Making the AI Communications Upgrade
Making the AI Communications Upgrade - Nick Shah

Making the AI Communications Upgrade

The concept of emotional intelligence and its role in leadership has always intrigued me.?

It’s often classified as a skill, and I agree, as it's something that can be improved upon. Emotional intelligence?requires being savvy about our own feelings (and how we express them)?as well as those of others. In some ways, the definition of emotional intelligence can be reduced to “the art of understanding people.”?

But what’s really been surprising to me over the past year is how AI systems, which don’t have emotions and cannot truly think (nor understand human behavior at this stage of development), can not only demonstrate traits of emotional intelligence but actually help us improve our own.??

Consider research published in the Proceedings of the National Academy of Sciences, which demonstrated how AI-generated messages could make people feel more effectively heard than human-generated messages on average.??

These AI responses were also deemed more accurate and made people feel more understood, even increasing the perceived connection between the responder and original sender.?

That is, until the sender learned it came from an AI. At that point, results flipped. Understandably, the subjects then felt less heard.???

This makes sense, as the responses were only mimicking understanding, however effectively. This of course is not full emotional intelligence but only the illusion, or mimicry, of it.?

But there are other ways we can harness AI to help us hone this precious skill.?

In today’s newsletter, I look at this perplexing capacity for AI to help us with our own emotional intelligence, and consider ways it can help us communicate, and lead, more effectively.??

??

Learning in a Virtual Mirror: AI-Powered Coaching for Leaders?

Leadership requires a lot of communication. This means being able to find and ask the right questions and, of course, listen to the answers you’re given.??

It also means dealing with conflict.?

But every conflict is unique, and learning these skills can be very difficult. I’ve been using AI for years in my communications with these goals in mind, both for aiding in phrasing and to assess my own approach.?

And while I had not yet done it in such a systematic way, I could really appreciate the findings shared by the Harvard Business Review in February (Research: How AI Helped Executives Improve Communication).?

In this study, researchers attempted to apply an LLM in a traditional trainer’s role, to oversee executive communications and provide analysis and recommendations.??

First, they picked a framework for evaluation, settling on John Heron’s six categories of intervention:?

  1. Prescriptive (offering advice)?
  2. Informative (sharing insights)
  3. Catalytic (prompting self-discovery)?
  4. Cathartic (helping process emotions)?
  5. Supportive (offering encouragement)
  6. Confronting (challenging behavior)?

?

Next, they created a custom tool (based on GPT-4) and put the process to the test with 167 executives from around the world and across industries.??

After pairing execs with the AI system and a human overseer for control and documentation, they?engaged in actual conversations from real experience, across a variety of settings, as the system observed.?

The tool was then asked to provide assessments. Specifically, the executives used prompts like:?

  1. What is my coaching style from Heron’s list?
  2. What approaches do I use most??
  3. What was effective and went well??
  4. What patterns do I use when asking questions??
  5. How can I improve my skills??
  6. How much talking/listening was done??
  7. What do you see about my style that you think I don’t see??

?

The results from AI were compared with those from the human overseers, and in the end the executives gave their own feedback.??

In general, the AI was far more effective the more specific the prompts got. It did great at picking from Heron’s approaches and even discovered things that surprised the researchers.??

And of course, there were instances of hallucination and vagueness.??

Overall, they evaluated this test of AI coaching for executives as follows:?

  • 55%: Found it both useful and novel, including some impressed by the granular detail, suitable and new conclusions, and effectiveness with tone?

  • 30%: Found it useful, but in ways they already expected?

  • 10%: Found it irritating, off-target, or unhelpful?

  • 5%: Thought it neither surprising nor useful (too general, too obvious, misaligned)?

?

Overall, 85% of the applicants (or some 140 of the 167) got value from this process of using an autocorrect machine without emotions to improve their own emotional intelligence.???

As improbable as it seems, this has also been my own experience.??

And while the researchers recommended anchoring the feedback for the AI (such as by using Heron’s framework), they found many of the executives preferred using AI as a communication coach to humans, as it was easier, more direct, more available, and less awkward.??

Still, the study also found their system (lacking real-world understanding) missed cultural differences and unspoken cues, making it limited in ways human trainers are not. But nevertheless, as a tool, or an option that’s ready for use when and how you need, it was found to have undeniable value in this arena.?

??

Using AI in Leadership Communication?

Beyond this kind of live coaching or training approach, I find AI to be highly useful at recognizing patterns and devising quick alternatives. And while the direct wording may not often be useful, it can absolutely help me notice things I may be ignoring.??

It can tell us if we’re being too vague, for example, or doing too much of the talking. It is a safe way to rehearse difficult scenarios.?

And a side benefit can come from trying to explain a situation sufficiently (and succinctly) enough for a prompt: it helps to clarify our own understanding and realize areas where we may need more detail or focus.??

Chatbots are increasingly effective at providing variations in tone, suggesting word alternatives, restating the challenging, or even pointing out our own unexpected assumptions.??

??

AI and Emotional Intelligence: An Unexpected Boost?

If you want to see this in action, I recommend taking an approach like the one in the research above.??

Think back on a challenging situation or interaction and explain it in a few hundred words.??

Also provide a sample of communications (suitably de-identified, of course), and then query the LLM. Ask for advice and evaluation, alternatives, and characterization of your own thinking and approaches.??

Something like this is quick and easy, has no risk (again, de-identified), and, in my experience, often yields surprising results.??

The current leading chatbots are quite good at exchange, as well, if you want to practice or role play a potential situation or just trigger new ideas or outcomes.?


Having AI-Enhanced Workplace Conversations?

Depending on what system you use, you likely have AI summaries of your team meetings being made already. And while these have been in the news for sometimes embarrassing reasons (divulging too much, picking up chatter after the meeting, etc.), they can also include surprising insights, especially in meetings you are running yourself.??

For use with emotional intelligence, it can absolutely improve your capacity to provide support to those in need (though of course this isn’t helping you with your own empathy unless you are learning from it).??

It’s also highly useful in preparing for conflict mediation, by offering alternatives in messaging, recognizing potential triggers, and suggesting a variety of potential approaches and responses.?


For Cross-Cultural Communication?

At PTP, we’ve seen the benefits AI can bring in helping offshore or nearshore workers bridge cultural differences. How about our own??

In all of these exercises with coaching and reflection, try asking the AI to reconsider the message from another cultural perspective.??

While AI systems can sometimes be guilty of perpetrating stereotypes, I’ve also found them surprisingly helpful with pointing these nuances out, ironically enough.?

For language work, taking advantage of multimodal systems, like ChatGPT’s Voice Mode, can be a big help. The big AI companies have long touted this, from Google DeepMind’s early applications of Gemini systems in translation to OpenAI promoting their ability to bridge research gaps.??

And while it’s had fits and starts, current systems can be highly effective for practicing exchanges in other languages.??

??

The Reality Check?

Okay, so now you may be thinking this is one of those rosy optimistic takes on the endless virtues of chatbots.??

No. AI still has its limits. For one thing, it continues to hallucinate, and I wonder with the study results above how many of those handful of executives who found it irritating ran up against a particularly egregious mistake.??

Using AI in these ways isn’t perfect and can be prone to being too vague and general. As in everything, the more specific the prompt, the more limited the framework or system for evaluation, the better. AI systems seem to be thriving most in the sciences at present, and I believe the structure that’s provided there has a lot to do with it.??

The more structure we can give it for our personal use, the more useful the responses are as well.?

And of course, as I led with, AI communication may only make us feel heard, when we may need to actually be heard for corrective action. When a human is not in the loop, it can be a potentially disturbing illusion.??

The prosaic, patterned speech produced by AI can also sometimes have the inverse effect and suggest that a speaker doesn’t really care and isn’t feeling for the situation at hand.??

Ultimately, it’s essential that AI is being used to aid my human judgment and improve my communication, rather than replacing it.??


Conclusion?

The MIT Technology Review recently ran an article profiling AI and human relationships. It considered personal stories of people using these chatbots to ease their loneliness, give partner and parenting advice, and improve language skills.??

And while some of these use cases can be problematic in certain contexts, using AI in communications can also bring a powerful, additional step of insight and personal attention to situations where it might otherwise be lacking.?

At PTP we work closely with AI beyond such a personal context, providing great people with the talent necessary for your organization to realize your own AI-powered innovation. We try to stay realistic on where it is and what it can really achieve, even as the frontier keeps moving.??

But by helping leaders recognize blind spots or improve awareness and refine messaging, it can also serve as a coach or surprisingly insightful assistant.??

If you’re not yet taking advantage of what AI has to offer in this area, I recommend you give it a try.?

??

References?

The AI relationship revolution is already here, MIT Technology Review?

Research: How AI Helped Executives Improve Communication, Harvard Business Review?

Helping the Client: A Create Practical Guide, John Heron Archive?

Using Artificial Intelligence to Enhance Coaching – A new study, Oxford Review?

AI can help people feel heard, but an AI label diminishes this impact, PNAS?

??

??

Check out other articles from PTP? — From Our CEO and on AI??

Get the latest updates on recruiting trends, job market, and IT, and expert advice on hiring and job seeking at The PTP Report.??

Peter E.

Helping SMEs automate and scale their operations with seamless tools, while sharing my journey in system automation and entrepreneurship

1 周

It’s fascinating how AI can analyze language and sentiment better than we often can in real time. But does relying on AI for emotional intelligence risk losing our own ability to read the room?

回复

要查看或添加评论,请登录

Nick Shah的更多文章

  • Solving the AI Data Shortage Before It Is a Crisis

    Solving the AI Data Shortage Before It Is a Crisis

    According to a new Dun & Bradstreet survey, nearly 90% of businesses are adopting AI into their operations now, with…

    2 条评论
  • How AI Is Shifting Jobs

    How AI Is Shifting Jobs

    Consider this: in the US, nearly 20% of workers hired today have job titles that did not exist 25 years ago. They will…

    6 条评论
  • Power Struggle: A Look at Data Centers and AI Growth in 2025

    Power Struggle: A Look at Data Centers and AI Growth in 2025

    “DeepSeek” is truly a magic word in tech this week. The Guardian’s Tuesday newsletter led: How an unknown Chinese…

    2 条评论
  • Digital Transformation in 2025: What Does That Mean?

    Digital Transformation in 2025: What Does That Mean?

    It’s 2025: what does “digital transformation” mean to you? Once it meant moving from paper to online or automating…

    4 条评论
  • Cybersecurity: The Human Link

    Cybersecurity: The Human Link

    Unless your company is in the security business, it’s likely many employees view cybersecurity as someone else’s job…

    2 条评论
  • Managing with AI Uncertainty

    Managing with AI Uncertainty

    This is a time of high uncertainty. Charles Schwab’s chief investment strategist Liz Ann Sonders tells Yahoo Finance’s…

  • Complacent Innovation: A Cautionary Tale

    Complacent Innovation: A Cautionary Tale

    In the news from November 1: NVIDIA replaces Intel Corporation on the Dow Jones Industrial Average. And looking at this…

    1 条评论
  • My Thoughts on Founder Mode

    My Thoughts on Founder Mode

    What makes great leadership? Most would agree on certain traits and behaviors like direction, devotion, drive…

    2 条评论
  • On Leading Distributed Tech Teams

    On Leading Distributed Tech Teams

    Everyone who managed their way through COVID-19 got a crash-course on mixed work environments, and the results…

    1 条评论
  • Nurturing Your Tech Talent Pipeline

    Nurturing Your Tech Talent Pipeline

    There’s an old truism called the Pareto principle (aka the 80/20 rule, the law of the vital few, etc.) that gets…

    4 条评论

社区洞察