I would wager that if I asked ten people to provide a short definition for Artificial Intelligence (AI), no two responses would be the same. There is, as of this writing’s date, no one legal definition of AI that is universally recognized. In fact, when we talk about AI, it’s important to recognize we are discussing a technology that is still in its infancy. Many countries are currently working to incorporate AI-related definitions into specific laws or regulations to help govern this emerging technology landscape. However, the truth remains that when people use the term “AI’, they are often-and perhaps unbeknownst to them-often talking about different things.
For the sake of this blog, we will define AI as “an umbrella term that refers to the use of computing machines to solve problems traditionally requiring human intelligence, including pattern recognition, predictive modeling, scenario analysis and process optimization”.
Here at Cority, we include an additional element in this definition when considering the role that AI plays in the EHS space. It’s our view that AI is about creating technology that augments human capabilities and improves overall productivity and quality of life. In other words, AI development shouldn’t be to replace humans, but rather to enhance and extend their abilities to solve real problems. In this blog, we’ll delve further into the capabilities of AI and how EHS teams can benefit from the technology, as well as, the relationship between EHS and AI for future opportunities.
What is Artificial Intelligence (AI)?
AI technologies are designed to complement human skills by automating routine tasks, providing insights from large amounts of data, assisting in decision-making processes, and enabling new forms of interaction and communication.
In this sense, we can (and should) think of AI not as a single solution, but rather as a spectrum of technology, each oriented toward a specific objective. From Basic Automation tasks (e.g., simple field auto-fill) to very advanced Cognitive Automation applications akin to IBM’s Watson‘s ability to write its own code without human intervention.
What many firms fail to recognize, however, is that the most advanced AI is not always the best approach. Cognitive computing offers a more transformative impact on the business, but that comes with higher complexity, limited availability, higher costs, and longer timelines to implement and optimize.
The Purpose of AI
A key thing for anyone exploring the world of AI to remember is that just because something is already available, relatively simple to implement and utilize, and without significant investment, doesn’t mean it’s not useful. It’s the Goldilocks problem; we want to ensure our selected software offers ‘just the right amount’ of AI functionality for our business’ specific needs.
Sometimes that means we should explore Task-Based, Machine Learning Automation tools like computer visions or natural language processing. Other times, more simplistic or basic rules-based automation is enough to reach the desired outcome.
Finally, when assessing the overall purpose of AI, we need to consider the relative strength of AI compared to human involvement. In this respect, there are three scenarios where AI and humans converge:
- Tasks in which AI excels and outperforms a human actor;
- Tasks in which AI falls well short when compared to a human actor, and;
- Tasks in which related outcomes are far improved when a human leverages AI to augment their innate skills and decision making.
As we consider the potential role of AI for EHS, it’s vital to remember that there are certain areas in which AI is simply not yet capable of competing with a trained, rational human. AI currently excels in repetitive tasks, along with collecting, aggregating and analyzing data, at a speed and scale that far exceeds human capabilities.
On the contrary, however, AI currently falls well behind human involvement in areas like assessing human intention and behavior or delivering front-line assistance (personal care or therapy). Yet, when AI is combined with human intelligence, curiosity, and rationality – it can be an incredibly useful tool, resulting in powerful augmentation and collaboration between man and machine.
The Intersectionality of AI & EHS
The opportunities that AI presents for EHS are vast and numerous – far too many to list in one blog post! However, as discussed in our recent webinar “Fantastic Voyage: Navigating the Growing Opportunities of AI for EHS Excellence”, we noted some core EHS uses cases that are ideally suited for AI integration:
Task Assistance:
This category is centered on AI-powered innovations focused on increasing process efficiency and optimization, specifically through automating repetitive tasks and functions, and streamlining workflows to increase user and program efficiency. Here, organizations leverage AI to remove repetitive, administrative task burdens from employees so they can focus on more value-added activities.
Insights & Artifacts:
Innovation in this category is focused on how to leverage AI to more easily and quickly detect, interpret and classify data patterns and associations from collected data and surface relevant insights to users to inform decision making and action.
Expert Solutions:
This use case focuses on leveraging machine learning tools, including computer vision, to collect and interpret complex data, to both reduce burdens on end users when collecting data, along with simplifying interpreting the results and providing prescriptive, guidance to support point-in-time actions.
Personalization & Engagement:
Lastly, AI can be immensely useful in helping to curate a more intuitive, engaging, personal and adaptive user experience within a given EHS software application to account for dynamic changes in an user’s preferences, interests, decisions and action history. Like algorithms within social media and ecommerce platforms that analyze past behavior and prompt users with options, AI offers similar potential to optimize employee engagement in EHS leading to stronger culture and performance.
Some examples of use cases that fit into each of these sections were discussed within the recent webinar, free to watch on-demand now! In the webinar, Cority’s experts also interrogated a recent Verdantix survey on EHS budgets, priorities, and technology preferences which found that analyzing leading indicators to predict safety risks is the highest priority use case for EHS professionals.
What about Generative AI?
With the rising popularity of tools like ChatGPT, EHS professionals are openly questioning the potential applicability of Generative AI technology in the EHS arena.
Generative AI is a broad term describing artificial intelligence technologies capable of generating text, images, videos, or other data using generative models, often in response to prompts”. Generative AI models typically learn the patterns and structure of their input training data and then generate new data that has similar characteristics. For example, it would be possible for AI to learn which types of incidents are most interesting to an EHS professional by examining existing incident reports based on repeat querying and training.
The Risks of Generative AI in EHS
While Generative AI offers incredible potential for multiple EHS use cases, it brings with it considerable risk if left unmitigated.
For example, we could ask an AI assistant to summarize 10 incident reports. This is an open request and doesn’t give AI much specificity, leaving room for error; it could misinterpret data, focus on areas the EHS professional isn’t interested in, or (at worst) begin hallucinating data. However, an EHS professional who is looking to augment their processes with AI would instead ask AI to summarize the same 10 incident reports to identify specific commonalities or differences. By asking better questions, Generative AI has more chance of delivering something useful.
Open-source tools present other issues for us to consider. If tools and data sources are not effectively regulated, the insights presented can easily be compromised. For example, a great use case for EHS professionals would be to benchmark their data against industry averages, goals set by governing bodies, or other institutions. Some of these data sources are public information, but unless it is vetted and proven legitimate, users could pull in false information. Let’s say an organization is using ChatGPT to pull medical data from across the web to provide information on respiratory infection rates – a great idea in principle, but it opens medical professionals open to pulling false information which could lead to compliance issues.
What does the future of AI hold for EHS?
The journey of AI within EHS is still unfolding– but there are plenty of opportunities and ideas to explore. Education, exploration, and investment is required to make sure the technology is right – and that organizations and professionals are ready for it!
To learn more about Cority’s approach toward AI, watch our on-demand webinar, “Fantastic Voyage: Navigating the Growing Opportunities of AI for EHS Excellence”, or reach out to one of our subject matter experts to explore what AI can offer your EHS performance, and how Cority can assist you at every step along the way.
