Black Box Thinking: The Psychological Barriers to Artificial Intelligence
In this bonus episode of Lessons Not Yet Learnt, host Kiran Kachela explores the psychological barriers that stand in the way of AI adoption. While many rush to declare they are using AI, the real challenge lies in human trust, control, and connection. From the black box problem to fears about job loss, Kiran highlights why success depends on putting people first, designing for explainability, and ensuring AI feels personal and transparent.
In this episode we cover
- Why AI adoption is shaped more by psychology than technology
- The black box problem and how explainability builds trust
- Why emotionless systems face resistance and how human like connection makes a difference
- Fear of losing control and the importance of keeping humans in the loop
- Why flexibility and personalisation matter for long term adoption
- The very real fear of job loss and how roles evolve alongside AI
- CI fixes to lead with humans design for explainability and build solutions people trust
About Our Host
Kiran Kachela – CEO and Founder of CI Projects
Kiran has spent over a decade driving transformation by simplifying complexity and empowering people to deliver lasting change. Through Lessons Not Yet Learnt she shares overlooked lessons, candid insights and practical reflections to help organisations avoid repeating the same costly mistakes.
Play this episode
Read the transcript
00:00:00:00 – 00:00:34:18
Kiran Kachela
Lessons Not Yet Learned is the podcast that uncovers the costly mistakes that businesses keep making and how to finally move past them through storytelling, candid conversations, and expert perspectives. I’m on a mission to challenge outdated thinking. Bring those overlooked lessons to the surface, and drive meaningful change one story at a time.
00:00:34:18 – 00:00:38:15
Kiran Kachela
Hi. Welcome to this bonus episode of Lessons Not Yet Learned.
00:00:38:17 – 00:00:53:19
Kiran Kachela
I’m your host, Kiran Kachela. In the last episode, I touched on the most talked about technology of our time, A.I. Artificial Intelligence. And I talked about this race to be seen to be doing it.
00:00:53:19 – 00:01:04:10
Kiran Kachela
But successful adoption actually relies on something far more complex than the technology itself. It relies on human psychology.
00:01:04:10 – 00:01:11:18
Kiran Kachela
psychological barriers are often the biggest contributor to resistance in AI adoption.
00:01:11:18 – 00:01:26:06
Kiran Kachela
Everyone wants to declare that they are using AI, but there’s a huge, immense pressure to make it work. And actually, what it really relies on is people. People to understand it, people to trust it, and people to adopt it.
00:01:26:06 – 00:01:37:14
Kiran Kachela
So in this episode, I’m really keen to explore what are some of these psychological barriers that are impacting the adoption of AI and what we can actually do about it.
00:01:37:14 – 00:01:58:21
Kiran Kachela
So one of the psychological barriers to AI adoption is the black box problem. And this is this lack of explainability. People not understanding what AI is doing in the background to actually come to the answer. So because people don’t understand it and they can’t explain it, they become resistant to it.
00:01:58:23 – 00:02:10:17
Kiran Kachela
It’s a bit like if we were sat in an autonomous car and all of a sudden the car decides to just stop, we would not really trust it. We wouldn’t understand why the car had stopped.
00:02:10:17 – 00:02:26:16
Kiran Kachela
But actually, if it did explain that the car has stopped because there’s an obstruction on the road a mile further down. That would help us to understand why the car was stopping, and it would help us to trust the solution better.
00:02:26:18 – 00:02:30:05
Kiran Kachela
So explainability is really, really key.
00:02:30:05 – 00:02:44:18
Kiran Kachela
The second psychological barrier is the emotionless barrier. Now AI by nature is very logical, quite black and white. It’s rule based, is data driven, and it’s devoid of that human warmth.
00:02:44:18 – 00:02:52:19
Kiran Kachela
We need to understand that actually this acts as a significant barrier because as humans, we connect better with emotion.
00:02:52:19 – 00:03:20:02
Kiran Kachela
Think about the early days of voice assistants. They used to be very cold. You’d get very straight answers with no emotion. Whereas today, if you turned on your Amazon Alexa, and ask that question, Alexa will now respond with taunts like, And it’s those short pauses and those human like acknowledgements that help break down barriers.
00:03:20:08 – 00:03:23:07
Kiran Kachela
It starts to build a bit of connection with the human.
00:03:23:07 – 00:03:47:00
Kiran Kachela
The third psychological barrier is this fear of losing control or autonomy. Now people resist to giving up the steering wheel. I’m one of them. I need to feel like I’m in control. And if I’m letting go, I struggle. So many people will see that as a barrier. They will resist AI because they feel like they have no control.
00:03:47:02 – 00:04:00:09
Kiran Kachela
And so the strategy isn’t about removing humans completely from the process. It’s actually offering manual overrides. It’s putting humans in the loop. I think that’s so, so key.
00:04:00:09 – 00:04:12:16
Kiran Kachela
an example of that is back at my house, I have a nest thermostat and it does learn what I like. It learns, my preferences around temperature and different times of days.
00:04:12:22 – 00:04:27:00
Kiran Kachela
But what I love about it is that it does have the manual override option. So I know that if I want to take control and if I need to take control, I can. And I think that’s something that we need to think of as a strategy when implementing AI.
00:04:27:00 – 00:04:35:24
Kiran Kachela
The fourth psychological barrier is flexibility. Many people see AI as black and white, okay.
00:04:35:24 – 00:04:55:19
Kiran Kachela
And so they feel that it doesn’t necessarily reflect their personal unique requirements and circumstances. They see it as quite generic tool that anybody could interface with, but it doesn’t really understand my unique requirements. And so I think personalisation is very, very important.
00:04:55:19 – 00:05:02:02
Kiran Kachela
If you switch on your Netflix tomorrow, and you take a look at, you know, what to watch.
00:05:02:08 – 00:05:20:09
Kiran Kachela
It says things on there like, you know, based on what you watched last time. So it’s showing you that level of personalisation and it’s doing the explainability. It’s explaining to you how it’s got to that answer. So it’s this combination of explainability with personalisation.
00:05:20:09 – 00:05:28:10
Kiran Kachela
And last but not least is the fear of losing their jobs. And we know that this is a barrier in AI adoption.
00:05:28:10 – 00:05:54:00
Kiran Kachela
This is something that is very real. You know jobs are changing. Jobs are evolving as a result of the introduction of AI. And I think the key here is, as I said earlier, this isn’t necessarily about removing humans from the process. It’s about adapting and evolving their roles. And so as part of the rollout, we need to be clear that AI is on their side and not be seen as a threat.
00:05:54:00 – 00:06:00:22
Kiran Kachela
So there are a few sci fixes from today’s episode. Firstly, lead with humans.
00:06:00:23 – 00:06:02:10
Kiran Kachela
Human first.
00:06:02:10 – 00:06:16:14
Kiran Kachela
if you work from backwards from the human, you can start to think about what’s in it for them and make sure that you’re really understanding all of those psychological barriers and how you can combat those for those audiences that you’re working with.
00:06:16:14 – 00:06:23:09
Kiran Kachela
The second one is design. It’s about the design for explainability and personalisation and transparency.
00:06:23:11 – 00:06:34:10
Kiran Kachela
So when you’re designing your solutions, think about how you actually explain what the system’s doing, what you need to make visible and how you make it not feel so generic and make it feel personal.
00:06:34:10 – 00:06:38:21
Kiran Kachela
And last but not least, make sure humans are in the loop.
00:06:38:21 – 00:06:42:12
Kiran Kachela
Thank you for joining today’s episode, and I look forward to seeing you on the next one.