Ethical Capability Building with AI | May 2025

Summary

In this insightful session, the focus is on the ethical considerations and challenges in AI, particularly in capability building. It discusses how AI is integrated into both private and public sectors, especially within Australia. The speaker from Acorn, with a diverse background in product capability, shares experiences from developing frameworks for ethical AI use. Highlighting the parallels with past industrial revolutions, the session underscores the need for responsible AI practices today. Key topics include the use of AI in decision-making, employee usage guidelines, data security, and transparency. With a humorous and reflective tone, the session also dives into real-world applications and the evolving landscape of AI ethics.

Highlights

• Started with fun banter about ethics and AI, making a heavy topic feel light! 😄
• Speaker shares a decade-long journey in capability building with AI. 🎢
• The importance of being a lifelong learner is highlighted. 📖
• Clever use of analogies with past industrial revolutions to discuss AI ethics. 🏭
• Emphasizes on training employees for responsible AI usage. 🏋️
• Discussions around the convenience of AI in the workplace. 🧑‍💻
• The role of AI in automating decision-making is critically examined. 🛠️
• AI's potential to enhance productivity and well-being is discussed optimistically. 🌟
• Transparent sharing of Acorn's internal processes for ethical AI use.🗂️

Key Takeaways

• AI ethics is crucial for both private and public sectors, especially in Australia. 🌏
• Past industrial revolutions offer valuable lessons for understanding AI's ethical challenges. 📚
• Data privacy, transparency, and responsible AI use are critical components. 🔐
• Balancing AI's power with human oversight ensures ethical decision-making. 🎛️
• Continuous education is necessary for keeping up with AI advancements. 🎓

Overview

The session kicked off with a lively mix of humor and insight, tackling the serious topic of AI ethics with an engaging approach. The speaker, from Acorn, took the audience on a reflective journey through the evolution of AI in capability building, underscoring the significant role ethics plays in this domain. With over a decade of experience, the speaker emphasized the importance of continuous learning and adapting to new ethical challenges presented by AI. 🌐

Drawing parallels between the current AI landscape and historical industrial revolutions, the talk highlighted how past challenges can inform present solutions. The session explored how navigating these challenges involves assessing data privacy, ensuring transparency, and focusing on responsible AI usage across sectors. These are not just checkbox exercises; they are imperatives for sustainable and ethical AI integration in both public and private realms. 🔍

With a blend of practical advice and visionary thinking, the session wrapped up by addressing the real-world impact of AI on productivity and well-being. It stressed the necessity of human oversight in AI-driven decision making and championed the need for ethical frameworks and compliance measures. Through this dialogue, participants were encouraged to adopt a proactive stance on AI ethics, embracing continuous education as a pathway to informed and responsible AI application. 🚀

Chapters

• 00:00 - 00:30: Introduction to Ethical Capability Building with AI The chapter 'Introduction to Ethical Capability Building with AI' begins with the speaker acknowledging the complex and sometimes contentious nature of discussing ethics, especially concerning AI. The focus of the session is the ethical considerations and the journey organizations have undertaken in recent years. This includes a particular emphasis on security and risk assessments conducted with both private and public organizations in Australia.
• 00:30 - 01:30: Background and Experience of the Speaker The speaker reflects on their journey and long-standing involvement with the company Acorn, highlighting nearly five years of contributions. They mention the company's growth, which they've witnessed firsthand, starting from a small team of approximately 11 people. The speaker hints at having had the opportunity to work internationally, and acknowledges the audience, some of whom have been with the company even longer than they have.
• 01:30 - 04:00: AI and Ethical Challenges The chapter titled 'AI and Ethical Challenges' begins with a background perspective from a speaker who works in the product team as the head of capabilities. This role involves contributing to anything seen in the capability module and assisting with implementations around capabilities. The speaker identifies as a lifelong learner with a strong interest in academia and theory and aims to apply this knowledge practically in the real world.
• 04:00 - 06:30: Historical Context of Industrial Revolutions This chapter discusses the historical context of industrial revolutions, focusing on niche areas such as work with organizations and the development of capability frameworks. The speaker references their experience and affiliations with AHRI and mentions involvement with the HR leadership council, which was part of the corporate executive board acquired by Gartner.
• 06:30 - 13:30: Ethical Considerations in AI Use The chapter discusses the importance of ethical considerations in the use of artificial intelligence (AI). The speaker shares their background in capability uplift and research, emphasizing their education in AI. They mention that AI will be used throughout the talk, prompting the audience to observe how AI is being integrated into the presentation. The chapter sets a tone of transparency and invites the audience to reflect on the ethical implications of AI usage.
• 13:30 - 17:00: AI in Workflow Automation The chapter discusses the skepticism around the use of AI, especially in terms of privacy concerns with recording capabilities, emphasizing transparency by declaring AI use. It transitions to exploring mixed perceptions about AI, particularly in capability building, and aims to gather thoughts and words associated with AI, which is crucial for the subsequent discussion.
• 17:00 - 28:00: Challenges and Perspectives on AI Ethics This chapter addresses the topic of AI ethics, examining both the challenges and different viewpoints associated with it. The dialogue reflects a lack of understanding about using AI for building capabilities efficiently. There is a sense of excitement and balance in the discussion, suggesting a workout or exercise metaphor. The speaker is not allowed to inquire further during the conversation but plans to discuss it afterward for clarity. Additionally, there is mention of 'infancy' and 'super nightmare,' indicating contrasting sentiments or stages related to AI ethics.
• 28:00 - 30:00: Conclusion and Availability for Further Discussion The chapter explores the origin and motivation behind the project, emphasizing the swift and iterative process of writing documents and addressing questions.

Ethical Capability Building with AI | May 2025 Transcription

• 00:00 - 00:30 hi everyone good morning it's always um always  fun to be up first and talk ethics right can be   a bit of a contentious issue um I'm going to  talk uh there's AI in there i'm going to talk   more around today um the sort of ethical thoughts  that we've had and the journey that we've been on   through the last few years um what really prompted  this uh session was more going through um security   and risk assessments with organizations both  private and public both here in Australia and
• 00:30 - 01:00 overseas as well so bit of a journey around  that um for those of you um who don't know me   or know me from zoom or there's various different  things quick who am I so I've actually been with   Acorn coming up to five years now uh I know  some of you have been with us even longer   so thank you for that but um I remember when we  were about 11 people DC probably when I joined or   something like that so I've really seen the growth  and fortunate to have been part of that um from a
• 01:00 - 01:30 background perspective so I work in the product  team the title is head of capabilities what   that actually means is anything you see in the  capability module I've been part of but you also   um may um have time with me for things like um  supporting the implementations around capabilities   etc um in the spirit of lifelong learner um  I'm very nerdy i'm into the academia i'm into   the theory and I like the bridge to bring it  into the real world as well and the practical
• 01:30 - 02:00 side of things um I'll say to my Acorn time I'm  also a member of AHRI for you for you all who   are members of AHRI and I'm also a facilitator  now since early 2024 I think on developing a   capability framework so having looked at so  many having worked with so many organizations   etc AHRI got in touch and were like it's  a really niche area can you help um and my   grounding in this whole area for the last sort  of 10 years does everyone know CB they were the   corporate executive board bought by Gartner but  they used to have the HR leadership council they
• 02:00 - 02:30 used to do capability uplift so that's where I got  my education basically on it so very much grounded   in research so that's who I am all right before  we jump in in the spirits of ethics in AI and   declarations I am using AI right now I'm going  to use AI throughout this talk to support this   talk and I'm going to ask you at the end how you  think I am using AI so keep a eye on it it could   be anything like what is he doing right now for  AI i don't have smart glasses on obviously you
• 02:30 - 03:00 can see that I don't believe in those because you  could be recording someone at any time but yeah   I'm declaring that I'm using AI right now and that  will finish at the end of my session and I will be   asking you cool all righty um first of all what  I usually do here is we see all these different   words and they're really like mixed on um the  good and different things like that does any is   anyone caffeinated enough to share a word or two  just shout them out on their thoughts generally   about AI and capability building it's sort of  super important for what I go on to next so I
• 03:00 - 03:30 wouldn't have a clue how to use AI for capability  building okay efficient efficient excited balanced we're going to do workouts i'm not allowed to  ask you what you mean by it but I'll we'll chat   to you afterwards so that I could unpack  it a little bit more anybody else infancy   sorry I miss I miss infancy what do we  have yep anybody else super nightmare
• 03:30 - 04:00 i hear you to go back to what triggered  the creation of this thing writing the   documents answering the questions  absolutely yeah yeah rapid rapid   iteration yes crazy quick right yeah uncertainty all right well I might might move on to the next  one um and I want to go on a bit of um we'll do a
• 04:00 - 04:30 bit of a a quick journey first so um when I think  about the ethical challenge um you know uh rapid   interaction rapid uh building it's moving very  quickly the it's very hyperbolic there's a lot   that I read in the news it's very clickbaity  and things like that what can it do what can't   it do there's a lot of myths out there there's  also a hell of a lot of truths about us being
• 04:30 - 05:00 good custodians of the technology as well so the  reality is though it's already shaping everything   we do today and has been for a while now um so if  we're not thinking about the ethics more broadly   um we really should be and and consider  considering that so um to quickly go like it's you   shouldn't always look at history as a precursor to  performance in the future but um we're going to do   that a little bit because we have sort of been on  this journey in a roundabout way before right so
• 05:00 - 05:30 if you look at the first industrial revolution  a couple of different definitions out there but   more broadly it was moving from just hand labor  to uh mechanical making it um machinery get going   all those sorts of things and the impact with  this was in a capitalist society we can really   work people this is good so um there's a challenge  there that we got to think about ethically and the   outcome from an ethics perspective was things like  child labor laws so that was something that we
• 05:30 - 06:00 went on um as a society before then we got to the  second industrial revolution so we got steam power   electricity we got all these elements and then the  consideration became about well we got these big   machines now like what on earth are we going to  do with this side of things and that's where we   actually found occupational health and safety laws  came in and that side of things so so we reacted   then we go into the third industrial revolution  which um actually one could argue started around   the 1950s we began to get data we going to get  digital transformation it really kicked in sort
• 06:00 - 06:30 of the ' 90s as we got the web and then again  we all acted as a society and and governments   etc and we now have things like GDPR and the  Privacy Act and and all that good stuff as   well and that leads us to where we arguably are  today which is the fourth industrial revolution   and there's some argument if we are or if we're  not um within that um artificial intelligence   is very central to that machine learning um that  digital age in a in a nuanced way that we're going   into and what does that actually look like so um  that's what we've been thinking about a little
• 06:30 - 07:00 bit and that's what I'm going to talk about and  share what we've learned especially in the last   two years um as we've really focused hard in  this area so if I was to distill it down I lump   um the ethics in AI in sort of three areas  that I observe in the conversations I have so   number one at the top there is how are employees  individually using AI and what I mean by that is
• 07:00 - 07:30 uh most government is Microsoft shop so you  utility of copilot you may know chat GPT or   Gemini or different large language models of AI  but like how are they using it are they doing   it the right way like what are the dynamics around  that so I'll go there in a second so that's bucket   one then bucket two is how you look at uh an  Acorn so SAS um software as a service provider   so any technology you use like how are they using  AI are you aware of it are you clear um does it
• 07:30 - 08:00 align with your expectations um and then the third  bucket that is actually commonly overlooked until   you scratch the surface a little bit is I put  in here automation of decision- making um it's   actually AI and workflow automation so you're  seeing more and more of disparate systems being   connected so people will talk about APIs and web  hooks and getting data from all these places you   can now embed AI as a layer in between that  and we need to think about that decision-making and
• 08:00 - 08:30 the impact of that and what that means for our  organizations so they're the three key buckets   but we'll just break them down a little bit one  at a time so how are your people using AI and   um before I click through my bullet points I  was reading something the other day which was   um some larger enterprises no longer know for  sure who are the highest performing graduates   but they do know which graduates are the best are  using AI which is a really interesting take right
• 08:30 - 09:00 because they're grappling with that so um what  are the things that we look at so um training on   responsible AI use what does that look like what  is responsible AI use how does it align with the   values of your organization because ethics and  values are are properly aligned is it declaring   and things along those lines do you have um clear  objectives for AI use like what is the purpose of   this making sure people aren't over reliant um I  and we're very very um progressive with our use
• 09:00 - 09:30 of AI and I use it a lot um for my daily work and  we talk about how we use it and stuff like that   the one time it was down I realized I was a little  bit over reliant on it um rewriting my emails for   example it turns out I got very lazy at writing  emails so it's it's it's a good one to call out   so you want to avoid that side of things um and  then because I spend a lot of time looking at   capability frameworks with my job and stuff like  that you got to make sure that there's ensuring
• 09:30 - 10:00 ethical alignment in there as well um it's  it's a hot topic with people you've got early   adopters you've got your late adopters the other  thing I think about um that's not on here is um   fairness and equity as well um you know I was at  dinner with my mother-in-laws the other day and   we talked about it and she's coming to the end of  her career she's a school psychologist and she is   completely overwhelmed with what they're pushing  and when she told me what they were pushing it   was like literally a little button on your email  in um Gmail that um t like tidies up your email
• 10:00 - 10:30 or something like that it was very entry level  when she was feeling overwhelmed and that's zero   reflection on her it's absolute reflection on the  training and the fairness and the equity and all   those things that I needed so so that's bucket  one but before we go into number two um I see   this a fair bit and I don't know if it's generally  um a thought of everyone's but um this belief that
• 10:30 - 11:00 um you are uploading company secrets and you are  uploading um private information etc yes if you   log into a chat GPT and you um just do it without  a login and any follows that that that can that   can be a possible thing you're doing but a lot  of the models the large language models of the   AI you can turn this off you can do it in a secure  way and it goes back to my further point about how   are we training people are we showing them how to  use it responsibly so that they can get the most
• 11:00 - 11:30 out of it um I was actually just on a call before  this and I brought up a piece of research that   I'll give you the link if you need it afterwards  but it just came to my mind earlier is it's called   the cybernetic teammate bit of a mouthful but the  cybernetic teammate is the first research to come   out of Harvard that shows that AI used properly by  an individual by one person outperforms a team of   four in thinking if it's trained and used properly  and I think we've all sort of been waiting for   research to validate uh that sort of thing so um  anyone wants it I'll share the link but bit of a
• 11:30 - 12:00 myth you you can use it you can in a careful way  share limited information um that doesn't mean I   advocate for you putting the financial budgets  in and stuff like that or anything ridiculous   like that all righty so companies like Acorn how  are we using AI um what are we doing about it how   have we thought through it all these things so  first of all understanding how we implement AI
• 12:00 - 12:30 um bringing that to life a little bit is if I  was a new buyer is AI immediately switched on and   exposed to everything the answer should be no you  should not in any way for whatever platform you   procure AI should be able to be um not within the  platform as a starting point right so that's the   first thing you need to understand that you need  to know where it is in the platform you need to   know the job it's serving as well um might sound  like an obvious one to the uh security nightmare
• 12:30 - 13:00 and stuff like that ensuring data privacy and  security um it's incredibly important if you   ask any of a provider you all the different  systems you have they will be able to tell you   instantaneously in one line how they do that if  it's someone who's not on the product side like me   um they might not know but someone like me or DC  or someone like that should be able to answer you   instantaneously that should be at the core of the  ethics of what we're doing transparency so what I
• 13:00 - 13:30 mean by clarifying AI transparency is um uh what  its role is how it's used is it a drafter is it   a writer is it a recommener is it a thinking  like all these different things um but where   in the platform is it is it clear is it clear  to admins only is it clear to users you have to   think about all these things as well so another  question that you can ask of providers um and   then not only that like they're like why behind it  like are they just putting it in because we're a
• 13:30 - 14:00 private software company and AI if we put AI in  our URL we'll grow quicker it's like you got to   have thought about these things very carefully and  properly and you want to get under the hood to the   foundations behind that um so these are actually  old documents now they're not designed they're   now designed and very flashy and stuff but when we  started on on all this work you can see here we've   got a responsible use policy we've got how we work  with um data we've got all these elements and um I
• 14:00 - 14:30 led all the iterations of writing conduit between  our um data um data privacy officer CTO and Sam   you were involved so like a lot of stakeholders  but I sort of led it and pull it together and   we had to think very carefully around um like  responsible use like what are the ethics and the   guiding principles um and it's very simple how we  actually did it so originally we started with the
• 14:30 - 15:00 um CSRO's principles um so that they were a  really good benchmark for us to really settle   in and and define what we wanted to use and what  we what we would communicate with the market and   then industry released um uh some AI ethical  principles and guidelines last year they're   really good there's eight of them um fairness of  society well-beings in there transparencies in   there accountability at any stage can you contact  people but it gives you that grounding as across
• 15:00 - 15:30 the board on how the ethics play into what you  are doing to make sure that you're doing the right   things and you've got those guiding principles i  like them personally because they're principles   and that means that I don't have to follow some  arbitrary rules that don't necessarily translate   to our environment but principles are broad enough  that we can adopt them ourselves as a foundational   way so you can access these documents anytime you  can ask us for them and they underpin everything   that we do and these are the ones um that get us  through the security assessments so um we can we
• 15:30 - 16:00 can talk about that as well all right then that  third bucket that I mentioned was around AI AI   and workflow automation so that is connecting  different systems AI embedded in the middle how   does that look and work is it automating decision-  making what do I need to think about etc so um I   think a key bit here is um mitigating the biases  so um there is a risk around that um of course
• 16:00 - 16:30 uh you know AI is fantastic it's all very exciting  um but humans still built it right so   there's still like there's going to be some  sort of bias in somewhere so how can we do that   um my personal view on it is that AI is an augment  augmentation tool not an automation tool I think   there should be a human in the loop at the end  of everything and it should equally be a drafter   um I don't think it should be empowered to do  everything on its own that's just what I believe
• 16:30 - 17:00 uh maybe the technology will improve and  and all that sort of thing but I feel quite   strongly about that i think it should speed up  the grunt work do the low value work and free up   every person to do more high value in their work  and I think we'll see more workplace well-being   with that as well you do need to prioritize  data governance of course goes without saying   um goes back to the security side of things as  well um and then again just that ethical oversight   layer if we are connecting disparate systems with  AI embedded in the middle what does that look like
• 17:00 - 17:30 how does that work etc the video I'm about to show  you is the CEO of a company called Anthropic and   Anthropic are the company behind Models Like  Claude if you've heard of that um and this was   from a podcast I was watching like November or  October last year so I have to state these things   because that's old in AI terms but um there's a  bit of noise in the background because to go on
• 17:30 - 18:00 the tech side of things I just screen recorded it  because I didn't know how to clip it because like   that so bear with me but I thought it's really  good to hear how because we're dependent on the   large language models so the database to do our  work so I'm interested in it's like supply chain   ethics in supply chain i'm interested in what they  think um and I thought this one caught it best   say and they're then tested both internally  and externally for their safety particularly   for catastrophic and autonomy risks uh so uh we  do internal testing according to our responsible
• 18:00 - 18:30 scaling policy which I you know could talk more  about that in detail and then we have an agreement   with the US and the UK AI safety institute as well  as other thirdparty testers in specific domains   to test the models for what are called CBRN risks  chemical biological radiological and nuclear which   are you know we don't think that models pose these  risks seriously yet but but every new model we   want to evaluate to see if we're starting to get  close to some of these these these more dangerous
• 18:30 - 19:00 um uh these more dangerous capabilities for me  what's important when we talk about ethics is   there's loads of different models that we can use  with our own AI so large language models I want   to know that they're taking it seriously because  they're moving incredibly quickly and I want to   do my due diligence um and I just think it's  an important part so anyway underpinning those   three use cases is governance and compliance  as well um so obviously it goes without saying   adhering to the data uh protection laws um if  you have ethical AI policies we are at a stage
• 19:00 - 19:30 where as we do more of this and it becomes  uh part of our world in a even clearer way   than it is that this is key um the governance  teams that you have already where does this sit   etc um but the continuous education for them  like what does that look like how do we keep   up with it um things like that video I played you  like no they don't need to listen to the podcast   I listen to but you need to have your finger on  the pulse does anyone want to elaborate and share
• 19:30 - 20:00 why they feel like they're on the right track  what gives you that feeling um because in the   spirit of constructive conversation there isn't  a policy in place so um does anyone want to share i think for me just because it makes sense I  understand what what you're saying but there's   obviously more to learn and a long way to go  it's also like there isn't a blueprint per   se out there i was looking for that like I was  I was looking for like where is a there's like
• 20:00 - 20:30 something I can give you afterwards that's like  a a guide or something and there's loads of good   um APS level things out there but there's not like  a for the three buckets that I see how can you hold   your SAS providers accountable there's not yet  that that exists so the other thing is there   is a I'm having to figure it out moment as well  and I think when I think back to the goal of the   talk it's to shine a light on it share what we're  learning but this isn't one of those aha moments   where I've got all the answers either it's a it's  sort of like a shared journey with with a new   area anybody else yeah I feel like with the like  going through the upheavalss of society with the
• 20:30 - 21:00 revolutions like kind of like that riding the wave  everyone's had to go through in the past the first   industrial revolution then the web as well not  knowing y how to do it so I feel like this is just   one of those times where it is have to understand  that no matter what's going to happen so the more   we can understand the better it will be yeah I I  really good point is massively an education piece
• 21:00 - 21:30 and it's an education piece from every stakeholder  to share where they're up to like so we're a small   part of a very big system at Acorn but like we  should be we're custodians of one part and we   should all be playing a part to help people make  the correct answers and that's how we'll see more   policies come into play and stakeholder engagement  and things like that i think it's good we're we're   definitely on the right track i'm always a bit  cautious about how remove unconscious bias when   creating the underlying chaos and then it's very  very difficult but that's always in the back of my
• 21:30 - 22:00 mind like how do you redirect it without creating  more biases how do you I think about that I think   about that so like so we have so for capability  discovery you've got a job description recommener   that you put your JDS in and it goes into the  capabilities and then well like the bias is What   defines a capability which is to us a capability  is a group of knowledge skills and behaviors that   drives a business outcome is that a bias like but  that's how we set it up and it's like other people
• 22:00 - 22:30 might see it differently and then what parts is it  passing and reading and you know there's there is   there is there is these small elements to it and  um I think about the two I don't have an answer   obviously you can tell I don't but it's like I  think the best the the way I've decided the best   step forward is to acknowledge it and say this is  how we fought through it and have a good way to   articulate that so if people don't agree at least  they know and that's where the transparency comes   in um but yeah it is an interesting one i think  it's a bit of a slippery slope cuz we live in a
• 22:30 - 23:00 world of convenience and AI is very convenient  even in its like kind of still in its toddler   stage like where is the line of how convenient  can it be before people just start using it to   like overtake work of people mhm i think it's  just there's not enough rules around AI just
• 23:00 - 23:30 like as we have as like a community and even if  we do start implementing rules AI just continues   to build and build so there's just like with the  further um industrial revolutions they were kind   of set like they were like oh this is a safety  hazard don't do this anymore ai is much more broad   how do we regulate something that is just growing  as quick as it is and as like big as it can become
• 23:30 - 24:00 the the way I thought through it is the reason  to do like a a a deliberate conversation around   ethics and values because it's like um the other  thing with ethics and really scratching at it is   the um the tension it can create internally if  people have values conflicts and stuff like that   so it's not even even the ethics side of it isn't  a simple conversation when it comes to where the   line is and then if you scratch even deeper like  value misalignment is one of the biggest things
• 24:00 - 24:30 around what causes um psychosocial hazard and  psychological distress and things like that so   you really like can go down a rabbit hole if  you want to um it's yeah it it's going to be   interesting how it plays out i think you know when  you have robust discussion on it you are you are   right we live in a world of delayed gratification  and and and all those sorts of things um hopefully   it was in one of the other sides and I touched  on it but it's like how you train your people   and the guidelines we don't want to lose critical  thinking constructive debate we don't want to lose
• 24:30 - 25:00 all the things i don't think we will i just don't  know what it looks like right so um I I personally   think there'll be a line that will have to be  like a lot of things where we learn a harsh lesson   somewhere and I don't know what that will look  like unfortunately which is maybe a sad reality   but how do you think I've been using AI throughout  this session you probably looked at my clicker my   watch you know has you got anything on there  you couldn't even work the TV was that smoking
• 25:00 - 25:30 mirrors what do you all think have I been have how  have I been using AI please go uh aside from the   assets like the images and the ones I feel every  slide with text and the table was AI generated   um do you know what from memory I don't think  it was no I think I wrote it oh do you know   what I do a lot of it could have been I write  and then I get it to tidy up because I'm a I'm   a bit wordy yeah I'm a bit It's like what could  be said in 10 words was 20 was written and that
• 25:30 - 26:00 is it but I'm actively using it now agnostic  of slides still using it but it's a good one   but I do use it for I tell you what I did do  with it uh I dropped the slides in when I first   finished it and I asked the custom one I've  got like this is the out this is the audience   this is the outcome I'd like to drive to um this  is this is my experience so it needs to stay in   this lane um can you help me think through it um  it had all my notes in and stuff like that and it
• 26:00 - 26:30 uh critiqued the front so I changed the starting  point it was really good i'd overlooked something   so that's one not not the answer to now but I did  use it that way and I would really appreciate it's   like having a mini coach or something um you got  a microphone on so recording to make a transcript   for afterwards well I think I'll give you that  one but it's not the microphone so on the back of   my phone here is a device called Applaud you see  the little red lights on and what that's doing is
• 26:30 - 27:00 um that's picking up everything in here and  then we I spoke at the national HR summit   um a few weeks back slightly different talk and  what it's doing is it takes the transcript but   it also puts it into session notes it also puts  it into mind map and it does everything for me   and it's all in this little thing here and um  that I can also connect so workflow automation
• 27:00 - 27:30 I can connect to different things i can get it  to summarize so one of the things I actually did   at the National HR summit is we had a little AI  agent running and if they emailed me a certain   um subject line it would trigger the automatedly  send them the notes so just to give you just to   give you an idea on why AI is everywhere where  does it stop where does it start like I obviously   need to declare at the start that's recording but  it does show you and the other interesting thing   is with this as I just nerd on this stuff is if  I'm on a phone call I can actually double click
• 27:30 - 28:00 that and it can record the whole phone call now  that's a real ethical conundrum in a way right   because um uh you should declare it every time  um but I I don't know if everyone will right so yeah it's so yeah and but then also are the notes  correct you have to go through all those things   so cool yeah there you go bit of experience for  you on that um we touched on it there but any
• 28:00 - 28:30 questions for me just generally about ethics AI  how we think through it with you i just feel like   kind of a lot of catching up to you like a bit of  a dinosaur it's Yeah I It's funny right i'm not   going to tell you how to feel but it's like it's  going so quick and it's incredibly hard to keep up   on keep up with even if you work on it 38 hours  a week so it's like the movement is incredible   the other thing I think about is yeah I read all  the main news there's a lot of noise from not our
• 28:30 - 29:00 software areas but a lot of money going in which  creates a lot of noise and a lot of stuff that   I read that I just know it's not agentic yet i  just know it can't do these things and then that   creates the clickbaits and all those things so  you I'm not trying to dismiss what you're saying   but you might be more further along than you think  um and then someone like me who has something like   this and works with it i also enjoy it and I'm a  nerd so I'd be more progressed in that regard so
• 29:00 - 29:30 yeah please obviously most of us in the room at  APS and we follow a lot of guidelines given to   us by usually not our area are you guys working  with the DTA or government to make sure that this   fits into all Yep yep so I was uh with the I'll  give you an example so I was with the APS Academy   yesterday um they're we're doing a roll out  with them they're going to be using in there we   went through full security and risk with them on  everything I've talked about i I wouldn't probably
• 29:30 - 30:00 be as comfortable if I hadn't been through that  process talking about it in this room i'd probably   do it in the private sector room um but yeah it's  making sure it aligns with everything can be used   can be approved um going back to like using the  industry principles and CSRO before that and   making sure we're up to date on all those things  um making sure the AI is in an IRA environment   stuff like that so um yeah um what I will say  is it's like I am constantly learning as well   so there might be little tweaks as I go I need to  make so it's like trying to keep my finger on the
• 30:00 - 30:30 pulse around that was there a question down this  way or I was just curious about where you stand   um on AI yourself like um as you're learning  about AI um what part of you is um saying I'm   um really excited to and really interested to  apply it in this setting but also I'm just curious
• 30:30 - 31:00 about the thought process I've got my Ben answer  there and I know it's I'm trying to think of my   Acorn answer at the same time so I I'll try and  blend them together right um for me we've talked   a lot around how the working week isn't designed  for the world we live in today got my wife and   I we've got a five and an 8-year-old there's 80  hours there we're crushed at the weekend all those   sorts of things and then we all we talk about is  productivity and this 4 day week that might exist   one day and stuff like that um then you see in  2023 the release of the psychosocial hazards and
• 31:00 - 31:30 the compliance we all have to have around that and  why does that exist and all this sort of thing so   when I peel it back what excites me is I believe  it is a route forward for efficient productivity   that brings us all back in line as a society for  better well-being that's what I believe at a high   level i'm not the smartest person in the room to  know how we do that but that's what excites me   why I like it why I like working with um Acorn  within HR is because you can get that direct   line in there a little bit more um but like at  a high level that's what I I believe it can give
• 31:30 - 32:00 us a path forward to that if we do it right and  we don't mess it up basically and I think every   there's a lot of custodians out there that have  a big responsibility to do it right so that's   where I stand on it what do you perceive as risk  the perceived risks so I'll stay stay with I'm a   glass half full not a glass half empty so I look  at the positive so the risks are more aligned with   the video that I shared with you and the CEO of  Anthropic with Claude so the risk here being if
• 32:00 - 32:30 those models aren't safely built the rest of us  can't build off of them so the risk is downstream   and it goes back to supply chain management  and ethics around that since it's always been   so that would be the biggest risk of them not  doing it right um that I would perceive but I   see that as very small at the moment because  of the approach that they're taking that's me   i'm always available i will talk about all of  these topics capabilities skills anytime how you   implement it what you're thinking more worldly  what am I learning what have we done from the
• 32:30 - 33:00 risk assessment security like all these things  it's my job to help so um please even if it's   just I get teams messages everything like that if  you want me to point you in the right direction   or that cybernetic research or any of that just  let me know and just confirming turning it off