AI, Neurodivergence and Being Understood with Elaine and Jay cofounders of Evro AI

One meeting can spiral into weeks of stress when you’re neurodivergent and constantly second-guessing what you said, how you came across, or what everyone “really meant.” Dr Jay Spence, a clinical psychologist and co-founder of Avro AI, talks about a different way to look at communication breakdowns: most damage comes from misunderstood difference, not bad intent.

Elaine Lee, Avro’s co-founder, shares the story behind the product. After years in high-performing tech roles, repeated burnout, and a late diagnosis of ADHD, dyslexia, and autism, she realised how often people are judged for signals that simply don’t translate across brains. Her reflections on masking, workplace friction, and repairing relationships through better understanding shape the values behind Avro AI.

We dig into what an AI meeting assistant can actually do in real life: sitting alongside Zoom or Teams, creating accurate records, then helping you prep, rephrase, and debrief with less guesswork. We also tackle the hard question head-on: does this kind of tool support authentic communication, or does it push people toward neurotypical “norms”? Jay explains why choice matters, and why the goal is a bridge, not a makeover. 

To connect with Jay and Elaine you can find them here 

And through LinkedIn

Elaine Lee

Jay Spence

Thank you to this season’s sponsors:

Terri Wyse & Rachel Helm
Together, they are offering webinars and workshops focused on EBSA for both parents and professionals.

08.07.26  School webinar

02.09.26 Parent/carer webinar

16.09.26 Joint Workshop

Podcast listeners receive £5 off using this code UNTYPICALPARENT5.

Link below to go straight to Insta

The Untypical Parent Podcast

The Untypical Podcaster

Transcript

Welcome To The Backup Team

SPEAKER_01 0:00

Welcome to the Untypical Parent Podcast. Podcast for parents in neurodivergent families. I'm Liz Evans and I'm your host. Here we talk about the messy and the sparkles, share ideas you can actually use and give you space to take what could work and leave what doesn't. There's no such thing as a manual that perfectly fits a neurodivergent family. So the untypical parent podcast is here to make the hard bits lighter and the good bits brighter. Welcome to your backup team. We've been expecting you. Welcome back to the Untypical Parent Podcast. Thanks for being with us today. I have a very interesting conversation today, a really interesting one. And I've got with me Dr. Jay Spence. And he's part of a company. I've written it down, Jay, because with my dyslexia, I couldn't get it into my head what it was called. So it's called Evro AI.

SPEAKER_00 0:53

Nailed it.

SPEAKER_01 0:54

Oh, good start. Okay, I've got it right, I've read it right, I've said it right. Great. Okay. So I had uh Jay and his colleague Elaine reach out to me and wanted to come on the podcast. And I was really interested in this because it's about AI. Now, love it or hate it, it's kind of here to stay. We know that it's here. And I was so interested in what Jay and Elaine were doing because it was very specific towards neurodivergent people. So, what I'm going to start with is Jay, will you tell us a little bit about you and Elaine and what we're also going to do is that we're going to acknowledge that Elaine isn't here as part of the podcast that we're doing live, like this and recording, but that's for her her own kind of well-being. Um, and we're trying we've we we think we found a way that we can bring Elaine in in a way that feels comfortable to her. My um editing skills are gonna come into play here. So we're hoping that as part of this podcast, we are gonna hear from Elaine, but she's not actually with us here today recording, but we're gonna get her voice in because it's really important that we hear Elaine's side as well. But Jade, tell us a bit about you and who you are.

SPEAKER_00 2:12

Yeah, I'm a clinical psychologist originally. Uh I still practice NCP patients and have um worked in neurodivision patients for many years. But in 2016, I started a mental health technology company, a startup based in Australia, and went down a rabbit hole of entrepreneurship. And that's been really about trying to figure out what works or doesn't work from a mental health technology perspective, um, and using a bit of my research background and using a bit of kind of tech that I have figured out along the way. And then more recently started to kind of jump into AI and yeah, I acknowledge what you mentioned before. Like it's I think it's really can be really polarizing, and I totally accept that for lots of people. You know, the things that are happening with us and loss of jobs are really concerning. The side that I sit on, I I see kind of what the benefits could be and what the potential could be. Absolutely acknowledge, like that there's lots of potential issues as well as we're going through this. And I hope that we do get a chance to kind of meet Elaine through this. I just watched some of the recording of her, and it's a it's a really special um part of her story. So hopefully that part you get a chance to hear a bit from her as well.

SPEAKER_01 3:25

I'm gonna hand over to Elaine. We've tried to be really creative, and actually, what Elaine has done has recorded uh her story and the background towards Evro AI, and I'm really pleased that we managed to find a way to do that because actually Elaine's story is really, really important, and I want to thank her for finding a way to be able to do it. But we we came up with a way, we did it, Elaine, and I'm gonna do my very best to cut this in seamlessly now.

Kindness That Rebuilt A Life

Late Diagnosis And Relationship Repair

Misunderstood Difference At Work

SPEAKER_02 3:57

Hi Liz, my name is Elaine, and I'm a co-founder of Everle together with Jay. I understand that you've been in touch with Jay, and um, and I want to thank you for um asking me and inviting me to do a brain dump on who I am and why we've started to build Everle. I read about your work and I really love the work that you do. I think that it is important work, and I hope that my story can shine some light on why we think this work is important and why we want to build Avro. So, to start off, um I spent my career building technology across fintech, health tech, banking, a lot of software. On paper, it looked strong. Good roles, serious systems, high-performing environments. I was high performing. At that time I was not diagnosed as a neurodivergent. I just tried really, really hard in life to overcome all of the shortcomings I have, which I later learned that they are called masking. Jobs after jobs, I have never failed to work my absolute best. Yet I burn out, which is really common. And it's a um it's a common trait that I have found across my over 20 years of working. I keep seeing some constraints, they're not talent related, they're not effort-related, there are a lot of communication um problems. I often felt misunderstood. So that is work, but on a personal note, I um I love outdoor. I um I've lived and worked in Australia for over 25 years. I I like to hike, I loved hiking, I love immersing in uh nature. But before that, I came from a really difficult childhood, shaped by trauma and instability. By the age of 21, um, I reached a point where I didn't think that it was worth living or life was possible. The future wasn't just dark, it was absent or not worth living. Then a show uh a stranger showed up randomly. She brought with her kindness, loyalty, faithfulness, joy, with practical help and belief, and she helped me rebuild a life from the ground up, reconnecting, stabilizing, entering a community that lived a very different life that I had known before. That's the moment where life changed completely for me. For the first time, I experienced kindness without transaction, love that is not just skin dip, simple joy that comes from contentment, encouragement that is without agenda, belief without proof. She showed me what it means to be seen and believed in. I learned that comp what compassion looks like in real life and not just described. I learned that one act of understanding can alter the life of a life. And because of that, I carried a simple conviction. People don't need fixing, they need somebody to understand them accurately and lift them up. I discovered that I'm ADHD, dyslexia, and autistic late in life. At the behest of my I got diagnosed at the behest of my partner who is a psychologist. Until then, I had simply pushed myself relentlessly through sheer effort, masking weaknesses, overcompensating more than others, worked harder more than others, um, prepared more than others, burnt out again and again. The diagnosis didn't really reduce, didn't reduce my confidence. In fact, it clarified my experience. It helped me understand my thinking patterns, my behavior, and the all the patterns that I fought for many years. It also importantly transformed the way I understand relationships and communication itself. My partner and I communicate very differently. Our styles clash. It's easy to misread and misunderstand each other as uncaring. You don't care. You don't listen. Are you here? In reality, we were both trying, but our signals just didn't land in each other's minds really well. Once we understood how our minds work, how we're different, how we filter, express information, everything changed. Not instantly, but steadily over time. Empathy increased, repair became faster, compassion replaced accusation. Friction started to reduce, connection started to strengthen much deeply, much more deeper. Not because either of us has changed that much, but because we learned how to meet each other better in the middle. I then looked back at my relationship with my mom through the same lens. Many behaviors that have caused a lot of really deep hurts in the past now make sense neurologically. Misinterpretation that created a lot of hurt, where her behavior was difficult to understand, where her struggles had been mistaken as indifference, lacking love, where her limitation was misread as neglect, selfish, self-centered. That reframing and now thinking of her behavior through the lens of autism brought grief and also enormous healing. Now I'll explain why grief, but let me explain the healing. I can now see that as an autistic uh through behavior where her love is expressed in another way that I didn't get before, and now I do. Grief for the time that we lost that I could have understood her better and built better relationships. But healing now because I see her through a totally different lens. This is how much pain misunderstanding can manufacture across a lifetime. This realization brought a powerful lesson to me, which is most relationships are damaged not because of bad intent, but it was caused by a misunderstood difference. I see because of the relationship or my experience with family and with my partner, I see the same pattern playing out in organization workplace after workplace. I observed the same recurring situation where senior leaders believed that they gave a clear direction, but their team left with five different interpretations. High performer speaks directly to move things forward, but were interpreted as being aggressive. A thoughtful contributor stays quiet to be precise, but was judged to be disengaged or non-performing. And feedback intended to support was received as criticism. This happens a lot. Silence intended to respect was read as resistance. We speak the same language, the same English language, but we fail to understand each other because of upbringing, cultural differences, religious differences, communication style differences, working style differences, all sorts of differences. But research consistently shows that misalignment rarely comes from missing information. Escalation rarely comes from bad intent. It comes from assumptions mismatched, partial understanding, and defensive interpretation. But our modern workplace still relies on human manually guessing how and what another human being tried to say under time pressure, under cognitive load, and emotional stakes. This is why Jay and I want to build Avro. Our mission is to help build human potential through shared understanding by facilitating alignment and understanding. Because differences is not the enemy of performance. Misunderstood difference is.

Using AI Safely As Support

SPEAKER_00 16:55

I would talk about teams that I was working in because I was working as a consultant for the last couple of years, and we would kind of like break down situations where a project didn't land or friction emerged, or things that we thought were kind of solvable tended to just kind of spiral. And we were looking at it specifically through a perspective of like, okay, how can we fix this? And we went round and round for a long time until AI came along, and then we saw the power that it did have to be able to intelligently think about communication in a way that we hadn't seen previous technology be able to. The key of it was really starting to see that the the this moment that AI was presenting us was potentially a way to be able to take all of our individual communication and biases, the lens through which we see the world and the the or the history that we bring into a conversation like this, and the same thing on the other person's side, understand that person's biases, history, nuances, character, personality, all of these factors, because AI is capable of holding so much of that information at the same time. It started to kind of show us that there was this way of potentially seeing things much more objectively. You could kind of almost hold this central space of saying, I understand Liz, what your background is, and I understand Jay what your background is, and I can help the two of you to communicate because it wouldn't know what your background biases are, what knows what my biases are, and it can start to help us to say, all right, if you want something to land really well with Liz, say like this. If you want something to land really well with Jay, say it like this. And so that's what the tech does. It's a it's a meeting assistant, it sits next to a meeting and it uses some pretty sophisticated AI that I don't know whether I even understand fully, although I think Elaine does. About how it can start to help us to communicate more effectively with each other.

SPEAKER_01 18:57

Okay.

SPEAKER_00 18:57

Yeah.

SPEAKER_01 18:58

Because I suppose we touched on it just slightly earlier, was around different feelings that people might have around AI. And it it does have a very, like you said, a very polarized there's not many people I get that sit in the middle. They've either got a kind of a worry and a fear about it or they're truly embracing it and going for it because it's been, you know, brilliant for them. Um and I think I'm of the era of you know, Terminator. So there's that worry. The computers are coming and they're gonna take over, and we're all gonna need what was his name, John I can't remember his surname now.

SPEAKER_00 19:34

Yeah, yeah.

SPEAKER_01 19:35

Yeah, it'll come to me halfway through this conversation. Connor, John Connor.

SPEAKER_00 19:39

Oh, well done, well done. Thanks, Brad.

SPEAKER_01 19:41

We're gonna need a we're gonna need a John Connor at some point. But I think when I was looking into this, because this really interests me, because I use AI a lot as being dyslexic. And I think I was looking into it, and there are different types of AI which I hadn't really realized. There are kind of AI that helps you do a specific thing, like Grammarly, which I use a lot, which helps me specifically with my spelling, my grammar, give me suggestions around that. But then there's kind of this generative AI, which I think I'm getting right, and somebody might go to me, you are talking rubbish, but this is my bit of understanding. Is there's a generative AI, which is more like chat GTP, Claude, that kind of thing, where it's actually learning about you and it's offering solutions and it's being a bit more um creative, I suppose, than just solving a particular issue like grammar. So I'd kind of got and then the then the list went on, Jay. I looked and there was like all these different types of AI that I didn't realised. Yeah, and I just thought that's enough. I've got two. I'll remember those two, but they'll remember the rest. So I didn't go down that route. But I think as a dyslexic person, I get quite um I found it quite difficult because in the social media space that I work in, is that we have to write quite a lot. So you you know, we're putting posts out, all that kind of thing. And there is a very strong feeling and a very vocal feeling that if you use AI to help you in those, people will tear you down for using AI. And I think it's how AI is used, isn't it? And this is the key, and this is why I'm kind of wrapping back around in a very long way back to you. So I'm saying again that it's how you use the AI. Now, I agree, if you're gonna use AI just to write something and you've not thought about it, and you've just said, write me think this, and it's just spat something out, and actually it's not you, then okay, I can kind of see where they're coming from. But if you're someone like me who's dyslexic, I write it and then say to AI, can you help make sense of what I've just written? Because what if I do it in Word, Word will go, What? It doesn't even recognise some of the words. So AI for me has been has given Given me a voice on paper that I never had before, which has been massive for me. And I suppose what I'm really interested about with what you've created, Jay, is that this is potentially making life easier for neurodivergent people. And that could be, I have a lot of parents that listen. So it could be parents in meetings. You know, the amount of meetings that we have to go into as parents when you're neurodivergent and possibly sometimes you're a parent on your own with a multidisciplinary team meeting with many different professionals. That's terrifying for some parents. And then being neurodivergent on top of that is even more difficult. So I suppose that's why what really uh interested me in what you're doing. Kind of at what stage are you at, Jay, with this development? Are you out there doing it or is it on its way?

SPEAKER_00 22:56

We have just released an early version. And so we would love it if any of your listeners are interested in being part of the early adopter group. Um, please get in touch and I'm sure it'll be in the show notes.

SPEAKER_01 23:07

Yeah, we'll put that in the show notes.

SPEAKER_00 23:09

We we also kind of saw it as something that levels of playing field as well. So I I absolutely could see ways in which it could be problematic for people, and uh definitely there's uses of AI, which which I think can um be very risky, but I think that the potential side of this as well is a lot around a lot of people similar to what you were talking about, have found ways to make it theirs, and also theirs in a way where it's giving them some sense of a seat at the table or a new way of communicating or opening a door that was previously not open to them. We had found a whole bunch of people who were in some of our early groups of users who were taking a lot of meeting transcripts to prepare and using AI to analyze those meeting transcripts in order to prepare for the next meeting. And so they might take, you know, 10 of their previous meetings, use AI to kind of analyze that and build up a script to be able to come into the next meeting and know what to say. And so they might have a meeting with Bill or Sarah or whoever it is, and if they've got 10 meetings before with Sarah, then you can prompt using ChatGPT to upload those transcripts and say, hey, here's my last 10 meetings. Our meetings are you know in a business that sells this or does that? Can you please give me some good ways of communicating? And what ChatGPT and some of these other LLM models can do is they're really good at taking apart all of those different transcripts and saying, okay, I understand this. This is the basics of your company, I understand what Sarah's trying to do, I understand what you're trying to do. Here's how you would phrase things with Sarah. If you want to win-win, which is both for you, for Sarah, and for the business that you're working in, here's how you would go about this. And that's that's kind of where the genesis was came from for everywhere. Is we found that initially we saw Elaine was doing this, so she was prepping for her meetings in this way and using it, and then coming back and saying, This is really helping me. This is helping me to express myself in a way I've never been able to before. This is helping me to not be so stressed when I'm coming into a meeting. Or sometimes for Elaine, she would come out of the meetings and not be sure, like, was that person frustrated? Did I miss some social cue? Did I phrase that badly? How did I come across? And suddenly she's got this tool of AI that can sit there and will quite objectively say, Yeah, so and so was like you were monologuing for 10 minutes. They were, they were dropping off. Their non-responsiveness probably was you talking too much. Next time you will try like coming in this way or changing it to that direction. Well, you might phrase it in this way, or you know, with J, you might say it that way. And then that's essentially what Evro became. It was this it was a way to be able to automate all of the things that we found people who had ADHD or dyslexia or autism were doing already that takes a lot of time and a lot of uploading and transcripts and doing it in a quick button, one button to get the information you need.

SPEAKER_01 26:08

Yeah. And I suppose what people might be thinking when they hear this, and again, I'm kind of maybe playing the devil's advocate a little bit, is that they might be thinking, Well, are you trying to teach neurodivergent people to be less neurodivergent? And I don't I don't get that sense from you. I get a sense of actually it's, you know, say for example, it's just different ways of communicating, and it's enabling that person to be to use a communication style that makes them they feel understood, but equally understanding the other person as well, because there is this kind of uh missing sometimes, isn't there? We kind of miss what the other one's trying to say. Um, and I think we do that with the I think you know, my son's neurotypical and he does it as a neurotypical, will still miss uh what I'm trying to say, or my son who is autistic is trying to say, and he'll misjudge us. Um, you know, sometimes, like you say, my son is um he's very direct, which some people can go, ooh, that's a bit that's a bit harsh. Um, and I don't want him to not be like that, but I also want him to be aware that it could cause a reaction in other people and he just needs to be mindful of it. And sometimes we need that directness because actually it kind of cuts through all the crap. Yeah, that sometimes we and you know, if you've got a flighty brain like maybe you're I you know, you and I do Jay, is that we can go off around the houses, he can cut through it, and we need that sometimes. But I get that I think I suppose I just worry whether some people might go, yeah, but is that making me less neurodivergent? And I don't could you talk could you talk to that a little bit?

SPEAKER_00 27:49

Yeah, it's it's actually something that Elaine and I have thought a lot about, like the issue of masking of authenticity and um Elaine being neurodivergent herself and me being a clinical psychologist who's worked in the area, we uh we understand that there is this tension here of like, you know, are we norming the communication advice that we might give on a neurotypical model where we're just trying to force everyone back into some way of being more neurotypical, or are we allowing people to find an authentic way of expressing themselves that doesn't lose who they are and force them into some other way? I like I I wouldn't say by any stretch that we've perfected that because I don't think anyone really has perfected like the way to that we can meet in the middle of these two conversations. What I can say is we've built the models in a way to make sure that it's they are always respecting the choice that the person has to choose to find the way that works for them. I think it comes down to certain situations. Like what we find from the users who are using Everestopar is what the same user might have a time where they want to script something in a way that is very neurotypical for a specific neurotypical person where they can't really get their message across unless it's phrased in the way that that person can hear it. And for that those that user, that that same user might then dial back that there's a uh a feature in every that kind of allows you to be more of direct or less direct or whatever you need to do. Okay, dial that back because they know with the other person, like, oh, my direct stuff works really well here. Yeah, and I like this. And that there might be talking between two neurodivision people in these contexts. And so this is what AI is quite good at understanding. It's it knows, you know, who is this other speaker, who is this person, because it's got this huge backlog of meetings that the person's been in. And so it's learning over time who is Liz, who is Jay, how do I connect? And the part that we're going for here is not to say, let's try and make everyone talk the same language. Because I agree, we would lose like creativity, the healthy conflict, healthy criticism, all of the other things that different communication styles bring in. But at the same time, we want to make sure that it's not so disparate that you've got a situation where someone like your son might talk to a neurotypical person, and that person might say, I have no idea what this is about. This guy's just rude, I'm out of here. I think that's kind of where it's too much of a rupture. Whereas an AI might be able to come in and say, Oh, that person left because it was too blunt. If you want to rephrase it, here's how you do it. Now that's his choice then. He may choose say, I don't want to I don't want to change.

SPEAKER_01 30:31

I'm gonna be me. Yeah.

SPEAKER_00 30:32

If you're also saying, like, hey, I really need to connect to this person, it's super important to me, how do I do it? Then I think that's a bit different.

Real Time Feedback Or After

SPEAKER_01 30:39

Yeah. And does it only is it something that only works so it works in the meeting in the moment by recording it?

SPEAKER_00 30:49

It sits next to a meeting. So people can turn it on so it's public and people know that they're using it, or they can turn it off so it's completely private. And so you might have a Teams meeting or a Zoom meeting or a WebEx or whatever it is. It sits next to the meeting. You you do kind of press record so that it does kind of listen in and transcribe the meeting. And then what it does is you people either use it for real-time feedback. So if somebody's autistic and they might miss a lot of social cues, you can give some feedback around saying, like, uh, yeah, you you're talking a bit too much, you might want to slow down because you might.

SPEAKER_01 31:22

In the moment, it can do that.

SPEAKER_00 31:24

Yeah, it can tell people like a track light system of like, you know, that's what you're saying right now is drifting or it's being a bit too unclear, just slow down. And some people love that, other people it's way too distracting for them, so they turn that off. Yeah, I could imagine the water being it's not for me either because I I like to focus on the person. I don't want to see things.

SPEAKER_01 31:47

Yeah.

SPEAKER_00 31:47

But some people like it because they don't want to miss that, they don't want to they hate that feeling of like, am I kind of being left out as something here that everyone else is catching?

SPEAKER_01 31:55

Yeah.

SPEAKER_00 31:56

Um, yeah, I the the features that I more prefer is like after a meeting, it will sit down and and I run it, of course, with all my meetings with Elaine and with our software team who are from all kinds of different cultures and places. And it will say to me things like, Ah, when you had that meeting with your software energy from Pakistan, here's three things that might be working really well. But actually think about these three things next time because you might be having a you've got a couple of blind spots here, Jay, that you may need to think through next time.

SPEAKER_01 32:24

Okay, okay. And and this all came from Elaine struggling with these with meetings.

SPEAKER_00 32:33

Yeah, it did. It c it came originally because Elaine was doing all of this herself. So she would take these meeting transcripts and then she would load them up to Chat GPT, and it would take her a huge amount of time. And then we found that there was quite a lot of people doing this.

SPEAKER_01 32:48

Yeah.

SPEAKER_00 32:49

So it came out of both Elaine's kind of journey to you with using AI and experimenting what really worked for her, what didn't work for her. Um, and then generalizing that back out to like our early testing groups and figuring out what worked for them. It was really about trying to figure out how to automate a lot of what people were doing that was very time consuming and to make sure that it was performing in these ways where it was respectful of that person's authenticity and things like that.

SPEAKER_01 33:17

Because I can see a real use for this in, I know kind of our systems are slightly different, Jay, but in this country we have um we have EHCPs, which is our education, health and care plans, which these kind of were the old statements that kind of specified what the kids' needs were, what their provision is, uh and then every year we have uh annual reviews on top of that to just check how things are going. They're very, very stressful processes. Trying to get an EHCP is very difficult. Um, a lot of families often end up going to um special educational needs tribunal. Um and you know, as we know, uh neurodivergence tends to be run in the family. Um that if you've got a child with a neuro a neurodivergent diagnosis, there's a likelihood somewhere in the parent that someone's going to have one as well, probably, whether they know or not. And they're trying to navigate these meetings often in high stress responses, they've often got things that are very difficult. And you know, even I can think for me that that would be helpful that it it uh when you're in those high states of emotion, where again we can misunderstand things, misread things are you know at the best of times, our tone isn't always the best the best.

SPEAKER_00 34:34

We've dissociated or all the other things that can happen.

SPEAKER_01 34:37

Yeah, all of that. And I think a lot of the families, you know, I I work, we talk about a lot of trauma with regards to being gas lit and all sorts of stuff. That actually something being very objective like that, because although we're in there doing the best for our kids, we are emotionally involved, and it's not always I can't always hear what other people are saying because I'm so stuck on something that I don't feel people are hearing. And I can see a real uh use for what you're talking about for parents in that capacity and in in being in those meetings, um, as well as kind of in a work situation as well, with lots of parents do what you know work as well. So for me, it's really interesting. I can't I think it's it's a really interesting way of looking at using AI.

SPEAKER_00 35:25

Yeah, I I mean I think that we might be sitting at the start of something that could uh have a lot of follow-on companies. Uh and while we're doing work right now, like my hope is that other people will kind of pick this up and run with it and have applications for things like, you know, I can imagine that there might be a time in the future even though this is a bit oily in at the moment. But, you know, if you're having an argument with your partner, that you can use an AI to be able to help level it out and help to not be just a referee, like I want an AI that says you're your wife's right. But if you know, um I I've done years of couples counselling with my wife, and uh and one of the things I really loved about that process is my couples counselor never never told me or her who was right or wrong, which was pretty impressive. I'm sure I was right. I'm sure that's what she was thinking. She she never said it. I'm sure Karen would say the same thing about her. But what she did an amazing job of was it wasn't her job to tell us who was right or wrong. It was her job to help us to understand what were the biases, what's getting triggered, how do you resolve it? And what I've seen in my profession in the world of therapy is if you said to me three years ago there'll be an AI therapist out there doing therapy at the level of an of a good clinician, I would have said, that's no way. We'll be at the end of the line. Thanks very much. But there's AI out there that's doing the same job, like at the level of what I would say an intern psychologist can do, which is not bad. Like it's not bad. I wouldn't say it's great, it's not at the level of a great therapist, but it's in a short period of time it's got to intern level. You know, odds are in five years' time it's gonna be at the level of a good clinician. And so what that says to me is it can it can do what that couples counselor are doing. Like there will be applications of AI that can sit down in the same way that a good couples counselor will do. And if you have a fight with your partner, it's not gonna say who's right, who's wrong, but we'll help both partners to understand and resolve. And that I think I'm excited about that because seeing a couples counselor is prohibitive for most people. There's a shortage of the and if we've got AI that sits in our computer for 20 bucks a month that can do that kind of thing, I think we might have a world that's got potentially less fights in it. We'll see.

SPEAKER_01 37:51

So is that you then out of a job, Jay?

SPEAKER_00 37:53

I think I could be, but I might be joined by a lot of other people today. That's okay.

SPEAKER_01 37:59

And I suppose what's interesting as well is I think in is even just the language, because a lot of parents, you know, going into um SEN, the SE world of SEM, which is special educational needs, is that the language and it's the acronyms and the education blows my brains. Um probably the other way around, health can blow people's brains with the inacronyms and the short version of this. And and again, in meetings, that's that's really difficult, isn't it? I think when you're people are are firing stuff at you, you're maybe neurodivergent, uh overwhelmed by the situation, and we don't always have the voice to be able to say, I don't understand what you're talking about. Yeah, yeah. Um that can be that's that takes a level of confidence and bravery to be able to do that in a meeting. Um, and I just wonder again whether there's you know the background for that and AI and how that could be used. It could just be huge. I don't think you're gonna be out of a job doing this one, Jay.

SPEAKER_00 38:58

Yeah, well, like I I I think that the impact could be really big as well. And there's two levels to that one that I am personally excited about. Like, one of them is that that you know, I think we've all sat in a meeting where you can hear different parts of it almost. Like, you know, we all take away something different. Yes. First generation meeting transcribers and note takers are actually really good at that. So, like, there's lots of free ones out there at the moment um that can do the transcription and they're like an accurate system of record, like it can tell you who said what and when. I think that that's what the first gen note taggers were good at. I'm thinking that some of the second generation's one like Evro, and you know, potentially like others that might come along. The next layer to that is almost like it's that thing where you can sit in a meeting, hear the same thing, both people can kind of quote almost verbatim what happened in that meeting, but our own backgrounds and biases mean that we can p we can still misinterpret what's going on. Yes. And that's the part that I'm really interested in. Because a as a psychologist, that's where I see lots and lots of breakdown in relationship occur.

unknown 40:02

Yeah.

SPEAKER_00 40:02

Like I if I think about personal instances for me, I might have been in a workplace and you know, I I'd had a difficult colleague, and I'd go, oh gosh, this person's really driving me mad. And like I'd be in these meetings and I'd hear things in a certain way, and I'd perceive them as being adversarial, and then something might shift, and then I'll come out on the other side and I'll see exactly the same behavior. The same behavior as the day before. But because of the shift that I've gone through and understanding that person, I can just perceive the same thing, go, it's not adversarial at all. She's doing that because she's got this agenda over here that I didn't even know about.

unknown 40:37

Yeah.

SPEAKER_00 40:38

And it's then my that's the thing that I think that's the next layer. AI is capable of doing some of these things now, which is really exciting to a psychologist because it means you've got this way of helping people to see past their biases if they're open to hearing it.

Features Roadmap And Broader Impact

SPEAKER_01 40:53

Yeah, yeah. This fascinates me. This really does fascinate me. And I think so. Where you're at the moment, Jay, it's at this kind of you're doing these this testing, is that what you're doing with it? You're running, so you're looking for people that would be interested to come and have a go with it and and trial it.

SPEAKER_00 41:11

Yeah, so we've gone through the first rounds of testing like enough to know that it's safe, it's not going to do anything outrageous and um the same level of safety that you would expect from kind of a good AI tool. Now we've got into the next level where we've kind of built the first round of features. So the first round of the things that you would get in a normal meeting note taker, so it can do the summary, it can do the action items. And then we've just started to build out the first round of reporting. And so you click a button at the end of the meeting and it will start to give you a breakdown of like, okay, in your meeting with Liz, here's some things that went well from a communication perspective, here's some things to think about for improvement for next time. And then if you and I were to have a few meetings, there's another option to kind of click another button and it will do the prep for the meeting. So say, Are you about to have a meeting with Liz? I read through your last three meetings. You know, here's what you might want to think about as your agenda, and here's how you might kind of want to phrase things with Liz. And like I said, there's ways to be able to configure that so it can be more looking at how that's going to work for you, or more kind of authentic to how I would like to phrase it myself. Yeah.

SPEAKER_01 42:15

Yeah, brilliant. And if people want to do that, we'll leave um some notes in the comments for everybody so that they can find that. And then what happens after that, Jay? So you're doing this next bit. What the what's the kind of what were you where are you going next?

SPEAKER_00 42:32

Yeah, so like the the process for us is really part of the mission. The mission is Elaine and I would really like to to help people to communicate better. Meeting Note Taker for us is a really great way to do it because it's something which we use every day. We're we're all in meetings constantly. We felt as though it was the right place to put a tool that for people that are really interested in communication or that might kind of have a lot of friction, or you know, whether or not it's a problem, or you might just want to improve, that it just felt like a natural place to start to put the tool up. But the mission for us is to be able to take it to wherever we feel as though it's gonna have an impact. We feel as though the right place to start is neurodivergent um uh people because of Elaine's background experience, a lot of her kind of DNA and experience has gone into this, and a lot of my knowledge and expertise from working with neurodivergent clients. Um but you know, I I I I think it's gonna need to be for both neurodivergent and neurotypical people eventually, because it's otherwise, you know, we're just gonna reinforce these stereotypes and differences of chain people. Eventually, I would really like to see it kind of being used by both sides. So it's kind of helping both neurodivergent people to understand how to communicate with others and neurotypical people to realize like there's a lot that could happen on the neurotypical side to adapt to what's happening with the neurodivergent person. Yeah.

unknown 43:58

Yeah.

SPEAKER_01 43:59

Okay. So that's That sounds amazing. So, what we'll do is we will leave all this information in the show notes. Is there anything that I have missed, Jay, that being a not very techie person around AI that you felt was really important that you wanted to bring? We're obviously going to see if we can edit in Elaine's voice into this as well, because I think that would be really great to hear from Elaine. So we'll that might pop up somewhere in might have already popped up if you've got to the end. It might already have popped up and my AI, my uh computer genius has shown through shone through, and I am uh doing a brilliant job and have done it. Um is there anything kind of just a last thing that you want to leave our listeners with just to say or leave them with at this moment?

SPEAKER_00 44:44

Yeah, look, I I think my parting comment around this would be like if AI is something that you feel really cautious about, um, while I do really understand that, I think that it's also important to try to understand it. And that there are ways to be able to kind of start to get across using tools in a really safe way, especially if you've got someone like yourself who's friendly and understands it to be able to help you through like what it's capable of. Because I I think it's a bit more than you know the end of art and all the other things that that that are happening that are alarming. There are so many ways in which I'm seeing it help um Elaine, help users. As I said, I still work with clients and I've had clients who have come in in at like tears, like that they've been really emotional to my new reversion clients, mind you, who don't tend to do that very often. Because they've found ways to use AI that work for them.

unknown 45:43

Yeah.

SPEAKER_00 45:44

And they're seeing that there are things that never happened for them before about decoding or translating or expressing that were off the table and are now available.

unknown 45:58

Yeah.

SPEAKER_01 45:58

And I can really relate to that, I think, as I said before with my writing, it's given me a voice on paper that I never had before and that I would always avoid. And actually now it's giving me it's opened up another avenue for me, which I wasn't there before. Um, so I get very defensive of times at times of AI, still whilst thinking, oh my god, Terminator could be round the corner.

SPEAKER_00 46:21

I'm so impressed that you remember John Conner, by the way. That's amazing.

SPEAKER_01 46:26

It's because I like the motorbikes, I think. So I was quite into it, and I thought, yeah, yeah, Terminator's just I love Terminator movies. Anyway, I've gone off. Um, I've done quite a lot of stuff. I've stayed on topic. Jay, thank you ever so much for coming on. It was a really interesting conversation. Please thank Elaine for us as well as um from the podcast and the listeners. Thanks so much for having me on. So that brings us to the end of the podcast recording with myself and Jay, but I just want to leave everybody with a comment from Elaine about imagining if. And I'm gonna just hand it over to Elaine for the last bit. Um, and just want to thank everybody, Elaine included, and Jay. Thank you ever so much for being on the podcast. And I'm gonna hand over to Elaine to take us out. Thanks, Elaine.

SPEAKER_02 47:13

Imagine a world where being different um does not mean being misunderstood, where clarity is assisted or facilitated, where more people are able to contribute at their full capacity because um they are truly understood, people get them. The vision is ambitious, the journey is long, the work is demanding and values-driven, and the outcome, hopefully measured in human potential unlocked or lives changed, is worth it. This is why I'm building Avro. I hope my story is one that can showcase why this work we we find this work important to us. And I would love to talk to you one time or one day and to share and understand the work you've been doing. And I think it would be wonderful to hear. Thanks, Liz.