Artificial intelligence is a vast and complex topic. Although there are many definitions of artificial intelligence, one of the most straightforward is: “the capability of computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making” (Russell and Norvig, 2020).
Various standard machine learning (ML) techniques account for the majority of applied statistical learning efforts within the geosciences. Increasingly, large language models are being utilized in the production of reports and other technical documents. These and other techniques will collectively be referred to as “artificial intelligence”. The various models, algorithms and statistical techniques covered by the term artificial intelligence are incredibly powerful tools that have facilitated major advancements in the sciences. The tools may be commercial products “off the shelf”; they may be modified to fit specific applications, or they may be completely custom-built.
Consistent with Geoscientists Canada and the Member Regulators’ Codes of Ethics, the over-arching theme of this document is Public Safety. Regardless of their existing successes in related disciplines, responsible human oversight of these tools is essential.
The pipeline for application of any ML technique can be seen as having three main components, the first of which is Data. Collecting and curating data is often the most arduous and time-consuming task in the modeling process. The legal requirements around data provenance, acquisition, conditions of usage and security are evolving rapidly. It is incumbent upon all who “touch” data to be aware of data protection laws in all jurisdictions in which they practice. The second component of the pipeline is the application of an appropriate algorithm; the third component is the outputs which require oversight
and proper vetting. This last step is critically important and relies upon the expertise of a professional geoscientist. All three components will be addressed in this document.
Purpose
The purpose of this document is to provide an overview of the usage of artificial intelligence tools along with best practice considerations for professional geoscientists, and in so doing, to support the ethical, highly-skilled practices of Canadian Professional Geoscientists at their various career stages.
Typically, professional geoscientists will move through different stages in their career progression.
Starting with a purely technical job function, most practitioners will then undertake mixed technical/managerial roles culminating in a mainly managerial role or in a senior technical specialist position.
Those practitioners currently with advanced years of experience both technical and managerial, may include those who started their post-secondary training and careers without AI tools – and in many cases, without computers!
This dichotomy of expertise results in cohorts of geoscientists with very different skill levels related to the usage of AI and AI ecosystems. Each of these groups will have a different view of AI tools and a different understanding of the ethics behind their use. Early career professionals are less exposed to the risks and decision-making around AI outputs, whereas the advanced career professionals will be Usage of AI Tools by Professional Geoscientists in Canada taking full responsibility for their teams’ work products that have been generated with full or partial use of AI.
Ultimately, the professional geoscientist is responsible for their professional work and its outcomes.