Capgemini-配备Gen AI的涡轮增压软件-组织如何充分发挥生成性人工智能在软件工程中的潜力（英）-2024

How organizations can realize the full potentialof generative AI for software engineering Table ofContent Organizations are reaping multiple benefitsfrom leveraging generative AI for softwareengineering. •Organizations are utilizing these productivity gains oninnovative work such as developing new software features(50%) and upskilling (47%). Very few aim to reduceheadcount (4%). ExecutiveSummary •Generative AI is having a positive impact on softwareprofessionals’ job satisfaction. •The leading benefits for organizations are enablingmore innovative work, such as developing new softwarefeatures/services (observed by 61% of surveyedorganizations), improving software quality (49%), andincreasing productivity (40%). •69% of senior software professionals and 55% of juniorsoftware professionals report high levels of satisfactionfrom using generative AI for software. •78% of software professionals are optimistic aboutgenerative AI’s potential to enhance collaborationbetween business and technology teams. •Organizations using generative AI have seen a 7–18%productivity improvement1in the software engineeringfunction as per early estimates. This is highest forspecialized tasks such as coding assistance2(34% as themaximum potential for time savings with 9% on average)and creating documentation (35% as the maximumpotential for time savings with 10% on average). Thisresearch analyzed time savings in various softwareengineering tasks using generative AI tools and not costsavings which can be significantly different. •Generative AI is expected to play a key role in augmentingthe software workforce with better experience, tools andplatforms, and governance (assisting in more than 25% ofsoftware design, development, and testing work by 2026). Generative AI adoption is at an early stagebut will accelerate sharply. ExecutiveSummary •Adoption of generative AI for software engineering is stillin its early stages, with 9 in 10 organizations yet to scale. •Coding assistance is the leading use case, but generativeAI also finds applications in other software developmentlifecycle (SDLC) activities (test case generation,documentation, code modernization, UX design assistance,etc.) •27% of organizations are running generative AI pilots,and 11% have started leveraging generative AI in theirsoftware functions. •Three in four (75%) large organizations (annual revenuegreater than $20 billion) have adopted (piloted/scaled) generative AI compared to 23% of their smallercounterparts (annual revenue between $1–5 billion). •Most use cases have yet to be adopted by a majority oforganizations (39% are focusing on coding assistance and37% on UX design assistance as top adopted use cases). •Adoption (including pilots) is expected to increasesignificantly in the next two years from 46% of softwareworkforce using generative AI tools today (for any kind oftraining, experimenting, piloting, and implementing, withauthorized or unauthorized access) to an estimate of 85%in 2026. Lack of foundational prerequisites andunofficial usage of generative AI posesignificant functional, security, andlegal risks. •Using unauthorized tools without proper governance andoversight exposes organizations to functional, security,and legal risks like hallucinated code, code leakage, andIP issues. ExecutiveSummary •27% of organizations have the platforms & tools, and 32%have talent prerequisites in place, to implement generativeAI for software engineering. •Over 60% lack governance and upskilling programs forgenerative AI for software engineering. •Of those software professionals who use generative AI,63% use unauthorized tools. •Nearly a third of the workforce is self-training ongenerative AI for software as less than 40% of employeesare receiving training from their organizations. How can organizations harness the fullpotential of generative AI for softwareengineering? retention, solving complex issues, and collaboratingwith business. •Identify requirements for new capabilities andsource them. ExecutiveSummary •Prepare for generative AI use by delivering technologyprerequisites: •Select and prioritize high benefit use cases. •Build a repository of platforms and tools for a seamlessand augmented software engineering experience. •Mitigate risks around security, IP/copyright issues, andcode leakage using a thorough risk management approach. •Privately and safely contextualize generative AIassistants with organization’s own content. •Transform your software organization to ensure optimalusage of generative AI: •Adopt a measurement protocol for generative AI impactmonitoring and use case prioritization. •Augment your software teams with a generative AIassistant. A majority of junior (53%) as well as seniorprofessionals (58%) believe that generative AI toolswill augment their day-to-day work within the nexttwo years. For instance, generative AI tools can helpjunior professionals learn faster and come up to speedquickly, while they allow seni