Why Every AI Company Needs a Coach

We can all use a bit of support to help us reach our desired goals.

AI is a rapidly evolving field, with new boundaries being crossed all the time.  As each new boundary is crossed, it is crucial to develop safe and responsible products that comply with existing laws and anticipate future regulations.  Just like humans, it can be challenging to correct an AI program's bad habits once it has been trained, especially as it ages and these habits become ingrained.  Companies simultaneously need to innovate on both the technical side of AI and the ethical and safety side of AI. This is especially true in the San Francisco Bay Area as a hub of both AI and strong, employee-friendly anti-discrimination laws. 

Coaching excels at helping clients at the cutting edge.  Consulting, mentoring, or other subject-matter experts require long track records to provide advice about what to do next. However, professional coaches provide process expertise, so they empower clients to figure out what to do themselves and do not need all of the answers.  Given AI's novelty, consultants and experts may not have all the answers because no one has done what AI aims to achieve.

A coach can accelerate growth in a sustainable way for AI companies in the following ways:

1. Strengthen Team Collaboration and Cross-Functional Communication

AI's novelty can lead to confusion about responsibilities within different organizational departments.  Organizational departments assume that another department will handle certain issues related to AI when that part of the organization thinks it is not their responsibility.  For example, engineers or data scientists might believe that AI bias is the responsibility of someone from HR, DEI, compliance, or the legal department, whereas one of those departments might believe the opposite.  Clarifying roles and responsibilities through conversations can prevent internal competition and overlooked issues, giving competitors an edge. 

Having a conversation about whose roles it is to do what regarding AI would help clarify the roles and responsibilities of every stakeholder (every employee and contractor).  That way, the company is not wasting time competing internally instead of externally or in the alternative, overlooking key issues since no one thinks about an issue since company leaders assigned no one the issue, giving competitors an edge who are thinking about these issues and making a better final product.

Team or organizational coaching can break down silos.

Instead, with some communication, different organizational functions can work together to make a stronger team and final product, as well as break down organizational silos.  As a result, effective communication breaks down silos not just for this one project but also for the improvement of company culture.  External team or organizational coaches have an additional benefit of enhancing that teamwork because they are not beholden to office politics and can provide an outsider’s fresh perspective when those inside an organization may be too close to the issue.  While the company may be able to turn out a product without a coach, it may make huge mistakes or errors because the company may not be looking at the big picture or different parts of the company may not be sharing information with each other that each needs to have the best outcome.

2.      Strategize How to Mitigate Bias in AI Development

Biased answers may turn customers to a competitor's products that give more relevant outputs.

An AI that makes biased decisions or provides biased outputs can not only create legal problems for the team that created it but also give a useless or less useful result for clients.  As a result, customers may go to a competitor’s product with a better answer or output that does not provide a biased answer or gives a more relevant answer because it considers inclusion or other user-specific information.  AI products may also lead to legal problems for users who rely on the biased decisions made by AI, which could lead to discrimination lawsuits, especially with AI-enhanced HR software that helps with hiring decisions.  It can be hard to overhaul an AI product once an AI model has already been created and trained.  A company may as well do the best it can up front instead of patching up a fundamental error later.  Once companies inevitably try to fix an issue that they could have prevented with some forethought or planning with the help of a coach, they lose precious time in the AI race against their competition and resources when a company diverts resources from product development toward reactionary legal expenses. 

3.     Detect Bias with AI Coaching

Bias detection is a crucial aspect of developing fair and effective AI systems.  Detecting bias involves identifying and addressing any prejudices or unfair tendencies in the data used to train AI models.  This process ensures that AI systems make decisions based on accurate and representative information rather than perpetuating existing biases.  With a coach, developers can reflect on their thought processes and think through how they can effectively build bias detection systems.

4.     Develop Policy on How AI Should Be Used and Developed

With anything new, policies about how a company may use or develop a product may play catch up with technological development.  With an organizational coach, company leadership (starting in the C-suite) can start to think through what guardrails they want to put in place with the input of various functions of the company to mitigate the risk of harm while capitalizing on the advantages of a new technology. 

5.      Achieve or Sustain a Competitive Advantage against Competitors

Build a better product by creating a healthy workplace culture that facilitates productivity and innovation.

By creating a positive work culture and thinking through potential pitfalls with a team or organizational coach, AI companies can gain or maintain a crucial competitive advantage against their competitors.  Other companies that do not have a coach to help them navigate issues related to bias may be bogged down by discrimination lawsuits and internal team conflict or miscommunications.  However, those with an AI bias coach can focus on the work they love best and devote resources back into their company growth instead of destructive and expensive conflict.  A study by Kanze, Conley, and Higgins (2020) found that startups that focus on fast growth without thoughtful consideration tend to break things as well, leading to more workplace discrimination lawsuits but failing to stop long enough to follow the law.  A professional coach can empower clients to take time to reflect and think about their actions and beliefs before acting.  At the same time, if AI companies are doing the same not only for their own work cultures but also related to their products, they also may face suits related to the bias in the AI they are creating.  At the same time, while other companies without coaches might be arguing over who should be doing what, walking over each other because there has been no coordination or communication, or underperforming, companies that work with an organizational or team coach can facilitate companies into working well together, improve communication, and foster creativity and a positive work environment.  Therefore, it is vital for companies to be clear about their AI strategy.