ChatGPT: Minimise Risk & Maximise Opportunity
Whilst I don’t wish to be, yet another person writing about ChatGPT, I feel it’s important to discuss how organisations can minimise risks and maximise opportunities with AI. This article focuses on one initiative you can implement today… educating your team!
The obvious education
ChatGPT, or any AI content generator, comes with risks but these risks may not be evident to your team. Some of these risks include:
Inaccuracies: ChatGPT will not always produce accurate responses. The software developers using ChatGPT to write code can attest to this. Any output needs to be checked for accuracy. It may also have gaps. For example, if you request ChatGPT write a contract clause, it may initially look reasonable but without context, it may leave you open to risks. If your team members are using ChatGPT, you must ensure they are checking for; (1) accuracy; (2) context relevance; and (3) strength of the content.
Tone and voice: While you can request ChatGPT adjust tone and voice, it will never mimic you perfectly. A letter of advice or educational article should be in “one voice” representing the individual and the company. Advise your team they can access ChatGPT as a starting point for articles, however, they must adjust the content to the appropriate tone and voice.
Bias: ChatGPT attempts to mitigate bias but acknowledges bias can still exist in the data. The following quote is from ChatGPT:
“It’s important to note that bias can still exist in the data that ChatGPT accesses online, and the model’s responses may not always be perfect or completely free from bias. It’s important for users to critically evaluate the information provided by ChatGPT, as with any other source of information, and to seek out additional perspectives and sources.”
No understanding of thought process: If you’re using AI software, you will never truly understand how it came to a particular outcome. This will become more evident as models become more advanced and sophisticated. We can only understand the inputs that the AI model has considered when generating its output. For your team, ensure they do their own investigations. Understanding the “why” behind any topic leads to deeper insight.
The “not so obvious” education
Some of the education I recommend extends beyond what some may perceive as an organisation’s responsibility. It could be argued the items below are an individual’s responsibility in their own personal development journey, however, these concerns impact organisational performance. I, therefore, recommend organisations consider the following:
Skillsets and fulfillment: Some software companies are using ChatGPT to develop code. These companies acknowledge the outputs are not always accurate and therefore ensure all code is checked. When software developers create code, it can be a very creative and fulfilling process. If these developers are now simply checking the code, the skillset shifts from “doing” to “checking”. A QA role is not as fulfilling as a creating role. For many organisations, this won’t be an immediate problem, it will, however, be a slow-burn over time. Organisations should consider this as they implement technology that results in skillsets pivoting.
Resilience and focus: Some argue that younger generations have less resilience. There are several reasons for this argument but regardless of generational differences, AI can negatively impact our resilience and focus. “AI negatively impacts resilience and focus by contributing to a culture of constant distraction and dependence on technology” (ChatGPT). Algorithms are designed to maximise engagement and can distract individuals from meaningful productive activities. Additionally, algorithms can contribute to a sense of information overload which slows down productivity and prevents quick decision-making. We, therefore, need to equip our teams with tools to increase resilience and improve focus.
Comfortable being uncomfortable: Change is inevitable with the increasing integration of AI into our lives. To adapt and thrive in this changing environment, employees need to become more comfortable with being uncomfortable and embrace a growth mindset. Again, organisations can invest in team training to adopt this mindset. Additionally, you can create a culture that fosters a growth mindset.
Curiosity and learning: As new technologies emerge, and as people have “answers at their fingertips”, it’s integral that we encourage curiosity and learning. Cultivating curiosity ensures employees challenge assumptions and dig deeper into their understanding.
Your to-do list
Training: Revisit your training programs. Ensure your training covers these items and ensure you’re constantly reassessing training. As technology evolves, your training also needs to evolve.
Culture: Assess your current culture. Ensure; (1) initiatives are conducive to your desired culture; and (2) your desired culture incorporates growth, curiosity, learning and resilience.
Policies: Develop a policy on the use of ChatGPT in your organisation. Ensure your team members don’t input confidential business or client data to ChatGPT, and ensure they’re analysing the responses for accuracy and context relevance.
If you’re unsure where to start with these initiatives, reach out!
ChatGPT, AI and other technologies can bring enormous benefits in terms of efficiencies, however, these technologies also come with risks. The potential downside includes data privacy issues, bias, a lack of resilience, and attention issues. It’s important to approach AI with a critical eye and willingness to ask questions and challenge assumptions. By educating your team, you can maximise the use of technologies and minimise the potential downsides.