If feedback is a gift, can AI help us give more
02 January, 2024
By Amina Crooks
We are often told not to care what others think and that one of the benefits of aging is leaning less on the crutch of external validation…but — yes, you knew there was a “but” coming — we cannot help ourselves. Wanting feedback and more specifically constructive feedback is an innate and visceral human trait. Why? Well, positive feedback makes us feel good, it gives us affirmation and recognition for our efforts. And, negative feedback, while not so warm and fuzzy, is appreciated in our attempts to better ourselves and attain our aspirations.
Regardless of whether we like feedback or not, humans need feedback because like all biological systems, feedback loops are designed to maintain homeostasis. Feedback loops correct any deviations from homeostasis using two loops — positive and negative feedback loops. The positive loop is easy, the messenger and the recipent both feel great, pat on both backs. The negative loop, on the other hand, not so much.
The dilemma is that most of us don’t know how to deliver negative feedback in an effective and constructive way, so we hold back.
And this really is a tragedy, because one could argue that our blind spots and failures are where we have the most room to grow. Research has shown that while we desire feedback, “people consistently underestimate others’ desire for constructive feedback and therefore don’t provide it” (1).
Seems odd right? If we, ourselves, desire constructive feedback and we know constructive feedback helps others learn and perform better — why do we avoid giving it to others? Could AI help, and if so, how?
To examine where AI could be of assistance, let’s examine where we as humans fall over. I have identified 3 key reasons why people do not provide feedback and where AI may or may not play a role.
#1 We don’t fully recognise the potential of our input to improve others’ outcomes.
For me, this is probably the most important one for humans to start stepping up to. AI will be great at recognising the potential of its input to improving outcomes. AI will be consistent in always providing feedback, whether we like it or not and it will:
go beyond analysing data to uncover patterns, trends and correlations;
generate personalised coaching or training to individuals or groups;
monitor outputs quickly to derive recommendations, continuous improvements and actionable insights; and
if data inputs are reliable and accurate then it is possible that AI-generated digital coaching could remove existing human biases such as gender bias, perception bias, or affinity bias (gravitating towards people who appear the same as you).
But, there are still ways that humans can add immense value and provide feedback to others. Firstly, humans observe what is not always measured, especially when it comes to emotional intelligence and soft skills. Secondly, AI does not possess some of the critical qualities to look for in a human life coach, such as being passionate, empowering, authentic, motivating, inspiring and importantly, walking the talk. We have experiences to lean on that enable us to give real, lived through advice.
#2 Fear of negative outcomes, such as the other person feeling embarrassed or upset.
Let’s be honest, most of us are crappy at giving constructive feedback. Why? Because constructive feedback implies it has both positive and negative loops and as mentioned above, the negative loop is not a easy one to communicate. Giving constructive feedback requires IQ and EQ and often a set of other skills like patience, diplomacy and carefully selected words. While people may want constructive feedback, we are all wired to minimise danger and maximise reward, which effectively means we are hoping for positive and not negative feedback. Evolutionarily, this wiring is designed to keep us safe from anything perceived as a threat. When we feel threatened, we shut down. That does not mean negative feedback isn’t gold. One of the mantras of the 21st century is surely how great it is to fail. “Failure is success in progress” [Albert Einstein].
Image taken Beyond Blue Consulting: https://www.bluebeyondconsulting.com/blog/this-is-your-brain-on-feedback-how-understanding-a-little-brain-science-can-make-a-big-difference-in-your-next-feedback-conversation/
Our wiring for minimising danger and maximising reward translates into this - for negative feedback to be constructive, it needs to be delivered as a reward and not a danger.
How to effectively deliver constructive feedback is a whole pantheon of literature , which I will not go into here, aside to point out one of my favourites — the shit sandwich. A ‘SHIT SANDWICH’ is a framework for giving some bad news, advice, or a rejection of someone’s idea. It plays on that wiring discussed earlier, by sandwiching at the beginning and at the end a “reward” (a positive comment), with the negative feedback (the part of the feedback that is “shitty” to hear), being delivered in the middle.
While current LLMs are often criticised for going off topic and not sounding human, future versions will be better trained and have specialised training to sound like certain people’s style. So in the foreseeable future, an AI assisted coach will help us to craft difficult conversations and suggest ways to communicate difficult news to people that is both personalised to our style of communication and also caters to the recipient’s preference and most effective means for receiving information.
There is also a burgeoning area of knowledge around digital nudging which capitalises on behavioural science and information technology and is providing tools for designers to change consumer’s behaviour and create habits.
#3 We underestimate others’ desire for such feedback.
Let’s unpack this one a bit more. It’s likely that knowing how to deliver constructive feedback is not the only barrier. We may simply lack awareness of how much our feedback matters to another person. This is a circular argument and presents the importance of feedback loops in feedback itself — meaning when we give feedback, we need feedback on whether it was impactful to ensure we continue to see value in giving it. How often do employees feel deliver feedback on the content or style of feedback that they receive from their managers?
It is also possible that we are aware of the impact of our feedback but we simply do not have the time or care enough to give it. All of which is why — feedback is a gift. Giving constructive feedback requires two critical elements that are demanding on the part of the recipient:
Time — which we all know is a scarce and limited resource. Good feedback requires thinking, observing and listening to the recipient to uncover what feedback will truly be valuable to them. Giving feedback requires a certain amount of kindness.
Overcoming our own feelings about ourselves. Giving feedback requires a certain level of confidence. We need to overcome our inner critic which makes us second-guess ourselves and underestimate the beneficial effect we can have on others.
Not sure if the above means that there may currently be an imbalance in the feedback out there with a excess of feedback being given by selfish narcissists, but let’s keep moving.
In both of the above, time and confidence, AI could assist, if computers were given the appropriate data. With the right data, AI could help with the discovery and delivery of feedback which could save time and hence help us to give more valuable feedback. If recipients provided truthful and honest feedback on the quality and impact of the feedback they received, it could help people determine where they need to lean in more; how, when and how they need to give feedback to have a positive impact.
Inconclusion. A critical part to our personal growth and ability to impact others positively is learning how to deliver constructive feedback effectively, which includes both positive and negative feedback loops.
Key takeaway — AI will have a useful role to play in digital coaching and training; but humans are not totally off the hook. Be generous with the delivery and the frequency in which you give feedback and keep seeking feedback on how to give feedback more effectively.
Goes without saying — feedback on this post would be greatly appreciated!
