完成

特邀社论:不要把完成任务误认为胜任工作

随着人工智能和在线培训的兴起,确保员工和公司将安全视为不仅仅是走过场,这一点至关重要。

一只机械手正在黑色背景的弹出窗口中勾选标有“我不是机器人”的复选框。这幅3D插图描绘了一个荒谬的场景:人工智能(AI)可以轻松绕过图灵测试。
图片来源:Dragon Claws/Getty Images。

十年前,石油和天然气行业正面临着劳动力老龄化、逐步退休的局面,与此同时,新一代员工开始承担复杂且高风险的工作。这一被称为“人员大更迭”的过渡期,引发了人们对失去实战经验的担忧,而这些经验只有多年的现场工作才能积累。

如今,这种转变已步入正轨,但安全方面却出现了另一种意想不到的威胁:那就是我们如何培训和验证那些将维持我们行业运转的人员的完整性。

采用新技术对于保持竞争力至关重要,而将培训转移到线上扩大了培训范围并降低了成本。然而,如果组织未能监控其在线培训的绕过方式,就会出现一个关键缺陷。

因此,虽然将培训转移到线上势在必行,但我们也不能忽视人们很容易点击浏览课程却学不到任何有意义的东西。

对于石油和天然气行业的工人而言,由于该行业存在诸多独特的、严重的安全隐患,导致工作场所死亡率居高不下,因此培训必须是一个可验证的过程,而不是旨在规避法律和保险责任的“风险规避”手段。如果企业仅仅为了应付合规手续而让工人对高风险情况毫无准备,这是不可接受的。

石油和天然气行业的劳动力正在发生变化。通过在线模块培训的新员工正进入一个真正的行业,在这个行业里,实际经验和敏捷的思维可能意味着生死攸关。但是,如果我们不改变在线培训的方式,这一次,下一代恐怕难以幸免。

当我们思考工作场所的危险时,我们往往关注的是一线风险,即直接接触危险因素的环节。但一线风险环节的失误往往反映出整个决策链中存在的潜在问题,其中最紧迫的例子就是培训不足。

随着人工智能(AI)的飞速发展,绕过在线培训不仅变得轻而易举,而且手段也极其高明。人工智能工具能够回答复杂的问题,生成看似合理的答案,甚至模拟人类的互动,使人们无需亲临现场即可“完成”课程,从而吸收关键的安全信息。

在 Cognisense,我们办公室的巴哥犬菲比已经获得了 200 多项认证,涵盖各种高风险行业。她最新的成就是什么?一张政府颁发的在线防御性驾驶证书。

菲比不会开车——她养的是狗——但从纸面上看,她却持有安全驾驶证书。这就是获得一些在线安全培训证书的便捷之处。

这些缺陷并非个例。2022年,英国石油公司位于俄亥俄州的一家炼油厂发生爆炸,造成两名工人死亡调查人员发现,该公司未能对操作人员进行适当培训,使其能够在异常操作条件下识别石脑油,并执行必要的停机程序。

石油和天然气行业另一个鲜为人知的危险是道路安全培训不足。许多公司依赖合同司机,而这些司机的道路安全培训缺乏监管。美国联邦数据显示,2023年德克萨斯州石油工人死亡事故中,交通事故占三分之二,这使得道路对当地社区而言日益危险。正如我们从菲比的案例中所了解到的,获得安全驾驶资格证书很容易被绕过。

这些案例暴露了培训环节的薄弱环节。我们看到课程未经审核、验证不足,以及一种重完成率而非理解力的文化。其后果不仅限于受伤的工人,还波及受影响的家庭和社区。

好消息是我们拥有解决这个问题的工具和标准。我曾与众多组织合作,以满足ANSI/ASSP Z490.1标准中规定的身份验证和真实参与要求。该标准是一项健康、安全和环境培训标准,最近进行了更新,以适应后疫情时代在线培训的广泛应用。

确保安全培训真正实现知识传递的工具和技术已在业内广泛应用。通过与各公司紧密合作,我们已通过验证其在线培训课程证明了这一点。

在我之前担任军方消防员的职业生涯以及多年在石油和天然气行业的从业经历中,我目睹了当工人无法在实际工作中依赖所受训练时会发生什么。当这种信任被侵蚀时,系统就会出现裂缝,灾难不仅可能发生,而且极有可能发生。

劳动力结构在变化,我们用来培训他们的工具也在变化。但有一点绝不能改变,那就是行业对安全的承诺。

雇主和监管机构有责任确保下一代石油和天然气工人能够拥有安全、成功和充实的职业生涯和生活。

罗伯特·戴 (Robert Day)是 Cognisense 公司的董事总经理,在监管合规和行业标准领域是一位杰出的专家,尤其擅长技术与培训/评估标准的交叉领域。戴的职业生涯始于一名军方消防员,这段经历奠定了他毕生致力于风险规避的基础。几十年来,他曾领导危险品应急响应小组,为大型保险公司进行风险评估,并为法律和监管机构提供合规策略方面的咨询。他的工作对跨行业、跨国界的立法和行业标准制定产生了深远的影响。如今,戴致力于帮助企业应对在线培训的复杂性,专注于验证项目的完整性,以确保员工队伍具备胜任能力和安全意识,能够应对任何挑战。

原文链接/JPT
Completions

Guest Editorial: Stop Mistaking Completion for Competence

With artificial intelligence and online training both on the rise, it's crucial to ensure workers and companies treat safety as more than a box-checking exercise.

Robot hand is checking the box with a label I’M NOT A ROBOT on a window popup on a black background. 3D illustration of ridiculous situation artificial intelligence (AI) can easily bypass turing tests
Source: Dragon Claws/Getty Images.

A decade ago, the oil and gas sector was bracing for an aging workforce edging toward retirement while a new generation stepped into complex, high-risk roles. This transition, dubbed “The Great Crew Change,” raised real concerns about losing hands-on, battle-tested knowledge that only many years in the field can teach.

Today that transfer is well underway, but a different and unexpected threat to safety has emerged: the integrity of how we train and verify those who will keep our industry running.

Adopting new technology is essential to remain competitive, and moving training online has expanded access and reduced costs. However, a critical flaw arises when organizations fail to monitor how their online training can be bypassed.

So, while moving training online is imperative, we must also not turn a blind eye to how easy it is to click through a course without learning anything meaningful.

For workers in the oil and gas industry, where distinct and acute safety hazards contribute to some of the highest workplace fatality rates, training must be a verifiable process, not a risk-washing exercise designed to keep legal and insurance liabilities at bay. It’s unacceptable for organizations to check a compliance box while workers remain unprepared for high-risk situations.

The oil and gas workforce is evolving. New workers trained through online modules are entering an industry where real experience and quick thinking can mean the difference between life and death. But unless we change how we approach online training, this time the next generation won’t be fine.

When we think about workplace hazards, we often focus on the sharp end, the frontline where direct interactions with hazards occur. But failures at the sharp end often reflect latent issues from the whole decision-making chain, the most pressing example being improper training.

With the rapid advancement of artificial intelligence (AI), bypassing online training has become not just easy, but remarkably sophisticated. AI tools can answer complex questions, generate plausible responses, and even simulate human-like engagement, allowing individuals to "complete" courses without even being present to absorb the critical safety information.

At Cognisense, our office pug Phoebe has earned more than 200 certifications across various high-risk industries. Her latest achievement? A government-issued online defensive driving certificate.

Phoebe doesn’t know how to drive—she’s a dog—yet, on paper, she’s certified to operate a vehicle safely. That’s how easy it is to earn some online safety training credentials.

These weaknesses are not merely anecdotal. In a 2022 incident at a BP refinery in Ohio, two workers died in an explosion. Investigators found the company failed to properly train operators to recognize naphtha during abnormal operating conditions and to execute required shutdown procedures.

Another example of an under-discussed hazard in the oil and gas industry is improper road safety training. Many companies rely on contract drivers who have little oversight over their road safety training. US federal data shows crashes accounted for two-thirds of Texas oil worker fatalities in 2023, making roads increasingly dangerous for local communities. As we know from Phoebe, receiving credentials in safe driving can be easily bypassed.

These examples expose weak links in the training chain. We see unchecked courses, inadequate verification, and a culture that prioritizes completion over comprehension. The consequences extend beyond injured workers to impacted families and communities.

The good news is we have the tools and standards to fix this. I have worked with numerous organizations to meet identity verification and authentic participation requirements outlined in the ANSI/ASSP Z490.1 standard, which is a health, safety, and environmental training standard that was recently updated to account for the widespread adoption of online training in a post-COVID-19-pandemic world.

The tools and technology to ensure safety training delivers real knowledge transfer are readily available to the industry. By working closely with companies, we’ve been able to prove this by validating their online training.

In my previous career as a military firefighter, and through my years in the oil and gas sector, I’ve seen what happens when workers cannot rely on their training in real time. When that trust erodes, cracks form in the system, and disaster becomes not just possible, but probable.

The workforce is changing, and the tools we use to train them are changing as well. What must never change is the industry’s commitment to safety.

It’s up to employers and regulators to ensure the next generation of oil and gas workers are equipped to lead safe, successful, and fulfilling careers—and lives.

Robert Day, managing director of Cognisense, is a distinguished expert in regulatory compliance and industry standards relating to the intersection of technology and training/assessment standards. Day began his career as a military firefighter, laying the foundation for a lifelong commitment to risk mitigation. Over the decades, he has led hazardous materials response teams, conducted risk assessments for major insurers, and advised legal and regulatory bodies on compliance strategies. His work has helped shape legislation and industry standards across sectors and borders. Day now helps organizations navigate the complexities of online training, focusing on verifying program integrity to ensure a competent and safe workforce that can meet any challenge.