生成式人工智能并不是唯一可以在来年改变公司的计划。解决运营差距和培养新的数字领导者也可以成为力量倍增器。
分析师报告告诉首席信息官，生成式人工智能应该在未来一年占据其数字化转型优先事项的首位。
此外，首席信息官所报告的首席执行官和董事会不希望被生成式人工智能抛在后面，许多员工希望在他们的工作流程中尝试最新的生成式人工智能功能。今年的技术宠儿和其他机器学习投资已经影响了 2023 年的数字化转型战略，董事会希望首席信息官经常更新他们的人工智能转型战略。幸运的是，许多人正在扩大预算以此。 数字化转型网（www.szhzxw.cn）
“我们调查的高管中有94%表示，他们计划增加对IT或新兴技术的投资，80%的高管表示他们计划在明年增加对人工智能的投资，”安永美洲技术、媒体和电信业务负责人Ken Englund表示。
尽管如此，首席信息官们应该将生成式人工智能计划添加到更长的转型计划二级清单中。毕竟，每个部门都面临着提高效率的压力，并要求自动化、数据功能和改善员工体验，其中一些可以通过生成式人工智能来解决。同时，首席信息官仍必须减少技术债务，实现应用程序现代化，并控制云成本。
正如每个 CIO 都可以证明的那样，对 IT 和数据功能的总体需求正在使他们的 IT 领导团队不堪重负。因此，在过度承诺太多优先事项之前，首席信息官应该重置和审查什么是转型计划，采取措施避免炒作夸大的目标，并与执行团队一起为未来一年设定合理的优先事项。方法如下。
设定转型的优先事项，而不仅仅是转型
在满足对技术能力的压倒性需求时，应敏锐地理解数字化转型和过渡性技术投资之间的差异。我在《推动数字化》一书中写道：“数字化转型不仅仅是技术及其实施。这是关于通过技术能力的视角来看待业务战略，以及它如何改变你的运营和创收方式。 数字化转型网（www.szhzxw.cn）
换句话说，首席信息官应该寻找影响业务战略的投资，并且是影响创收和运营效率的力量倍增器。许多技术投资只是过渡性的，将今天完成的事情升级为更好的能力，而不一定改变业务或运营模式。
其次，首席信息官必须确定企业领导者对 IT 部门的期望与现实之间的差距。在研究本文时，我发现 CIO 向利益相关者承诺能力，但实施和业务影响落后于预期。在许多情况下，这种滞后可以在任何承诺之前被预见和解决。
在考虑明年的优先事项时，以下举措对许多组织来说都是转型的好选择，这在很大程度上是通过帮助首席信息官缩小执行差距，同时提供持续的业务影响。
1. 创建这 6 个生成式 AI 工作流
首席信息官应记录其 AI 战略，以在规划有远见的影响的同时实现短期生产力改进。生产力的提高可能来自对嵌入提示和其他自然语言功能的平台和工具的试验，而长期影响将来自将公司的知识产权嵌入到私人管理的大型语言模型中。 数字化转型网（www.szhzxw.cn）
许多首席信息官已经在优先考虑生成式 AI 工作流，例如：
- 尝试使用 ChatGPT、副驾驶和其他工具
- 评估大型语言模型、嵌入和向量数据库
- 为组织定义生成式 AI 护栏、审查风险并翻译新法规
但是，首席信息官应该在这些 AI 工作流中增加三个执行差距。
首先，企业长期以来一直在努力改善客户、员工和其他搜索体验。对于希望提供生成式 AI 功能的 CIO 来说，提高搜索能力和解决非结构化数据处理挑战是关键差距。《2023 年企业搜索：无名英雄》报告发现，98% 的组织表示他们正在改进门户、CRM 工具、电子商务网站和在线社区的搜索功能。但 99% 的受访者也报告了技术挑战，将集成 （68%）、数据量和清理 （59%） 以及管理非结构化数据 （55%） 列为前三名。在改善搜索体验的同时缩小这些差距是促进 gen AI 的提示和自然语言查询功能的智能构建块。
首席信息官必须解决的第二个差距对于结果和可信度都至关重要。在《2023 年数据科学和机器学习现状报告》中，只有 18% 的受访者表示，他们至少有一半的机器学习模型投入生产。如果首席信息官不提高从试点到生产的转换率，他们可能会发现他们的投资者在实验的过程和文化中失去了耐心。
第三，在《CDO议程：2024年：驾驭数据和生成式人工智能前沿》中，57%的受访者尚未改变其数据环境以支持生成式人工智能。毫不奇怪，近 50% 的受访者认为数据质量是实现生成式 AI 潜力的最大挑战，这需要扩大数据运营和数据治理计划的范围，以包括非结构化数据源，并定义在 LLM 实验中应用它们的位置。 数字化转型网（www.szhzxw.cn）
这些工作流需要记录愿景、分配领导者并授权团队进行实验。首席信息官应该着眼于设定切合实际的目标，并传达哪些工作流是优先事项。
2. 通过缩小运营和安全差距来推动业务影响
数字化转型旨在提供竞争优势，通常通过新的数字产品、改善的客户体验和数据驱动的决策。这些举措因行业和业务战略而异，可能包括智能制造、数字健康、电子政务举措、可持续发展计划、数字孪生等。
这些计划的基础是数字化转型的核心能力，包括设计思维、产品管理、敏捷方法、DevOps实践、公民发展和数据治理。这些实践包括以运营和风险管理能力为后盾的领导力和交付实践，研究表明，正是这些能力，许多组织都落后了。首席信息官应考虑缩小这些差距在其数字化转型优先事项中的位置。
例如，在过去十年中，许多 CIO 推动了 DevOps 实践，例如 CI/CD、基础设施即代码、数据可观测性，以及缩小开发团队和 IT 运营之间的文化差距。在 2023 年 DevOps 基准研究中，表现最好的公司实现了显著的业务优势，其中 67% 的企业实现了按需部署频率，94% 的企业实现了变更失败率低于 15%。
但是，根据其他调查，尽管许多组织将其称为 devsecops，但实施 devsecops 安全最佳实践却滞后。
在 SANS 2023 DevSecOps 调查中，不到 22% 的受访者在不到两天的时间内修补并解决了关键安全风险和漏洞。虽然静态应用程序安全测试（SAST）是82%的受访者认为最有用的工具，但只有28%的受访者声称这些工具至少在75%的代码库中使用。 数字化转型网（www.szhzxw.cn）
在这里，《2023 年 DevSevOps 现状》报告强调了 CIO 在 2024 年数字化转型中应重点关注的几个差距，包括安全培训不足、应用程序安全人员短缺以及开发和运营工作流程缺乏透明度。
首席信息官应寻找其他运营和风险管理实践来补充转型计划。例如，麦肯锡估计，企业可以通过优化将云成本降低 15% 到 20%，而实施 finops 实践就是一个机会。另一个需要关注的领域是持续测试，特别是因为生成式人工智能和副驾驶可以提高代码开发的速度和代码生成器的风险。
3. 培养转型领导者以推动更多举措
对创新、企业级转型以及部门技术、数据和自动化能力的需求不断增长，这是当今 CIO 面临的重大挑战。与项目经理、架构师和其他在转型计划中发挥作用的领导者相比，技术投资请求的数量是相当普遍的。再加上积压的技术债务、安全改进以及影子 IT 转型的工作，这是一个过度投入的 IT 部门。
虽然并非所有这些举措都符合我对数字化转型的定义，但能够监督更多举措、标准化多用途平台并提供受监管的公民数据科学计划可能具有高度的变革性。转型的驱动力是培养更多的数字化转型领导者，或者我所说的数字开拓者。
Gartner 在 2024 年对人力资源的重中之重是领导者和经理的发展，因为人力资源领导者报告称，76% 的经理对工作职责的增长感到不知所措，并且承担的责任比他们所能处理的多 51%。我怀疑许多首席信息官在他们的部门内都在为同样的问题而苦苦挣扎，尤其是在要求数字开拓者领导和贡献太多计划时。
为了解决这个问题，首席信息官应该与人力资源部门合作，利用他们的培训和发展预算，并启动计划来培养转型和变革管理领导者。这样做将帮助 CIO 增加 IT 部门可以启动的计划数量，同时更快地交付结果并减少组织在变革中的摩擦。 数字化转型网（www.szhzxw.cn）
随着首席信息官展望 2024 年，在其数字化转型计划中优先考虑各项举措的适当平衡将具有挑战性。请记住，过度承诺和交付不足是失败的秘诀，因此，对于面临更大、更快影响压力的首席信息官来说，增加数字化转型领导者的数量并投资于学习计划是关键。
英文原文：
3 key digital transformation priorities for 2024
Generative AI is not the only initiative that can transform your company in the coming year. Addressing operational gaps and building up new digital leaders can also be force multipliers.
The analyst reports tell CIOs that generative AI should occupy the top slot on their digital transformation priorities in the coming year. 数字化转型网（www.szhzxw.cn）
Moreover, the CEOs and boards that CIOs report to don’t want to be left behind by generative AI, and many employees want to experiment with the latest generative AI capabilities in their workflows. This year’s technology darling and other machine learning investments have already impacted digital transformation strategies in 2023, and boards will expect CIOs to update their AI transformation strategies frequently. Luckily, many are expanding budgets to do so.
“94% of the executives we surveyed said they had plans to increase their investment in IT or emerging technologies, and 80% said they plan to increase investment in AI over the next year,” reports Ken Englund, leader of technology, media, and telecommunications for the Americas at EY.
Despite all that, CIOs should add generative AI initiatives to the much longer secondary list of transformation programs. After all, every department is pressured to drive efficiencies and is clamoring for automation, data capabilities, and improvements in employee experiences, some of which could be addressed with generative AI. Meanwhile, CIOs must still reduce technical debt, modernize applications, and get cloud costs under control. 数字化转型网（www.szhzxw.cn）
As every CIO can attest, the aggregate demand for IT and data capabilities is straining their IT leadership teams. So, before overcommitting to too many priorities, CIOs should reset and review what constitutes a transformational initiative, take steps to avoid hype-inflated goals, and set reasonable priorities with the executive team for the year ahead. Here’s how.
Setting priorities that transform, not just transition
Overwhelming demand for technology capabilities should be met with a keen understanding about the differences between digital transformation and transitionary technology investments. I wrote in Driving Digital, “Digital transformation is not just about technology and its implementation. It’s about looking at the business strategy through the lens of technical capabilities and how that changes how you operate and generate revenues.” 数字化转型网（www.szhzxw.cn）
In other words, CIOs should look for investments that impact the business strategy and are force multipliers, impacting both revenue generation and operational efficiencies. Many technology investments are merely transitionary, taking something done today and upgrading it to a better capability without necessarily transforming the business or operating model.
Second, CIOs must identify the gaps between the hype of what business leaders expect from IT departments and the realities. In researching this article, I found gaps where CIOs promised capabilities to stakeholders, but implementations and business impacts have lagged expectations. In many instances, this lag can be anticipated and addressed prior to any commitments.
In considering next year’s priorities, the following initiatives are good bets to be transformational for many organizations, in large part by helping CIOs close execution gaps while delivering ongoing business impacts.
1. Create these six generative AI workstreams
CIOs should document their AI strategy for delivering short-term productivity improvements while planning visionary impacts. Productivity improvements will likely come from experimenting with the platforms and tools that embed prompting and other natural language capabilities, while longer-term impacts will come from embedding the company’s intellectual property into privately managed large language models. 数字化转型网（www.szhzxw.cn）
Many CIOs are already prioritizing generative AI workstreams such as:
- Experimenting with ChatGPT, copilots, and other tools
- Evaluating large language models, embeddings, and vector databases
- Defining generative AI guardrails for the organization, reviewing risks, and translating new regulations
But there are three execution gaps that CIOs should add to these AI workstreams.
First, enterprises have long struggled to improve customer, employee, and other search experiences. Improving search capabilities and addressing unstructured data processing challenges are key gaps for CIOs who want to deliver generative AI capabilities. The 2023 Enterprise Search: The Unsung Hero report found that 98% of organizations say they are improving search capabilities on portals, CRM tools, ecommerce sites, and online communities. But 99% also report technical challenges, listing integration (68%), data volume and cleansing (59%), and managing unstructured data (55% ) as the top three. Closing these gaps while improving search experiences are smart building blocks for facilitating gen AI’s prompting and natural language query capabilities.
The second gap CIOs must address is critical both for results and for credibility. In the 2023 State of Data Science and Machine Learning Report, only 18% of respondents said that at least half their machine learning models make it into production. If CIOs don’t improve conversions from pilot to production, they may find their investors losing patience in the process and culture of experimentation.
Third, in the CDO Agenda: 2024: Navigating Data and Generative AI Frontiers, 57% of respondents haven’t changed their data environments to support generative AI. Not surprisingly, nearly 50% of respondents identified data quality as the top challenge in realizing the potential of generative AI, which requires increasing the scope of dataops and data governance programs to include unstructured data sources and defining where to apply them in LLM experiments. 数字化转型网（www.szhzxw.cn）
These workstreams require documenting a vision, assigning leaders, and empowering teams to experiment. CIOs should look to set realistic goals and communicate which workstreams are priorities.
2. Drive business impacts by closing operational and security gaps
Digital transformations aim to deliver competitive advantages typically through new digital products, improved customer experiences, and data-driven decision-making. Initiatives vary by industry and business strategy and may include smart manufacturing, digital health, e-government initiatives, sustainability programs, digital twins, and others.
Underpinning these initiatives are digital transformation core competencies, which include design thinking, product management, agile methodologies, devops practices, citizen development, and data governance. These practices include leadership and delivery practices backed by operational and risk management competencies, and it’s these competencies where research shows many organizations are lagging. CIOs should consider where closing these gaps falls in their digital transformation priorities.
For example, over the past decade, many CIOs promoted devops practices such as CI/CD, infrastructure as code, data observability, and closing the cultural gaps between development teams and IT operations. In the DevOps Benchmarking Study 2023, top performers realize significant business benefits, including 67% with on-demand deployment frequency and 94% with change failure rates under 15%.
But while many organizations call it devsecops, implementing devsecops security best practices is lagging, according to other surveys. 数字化转型网（www.szhzxw.cn）
In the SANS 2023 DevSecOps Survey, less than 22% of respondents patched and resolved critical security risks and vulnerabilities in under two days. And while static application security testing (SAST) was the top-rated tool for usefulness by 82% of respondents, only 28% claim these tools are used on at least 75% of their code base.
Here, the State of DevSevOps in 2023 report highlights several gaps CIOs should focus on closing in their 2024 digital transformations, including inadequate security training, shortages of application security personnel, and lack of transparency in development and operations workflows.
CIOs should look for other operational and risk management practices to complement transformation programs. For example, McKinsey estimates enterprises can cut 15% to 20% of cloud costs through optimizations, and implementing finops practices is one opportunity. Another area to focus on is continuous testing, especially as generative AI and copilots can increase the velocity of code development and risks from code generators. 数字化转型网（www.szhzxw.cn）
3. Develop transformation leaders to drive more initiatives
The increasing demand for innovation, enterprise-wide transformation, and departmental technology, data, and automation capabilities is a significant challenge for CIOs today. It’s fairly common to see more technology investment requests than there are program managers, architects, and other leaders with roles in transformation initiatives. Add on the backlog of technical debt, security improvements, and work to transform shadow IT, and that’s one overcommitted IT department.
While not all these initiatives fit my definition of digital transformation, being able to oversee more initiatives, standardize multi-use platforms, and provide governed citizen data science programs can be highly transformative. The transformation is driven by developing more digital transformation leaders, or what I call digital trailblazers.
Gartner’s top priority for HR in 2024 is leader and manager development, given that HR leaders report that 76% of managers are overwhelmed by the growth of their job responsibilities, and juggle 51% more responsibilities than they can handle. I suspect many CIOs struggle with these same issues within their departments, especially when asking digital trailblazers to lead and contribute to too many initiatives.
To address this, CIOs should partner with HR, tap into their training and development budgets, and initiate programs to develop transformation and change management leaders. Doing so will help CIOs increase the number of initiatives IT can launch, while delivering results faster and reducing the organization’s friction to change. 数字化转型网（www.szhzxw.cn）
As CIOs look into 2024, prioritizing the right balance of initiatives in their digital transformation programs will be challenging. Keep in mind that overpromising and underdelivering are recipes for failure, so increasing the number of digital transformation leaders and investing in learning programs are key for CIOs facing pressure to deliver greater and faster impacts.
本文由数字化转型网（www.szhzxw.cn）转载而成，来源于CIO.COM；编辑/翻译：数字化转型网宁檬树。
免责声明: 本网站(http://www.szhzxw.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。
本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。
