Human-machine teaming and the imminent AI revolution

As hardware and machine learning algorithms continue to improve at impressive speed, AI-powered predictive analytics has also started to boost the way we work together.

Generative AI, which transforms information to new content, can help teams to analyze and distill huge amounts of both historical and incoming data, identify patterns and trends, and provide insights that can inform apt in-the-moment team decisions.

But how? And what about the AI’s limitations and challenges?

AI teamwork makes the dream work

From labelling objects, animals and faces, translating text, and transcribing speech, well-trained AI systems have become capable of supporting knowledge-intensive teamwork through answers and suggestions that in addition are remembered by the system.

Sure, the AI system only “knows” facts and their relations through what it has been told, but its non-sentient bullshit is so good, it can actually remember what you said and conduct what feels like a fruitful conversation.

Combine its vast “experience” from billions of parameters with an advanced calculator, (you know, since machine learning can only give you its best guess), and you have a fairly mindful team member who is on the ball, and who is very skilled at things like math and software programming.

Five challenges and limitations

With generative AI on the rise, organizations that use it to innovate how teams work, will likely gain a competitive advantage.

However, it is paramount to address the limitations and challenges surrounding AI in order to ensure it is used both effectively and ethically. The main challenges are:

• Bias in the data: Garbage in, garbage out. Even the best AI systems make mistakes, despite them sounding more certain than a salesman selling sand in Sahara. If a system is trained on biased or incomplete data, it can lead to unfair or inaccurate results. This is a particular concern in high-risk, changing situations, where AI can lead to costly decisions.
• Accountability: It can be difficult to assign accountability for AI’s errors. Self-driving cars, anyone? These issues can create legal and ethical challenges for everyone involved.
• Transparency: Sophisticated AI systems are complex and opaque, making it difficult for both creators and users alike to understand how the systems arrive at their results. This lack of transparency makes it difficult for users to trust and accept AI-generated suggestions and decisions, with reduced team effectiveness as a result.
• Adoption by end-users: Be it due to change or ethics, many people are hesitant to use AI systems and may even be resistant to adopt them in their work. This can create challenges in implementing AI solutions and getting teams to use them as intended.
• Technical limitations: AI only imitates human expertise and creativity to a certain degree. This issue limits its effectiveness, particularly in situations where human expertise or plain common sense is significantly superior in tackling a problem.

With great power, comes great potential

How should you approach these issues as an organization as AI is fast becoming the general-purpose technology of our time?

The potential of AI will only be fulfilled if the technology is thoughtfully integrated into the social fabrics of your teams. Organizations should therefore use a step-by-step approach rather than a born-again adoption. Attention should be on augmenting the existing human capabilities. Who knows where you will end up in the process?

To start with, your team members should develop their AI literacy. This includes an understanding of how AI (primarily machine learning) arrives at analytical outputs and how to best integrate the strengths of an AI system in the team processes.

At the end of the day, it is the team members who do (and should do) the intuitive and global assessment of the situation at hand. At the same time, AI has the potential to provide teams with powerful tools and systems than can support and augment their decision-making and problem-solving abilities.

With the right governance in place, AI systems hold great potential in assisting and inspiring the endeavors of humans working together.

The author, Morten Juel Hansen, is also a PhD Candidate at BI's Department of Leadership and Organizational Behaviour.

References:

This article is featured in the upcoming 2023 issue of BI's Leadership Magazine.

Davenport, T. H., & Ronanki, R. (2018). Artificial Intelligence for the Real World. Harvard Business Review, 96(1), 108-116. Retrieved from <Go to ISI>://WOS:000418785600024

Jarrahi, M. H., Lutz, C., & Newlands, G. (2022). Artificial intelligence, human intelligence and hybrid intelligence based on mutual augmentation. Big Data & Society, 9(2), 20539517221142824. doi:10.1177/20539517221142824

Mollick, E. (2022). ChatGPT Is a Tipping Point for AI. Harvard Business Review. Retrieved from https://hbr.org/2022/12/chatgpt-is-a-tipping-point-for-ai?ab=hero-main-text

Stowers, K., Brady, L. L., MacLellan, C., Wohleber, R., & Salas, E. (2021). Improving Teamwork Competencies in Human-Machine Teams: Perspectives from Team Science. Frontiers in Psychology, 12(1669). doi:10.3389/fpsyg.2021.590290