have significantly enhanced their capabilities.Modern context-aware systems deliver highly personalized and responsive experiences,becoming a competitive advantage across industries and signaling their importance

250、towards adaptive and human-centric technologies.DData catalog:A comprehensive inventory of an organizations data assets.Crucially,it is built on well-organized metadata,which makes it easier for organizations to discover and retrieve a particular asset and then use it appropriately.Data clean room:S

251、ecure environments for organizations to share and combine data with each other without having to physically share their own data.Data contract:A formal agreement between two parties producer and consumer to use a dataset or data product.Data fitness functions:Automated tests that assess the quality,

252、consistency and reliability of data in real time.By continuously assessing key characteristics,these functions ensure data meets predefined governance standards and remains fit for use in evolving workflows,facilitating interoperability and trust across data systems.Data lineage:An emerging set of t

253、echniques to certify the provenance of data and to govern its use across an organization.This could prove transformative in the effort to track and enhance progress towards sustainability targets.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.45Data marketplaces:A system that enables the

254、 finding,buying,sharing and selling of data within and outside an organization.Data mesh:A data platform organized around business domains where data is treated as a product,with each data product owned by a team.To enable speed and drive standardization,infrastructure teams provide tools that allow

255、 data product teams to self-serve.Data product specification:A precise technical description of a data product that enables its provisioning,configuration and governance.Decentralized data architectures:Use of multiple data stores instead of singular,monolithic centralized stores.A good example is d

256、ata mesh.Decentralized personal data stores:A data architecture style where individuals control their own data in a decentralized manner,allowing access on a per-usage basis(for example,Solid PODs).Decentralized security:Rather than using traditional security perimeters that are a single point of fa

257、ilure,techniques such as zero-trust networks decentralize security checks across the network.Developer experience platforms:Platforms which provide the tooling to make it as effective as possible for developers to create,test and deploy software.They also help developers leverage data effectively.De

258、vSecOps:An abbreviated portmanteau for development,security and operations.This is an approach that includes security as a first-class concern,together with development and operations.Digital carbon management:Measuring organizational greenhouse gas(GHG)emissions and efforts to mitigate those emissi

259、ons.Establishing a carbon footprint and a program to determine it is an essential component on the journey towards net zero and is the first building block towards any sustainability strategy.Digital twin:A virtual model of a process,product or service that allows both simulation and data analysis.3

260、D visualization can be used together with live data,so you can understand what is happening to pieces of equipment you cant actually see.EEasing access to generative AI:Making AI easier to use by lowering the barrier to entry with shared context and other data that those who arent familiar with prom

261、pt engineering may struggle with.Edge computing:Bringing data storage and processing closer to the devices where it is stored,rather than relying on a central location that may be thousands of miles away.The benefits of edge computing include reduced latency for real-time systems and improved data p

262、rivacy.Its also possible to run AI/ML models at the edge too.Ethical frameworks:Decision-making frameworks that attempt to bring transparency and clarity into the way decisions are made,especially around the use of AI and potential bias in data.Evaluating and managing AI outputs:Ensuring the quality

263、,reliability,and safety of AI-generated outputs through evaluation frameworks evals and guardrails.These include systematic tests to measure performance and tools to enforce ethical and operational standards,helping businesses deploy AI responsibly and effectively.Evolutionary architectures:In contr

264、ast to traditional up-front,heavyweight enterprise architectural designs,evolutionary architecture accepts that we cannot predict the future and instead provides a mechanism for guided,incremental change to systems architecture.Explainable AI:A set of tools and approaches to understand the rationale

265、 used by an ML model to reach a conclusion.These tools generally apply to models that are otherwise opaque in their reasoning.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.46FFinOps:The practice of bringing financial accountability to the variable spending model of cloud computing.It in

266、volves a collaborative approach among teams such as finance,operations and development to manage and optimize cloud costs effectively.GGenAI computer control:A new capability of generative AI tools to execute and automate computer-based tasks through natural language.They enhance digital workflows b

267、y enabling intuitive,conversational interactions with operating systems and applications.Examples include Claudes“computer use”feature and Auto-GPT,among others.Generative AI:AI that creates text,image,audio and video from simple human language prompts.Green computing:Green computing is a diverse co

268、llection of practices and techniques that attempt to address the environmental impact of computation.It includes green cloud,green UX and green software development,all of which optimize systems,code and other part of technology infrastructure to improve computational efficiency and reduce waste.HHa

269、rdware security:The growth in smart devices and embedded systems have made hardware an even bigger target for cybercriminals and malicious actors.Ensuring hardware is secure is today a key step in ensuring security across the enterprise.IImpact funds:Impact funds,or impact investing,is a trend where

270、by investors target businesses tackling significant social or environmental challenges in a bid to both develop a solution and,in doing so,unlock substantial financial returns.Industrial XR:Using virtual environments to test and model desired physical outcomes in an industrial context.Integrated dat

271、a and AI platforms:Platforms designed specifically for machine learning,providing end-to-end capabilities such as data management,feature engineering,model training,model evaluation,model governance,explainability,AutoML,model versioning,promotion between environments,model serving,model deployment

272、and model monitoring.Integrating unstructured data:Set of techniques and tools for processing and incorporating unstructured data,such as text,images,and videos,into workflows and decision-making.Approaches like natural language processing,computer vision,and data indexing systems make this data mor

273、e accessible and actionable for businesses.Intelligent machine-to-machine collaboration:Technologies enabling the direct interaction of devices and information sharing between them,usually in an autonomous fashion.This enables exceptionally rapid decision making and action with little or no human in

274、tervention.Interfacing with AI:Establishing standardized methods for integrating generative AI into business systems using tools like LLM proxies and OpenAPI.LLM proxies act as intermediaries that simplify AI interactions,while OpenAPI defines clear,consistent interfaces for connecting AI models to

275、applications,ensuring scalability and ease of use.Internet regulation:The regulation of the internet has become more and more significant in recent years.This manifests itself in many different ways,from attempts to address harmful content,restricting childrens use of social media and rules about ho

276、w consumer data can be collected and used.KKnowledge graphs:A way to represent knowledge and semantic relationships between entities using a graph data structure.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.47LLLMOps:The practice of integrating LLMs into business operations,focusing on

277、 deployment,monitoring,security,and governance.This includes tools and processes for fine-tuning,performance tracking,cost management,and ensuring responsible AI use.MMindful screen interaction:A growing shift toward intentional and balanced device use,driven by increased awareness of screen time.To

278、ols like screen time trackers and focus apps exemplify this trend,supporting users in managing their digital habits.MLOps:A movement to bring DevOps practices to the field of machine learning.MLOps fosters a culture where people,regardless of title or background,work together to imagine,develop,depl

279、oy,operate,monitor and improve machine learning systems in a continuous way.Continuous Delivery for Machine Learning(CD4ML)is Thoughtworks approach to implement MLOps end-to-end.Model training optimization:Strategies and techniques to enhance the efficiency and effectiveness of machine learning mode

280、l training.Examples include retrieval-augmented generation(RAG),which combines data retrieval with generative AI for precise outputs;causal inference,which identifies cause-and-effect relationships to improve generalizability and reduce training data requirements;transfer learning,which leverages pr

281、etrained models for faster adaptation;and automated hyperparameter tuning,which optimizes model performance with minimal manual effort.These approaches are crucial for reducing costs,minimizing energy consumption,and accelerating deployment.Multimodal AI:AI model interactions that span different mod

282、es of communication.For example,a chatbot that understands and responds in both written and spoken language.Multimodal interactions:Systems that enable users to interact through multiple input methods,such as text,voice,image and gesture recognition.By combining these modalities,tools and applicatio

283、ns create more intuitive and accessible experiences across diverse contexts.NNext-generation cryptography:Forms of cryptography created in response to technological or societal challenges.Examples include quantum-resistant encryption algorithms,confidential computing with specialized hardware secure

284、 enclaves,homomorphic encryption allowing computation to occur on the data while it is still encrypted,and energy efficient cryptography.Next-generation robotics:The next generation of robotics is underpinned by advancements in artificial intelligence and machine learning.These technologies are help

285、ing to bring new dimensions of responsiveness and reasoning to robotics.Next-generation wearables:The next generation of wearables are getting smaller but also ostensibly smarter thanks to the increasing integration of AI.These devices ranging from the popular Oura to the Humane AI pin offer users n

286、ew ways to quantify the self.OOnline machine learning:A technique where algorithms continuously learn based on the sequential arrival of data,and can explore a problem space in real time.Contrasts with traditional machine learning where model training uses only historical data and cannot respond to

287、dynamic or previously-unseen situations.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.48PPersonalized healthcare:Understanding an individual patients genetic profile to identify potential issues before they happen and provide more effective treatments in response to existing conditions.

288、Platforms as products:A way of creating and supporting platforms with a focus on providing customer(user)value instead of treating platform building as a time-boxed project.Privacy first:Privacy first is a significant shift in business,organization and product strategy,where privacy operates as a co

289、re business value and offering.This shift moves away from the prior movement where“users are the product”,into a new realm,where building trust and transparency comes first.Privacy-enhancing technologies(PETs):A collection of technologies and techniques designed to preserve user privacy while enabli

290、ng secure and trustworthy interactions.Examples include anonymization,encrypted computing,differential privacy,decentralized identity(DiD)for self-owned digital IDs and verifiable credentials,and zero-knowledge proofs,which allow validation without exposing sensitive data.These tools play a critical

291、 role in safeguarding privacy in increasingly data-driven and interconnected systems.Production immune systems:Systems that monitor metrics across complex distributed systems and take corrective action if a problem is detected.They are often used for security,but increasingly also for resilience and

292、 recovery in the face of an outage.QQuantum computing:The use of probabilistic states of photons,rather than binary ones and zeros,to execute algorithms with significant speedup in specific problem domains.Recent advancements,such as Googles breakthroughs in quantum error correction,signal progress

293、toward scalable systems.However,these developments also raise concerns about security,as quantum computers could potentially break classical cryptographic protocols,driving interest in quantum-resistant encryption methods.RResponsible tech facilitation:Tools and techniques are emerging that support

294、incorporating responsible tech into software delivery processes,primarily focusing on actively seeking to incorporate under-represented perspectives;some examples include Tarot Cards of Tech,Consequence Scanning,and Agile Threat Modeling.SSatellite networks:High-speed,low-latency broadband for place

295、s where traditional fiber or wireless network providers wont spend the money to connect.Examples include Starlink from SpaceX,Kuiper from Amazon,OneWeb and Telesat.Secure software delivery:Security applied to the entire process of software creation,which in modern architectures includes the delivery

296、 pipeline used to build,test and deploy applications and infrastructure.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.49Semantic representational technologies:A collection of techniques aimed at helping machines better understand data.It aims to put meaning at the very center of data,so

297、 concepts,categories and relationships can be better understood by machines.For users,this can make it easier to search and manage incredibly complex data sets.Small language models:An alternative to large language models(LLMs)that are more lightweight and efficient.While they arent as powerful comp

298、ared to their larger siblings,because they require less memory and computational power they can be used in devices at the edge of a network.Smart systems and ecosystems:Networks of networks that use AI and ML to enhance a system to become more than the sum of its parts.For example,in a smart city,ne

299、tworks of cars and roadside sensors help speed the flow and safety of traffic.Software-defined vehicles:Automobiles where the core functionalities,features and user experience are primarily governed by software,rather than traditional mechanical and electrical systems.This approach enables increased

300、 flexibility,customization and continuous enhancement through remote updates,significantly transforming the vehicles capabilities and,in turn,the automotive industrys business models.Synthetic data:Artificial data that mimics real data.It is created algorithmically,expanding the potential size of a

301、data set without requiring further data collection.This has many applications,from drug research to testing,and also has the benefit of reducing the risks and challenges that come from acquiring new,real data.TTactile interaction:Tactile interaction is an emerging trend in extended reality.It uses s

302、omething called haptic feedback to enable richer and more immersive experiences where users can physically experience a virtual environment.Talk to data:Talk to data(T2D)is a technology that allows users to interact with and analyze data using natural language queries as opposed to,say,the kinds of

303、analytics and business intelligence dashboards that have become commonplace over the last two decades.It makes it easier to uncover insights and has a lower barrier to entry,giving more employees the ability to explore and ask questions about data.Touchless interactions:The ability to interact with

304、devices without touching.Specific technologies include hand tracking and voice and gesture recognition.UUnderstandable consent:Most terms of service(TOS)or end-user license agreements(EULAs)are impenetrable legalese that make it difficult for people without a law background to understand.Understanda

305、ble consent seeks to reverse this pattern,with easy-to-understand terms and clear descriptions of how customers data will be used.VVector databases:Specialized storage systems designed to efficiently handle and index high-dimensional data vectors,commonly used in machine learning and AI applications

306、.GlossaryLooking Glass Thoughtworks,Inc.All Rights Reserved.50Thoughtworks is a global technology consultancy that integrates strategy,design and engineering to drive digital innovation.We are over 10,000 Thoughtworkers strong across 48 offices in 19 countries.For 30+years,weve delivered extraordinary impact together with our clients by helping them solve complex business problems with technology as the differentiator.