Mentoring in a hybrid world - Human + Machine?

Connections - Midjourney - Antony Slumbers

Much is made of the problem hybrid, remote and distributed working causes for most companies around mentoring of new and junior staff members.

Often the response is that ‘we need to get people back in the office’’.

But is that the right response. Or the only option?

Well, it is ‘a’ response. But one that goes against the grain of how ‘normalcy’ is developing around the way we work. Put simply we cannot rely on mentor and mentee being in the same place.

Or at least not all the time.

So perhaps we need a ‘hybrid’ way of mentoring, as well as working?

Which brings us to the question ‘how can Generative AI help in the process of mentoring?'

I think, it can. A lot.

By following the outline of this proposed slide deck:

Slide 1 - Title: AI-Assisted Mentoring Program

Slide 2 - Purpose: Provide more frequent mentoring conversations through an AI assistant. Maintain human manager oversight via prompts and monthly check-ins. Goal is to enhance mentee development and growth.

Slide 3 - Mentor Preparation:

• Develop a series of prompts that are open-ended, thought provoking questions tailored to mentee. These would be prepared by individual mentors and become part of a ‘Prompt Library’.

• Schedule 30-60 min monthly in-person mentoring check-ins

• Review AI session summary reports before check-ins

Slide 4 - AI Mentoring Process:

• Manager provides prompts

• AI has regular mentoring conversations guided by prompts

• AI sends summary report to manager prior to their check-in

• Manager has in-person check-in equipped with insights from report

• Manager refines prompts based on insights and progress

Slide 5 - Sample AI Summary Report [*see below for a sample report]

Slide 6 - Action Steps:

• Managers create initial prompts for mentees

• Schedule first in-person kick-off mentoring session

• Begin regular AI mentoring conversations

• Provide feedback on experience and report format

Ultimately the aim would be to solve the Bloom Sigma Problem - as explained here:

‘Bloom's 2 sigma problem refers to an educational study by Benjamin Bloom in 1984 that found students who received one-on-one tutoring performed two standard deviations better than students in a traditional classroom lecture setting. This highlighted the significant benefits of personalised instruction and attention.

While AI mentoring has many advantages, completely replicating the effectiveness of human one-on-one tutoring is challenging. 

Some key considerations:

• Human tutors have deep expertise, wisdom and intuition gained from years of study and experience in their field that AI cannot fully match.

• The interpersonal connection between a student and human tutor can provide additional motivation, empathy and support.

• AI has some limitations in understanding full context and nuanced human communication/behaviour during complex discussions.

• The datasets used to train AI mentors may lack diversity leading to biases. Humans are better able to adapt across different personalities and learning styles.

That said, AI mentoring can get closer to the 2 sigma effect through methods like:

• Leveraging large datasets to train conversational ability and mentoring style.

• Incorporating multi-modal feedback - verbal, visual, emotional - during discussions.

• Maintaining human oversight and input to enhance AI mentor prompts over time.

• Tracking detailed metrics on learning improvements to refine approach.

• Utilising personalisation algorithms to tailor mentoring to individual needs.

While not a complete solution yet, AI mentoring can still dramatically expand access to the benefits of one-on-one instruction. With continual improvements, bridging Bloom's 2 sigma gap could become more feasible over time.’

So, in summary, it’s a fair amount of work to set up but I believe this could be an enormously powerful tool for companies to adopt. The development of the ‘Prompt Library’ is key. Each specialist within a company would need to tailor just the right prompts to catalyse an excellent learning experience.

Combined with regular in-person mentoring sessions we might be able to get to the best of both worlds.

Remember the 2 sigma effect. That is a goal worth pursuing.

What do you think?

-----------------------------

*Sample Report provided to each mentor prior to monthly in-person sessions:

Mentee: Sarah James

Mentor: Amanda Smith

Reporting Period: January 1 - 31

Topics Covered:

• Presentation skills development - gave practice run-through of upcoming quarterly presentation and received AI feedback

• Time management challenges - discussed struggles with prioritisation and setting boundaries on availability

• Career advancement goals - interested in leading a cross-functional initiative this year

Key Insights:

• Wants to work on voice projection and limiting verbal fillers during presentations

• Has difficulty declining or delegating extra tasks from peers

• Would like to gain experience managing a team for advancement

Progress Made:

• Practiced presentation 3 times and improved confidence each time

• Implemented 30-minute focus blocks in calendar to limit distractions

• Reached out to cross-functional counterparts to discuss initiative ideas

AI Mentoring Feedback:

• Appreciates consistent opportunity to practice presenting and get unbiased feedback

• Found time management tips helpful but still struggles with execution

• Suggest focusing next 1-1 on assembling team and pitch for leading potential initiative

Recommendations:

• Provide input on body language and slide design to elevate presence

• Discuss prioritisation frameworks to identify key tasks vs. "nice-to-haves"

• Share insights on navigating cross-collaboration and tips for leadership