In Davos earlier this year countries met to discuss climate change, the war and the economic crisis. So why was the talk over lunch about something else altogether : ChatGPT?

Some say the appearance of ChatGPT ranks alongside the invention of the printing press in terms of its importance to the world. But what about the world of HR? Join the Chairman of Legal Island, Barry Phillips as he discusses the significance of this major development and conclude with their top five examples of how ChatGPT can help you in HR.

As featured in Personnel Today, ChatGPT, and similar AI tools, are considered by many to be the most important development since the invention of the printing press” – Legal Island.

Remember to access all recordings of the Legal Island webinars, you must be logged into your Legal Island Employment Subscription Hub.

You can access the long form content Barry mentions in the recording here

Recording

Transcript 

Barry:  Well, good morning to everybody, and welcome to today's webinar, which is entitled "Understanding the Relevance of ChatGPT to the World of HR: Five Ways It Will Save Your HR Team Time and Money".

And can I say straight away thank you very much to MCS Group for sponsoring this webinar this morning.

My name is Barry Phillips and I am the chairman of Legal-Island, a company that I founded, that I set up in 1998, which, of course, means that this year we are celebrating our 25th anniversary. And in those 25 years, I've seen a huge amount of change and an awful lot of developments relevant to HR and compliance and employment law, but I'm not sure if I've seen anything as important as ChatGPT, anything as significant as this. And I hope I can convey just how important it is to everybody in the next 30 to 40 minutes or so.

I should flag up and say straight away I am not a technician. I'm not a techie. My background is in law. I qualified as a barrister in London so many years ago. I don't care to admit exactly when. But I then came over to Northern Ireland and specialised in employment law and also worked in Dublin for a few years before setting up Legal-Island.

Now, what I'd like to do this morning is just to start by asking everybody to put into the chat where it is that you're watching this webinar from. And if you also want to put in your name or the name of your organisation as well, please do that.

I'm delighted to say we've got hundreds signed up for this, this morning, so it'd be really nice just to see whereabouts you are throughout the island of Ireland or in GB or even further afield as well. So if you'd to do that now, then please do so. Put that into the chat, and I should be able to see the responses coming through in a moment.

I see we've got quite a few in from the UK, and Alan Corcoran in Dublin, and others joining us throughout the island of Ireland. You're very welcome. That's great to see, and I hope you enjoy today's session.

The other thing I'd you to do now, and I don't know that I'll ever have an opportunity to ask this of people again and it's quite an unusual thing to ask, but it's a great thing to be able to ask, and it's simply this. Would you put it in the chat if you have been lucky enough to see the Northern Lights this week? And if so, perhaps you would be kind enough to mention where it was that you saw the Northern Lights.

And I know they have been appearing over the island of Ireland and also in GB and indeed further afield this week. So I'd to take a moment to do that. That would be much appreciated.

I have a personal interest in the answer to that. And the reason for that is that I was trying to see them on Sunday evening. I had a tip-off from a friend of mine who is an amateur astronomer and he said, "Look, Barry, I think you've got a good chance of seeing them on Sunday".

So, like any normal family, we decided to sleep outside on Sunday evening under the stars. And so here we are, myself, my wife, and my 8-year-old daughter. Now, I have to say that is not a fair representation of the whole of the night, because after about half an hour, my wife very sensibly gave up, went back into the house, and crawled into a very nice warm bed on top of a very hot electric blanket and left me there with my daughter.

And my daughter asked me this question. She said, "Papa, would you to fly to the moon?" And I immediately replied to her and said, "Alicia, I would love to go to the moon, but I don't think I'm going to see it in my lifetime". But I added that, "I think you've got a pretty good chance of flying to the moon in your lifetime. And my guess is that by the time you're my age, there'll probably be something like, I don't know, Ryanair flights going round the moon and back. Such is the advancement in technology and transport and engineering".

And as I'm thinking of this, as we were both peering at the moon, it did get me very pumped up about development and technology and how quickly we seem to be moving forwards. Very excited about the life that she's going to experience, and I suppose lamenting a little bit all the things that she's going to experience that I'm going to miss out on.

But I do understand that not everybody shares that same enthusiasm about the level of development at the moment, particularly in HR.

And I was talking to a friend of mine who's a director of an HR operation in a very big company, multi-national company that has operations throughout the island of Ireland and GB and also in the United States. I was chatting to her about this development of ChatGPT, and she wasn't showing me much enthusiasm.

And eventually, she said to me, perhaps out of frustration, maybe a little bit of a hint of anger in there, "Barry, understand something. In the last seven years in this role as HR director of this company, I have had to deal with the consequences of a chaos monkey in the White House", her very words, "of Brexit, COVID, then a war in Ukraine, an oil crisis. And here you are telling me that I need to take note of this big development and I've got to work out its relevance to me". And she said, "My energy levels are really low at the moment, and I don't really think I have time for this".

I would say to everybody in attendance this morning that if you feel the same, I get it. I really do understand it, because the last five, seven years for HR I think have just been the most demanding that I've known certainly in 25 years of Legal-Island. And I understand that people are feeling that it just seems to be one thing after another.

But I hope that I can persuade you that this thing, ChatGPT, is to be treated as a friend rather than a foe. And it can help a lot of us in HR and improve HR departments as well as our own personal world.

I put this slide up simply to remind me that, for some reason, I seem to have problems saying "ChatGPT". So occasionally I slip into calling it ChatCharlie. So if I slip into that language this morning, then you'll know what it is that I'm talking about.

But before we do that, a quick survey. What I'd you to do is just to answer this question, and it's as follows. "Which of these game-changing products has been the most important for you in terms of improving your life or your lifestyle?" And if you could just respond to that now, then hopefully I'll be able to see the results of that, and you'll be able to see them too once you've responded as well.

And the big leader is my favourite and would've been my choice as well, which is the iPhone, because this thing has changed my life substantially.

And it's amazing, 25 years ago, if I had wanted to buy or have the capability to do what this thing would do, I would've had to have gone into Dixons or to Currys or somewhere on O'Connell Street or Grafton Street and basically bought a camera, a voice recorder, an atlas, roadmaps, and all the sorts of things that you can now get in this, in one place, and have paid an absolute fortune for it.

I was using ChatGPT yesterday to see what the cost of a 1-megapixel camera was 25 years ago, and it was near £2,000. And now this comes as standard at 12 megapixels. So it's much, much cheaper and it is just much more effective.

But with any development, with any breakthrough technology, there are always a few problems that are created with the breakthrough. And certainly, with this iPhone, it's changed one thing in my life, possibly not for the better, and that is that whenever I go with the family at the weekends to beauty spots or what have you, prior to this thing we would just enjoy the view.

Now I just seem to be taken up with taking photographs of my wife and my daughter in front of these lovely places or posing for photographs, or chasing down people who are willing to take photographs of the three of us.

It's a very different experience, and I don't think it's a better one from that point of view. But I'm sure if my wife was here, she would say, "Well, yes, but you're always on this texting people and emailing and so on".

So next thing is . . . if I can move this on. Yes. Okay, this is a game changer. And before I explain why, I just want to try and explain best I can what ChatGPT is.

As I said, I'm not a techie, so I asked ChatGPT what it is and this is what it told me. It said that it is a chat-based pre-trained transformer and is a powerful large language chatbot. And it's based on GPT-3, which has been trained on hundreds of billions of words from the internet, but also in other places as well, other databases that don't quite make it to the internet, just to add to that.

It's founded by OpenAI and a guy you may have heard of called Elon Musk and also Sam Altman. And between them, if you'd like, they could be said to have founded OpenAI, which has created ChatGPT.

Now, what do we know so far about this? It's had a very impressive debut. It has had a million sign-ups in the first month, far more than Facebook got, TikTok, Instagram, any others. So the traction has been very, very rapid and huge.

It is worth knowing there are alternatives out to this already. ChatSonic is something that a lot of people are talking about. Jasper Chat I've had a go at last week, and it was quite good. Bard AI, which is Google's response to this, was soft launched a few weeks back and hit the rocks almost immediately. And so they're going to have another go at that, I understand, a little bit later on.

DALL-E is the sibling of ChatGPT from the same organisation, but it works in a different way in so far as it's not designed to produce chat back to you, or written content back to you, but an image.

So you can go to DALL-E and you could put in the chat bar, "Produce me an image of a crazy family in the back garden sleeping under the stars", and it will come up with something. And you can play around with it and you can say, "Do it in the style of van Gogh", or what have you. It's worth a look at because it's quite crude at the moment, but it's going to get better and better.

On the side of getting better and better, ChatGPT 4 is expected out sometime this year, maybe next year. And depending on the IT journalists that you read, it's going to be 50 times stronger or it's going to be 500 times stronger or better. And I think that's important to notice, because in a moment, I'm going to talk about the limitations of this.

But just recall Google Translate. Google Translate, when I first had a go at it some 10 years ago with my wife who is Russian-speaking, we were trying a simple English sentence in it, which we would then ask it to translate in Russian. And my wife would almost laugh at it that it was such a poor translation. And we'd do it back the other way and I'd look at it and think, "Well, I know what it's trying to do here, but this is not right".

Now she uses it for professional translation work. She puts it all into Google Translate and she basically just corrects a few errors, and a translation that would've taken her two hours she'll do now in probably the best part of five minutes. So it's an indication of just how quickly these things do develop.

So what do we know right now about the limitations of this? Knowing too that it is going to improve, but right now what is it that we've got? What can it do and what can't it do?

Well, first thing is it's trained on out-of-date data. The data that it is trained on is up to date . . . I think it's September, possibly November, of 2021. So that's the first limitation.

But it doesn't mean that anything that happened last week it hasn't a clue about, and that's the interesting thing. That's because it also works off the input of the users and trains itself off that input as well.

So there is a chance that it will know about the war in Ukraine. There is a chance that it will know about the agreement of the Northern Ireland Protocol, if I'm not talking prematurely there. But the information it's likely to have is going to be lighter than anything that predates the end of 2021.

But this is the first red flag, and there's going to be a number of red flags in this area. The first red flag is that it is swallowing up the data that you are putting into it and it's not clear at the moment how it is looking after that data, and it's not clear exactly what it is doing with that data and when it may be reused.

So, for now, we have to say whatever happens, don't put personal data into this because there's a chance it will gobble it up and it will, pardon the expression, puke it out somewhere else, possibly for a user in another part of the world who has nefarious intentions with that personal data. We don't know. But that's the first red flag, and there are a number of others.

Another limitation is, for some reason, it makes things up. And this is an interesting thing. When you're trying to compare this AI machine to humans just to see how close it is, maybe it's this characteristic that resembles human beings the most, and this is our propensity to make things up if we don't know the answer.

This is different to Alexa because if Alexa doesn't know, Alexa will just say, "I'm sorry, I do not know the answer to that question". ChatGPT doesn't. And you would think that something so powerful would be able to sort of say, "My information on this is limited", or, "I do not know the answer to this". But no, it makes things up.

It's different there to Alexa, so in one sense, Alexa is a safer alternative in so far as at least Alexa will tell you when it knows it and when it doesn't.

Just to illustrate this, I want to step out and talk about something outside of HR in this case, just to illustrate how it can cause problems and confusion and mislead people.

John Lee was a famous criminal from the English legal history and folklore. He was convicted of murder in the Southwest of England in 1885 and he was sentenced to death for the murder of his landlady. And they tried to execute him on three occasions. On each occasion, the execution failed. And there have been all sorts of speculation since then in terms of why the execution did fail.

I asked ChatGPT, or ChatCharlie, on Monday, and the answer it gave on Monday was that . . . Pretty much it was saying it was divine intervention, which is what John Lee had claimed all along.

Then when I asked it on Tuesday, it came with this, and it said, "He had an unusually strong vertebrae". That's an explanation I have never seen anywhere, so I can only think that it made it up.

And then it said, "It's faulty equipment". And it is probably something to do with it, but it's more likely that it was down to an incompetent bungling executioner called James Berry, who was an alcoholic and he was probably highly intoxicated when he was called upon to exercise his dark craft.

John Lee became famous as the man they couldn't hang. They commuted his death sentence to life in prison. He came out 20 years later, and in the early years, he made a living out of making guest appearances in pubs in the East End of London with people queuing up to shake the hand of the man they couldn't hang.

When I asked ChatCharlie "Where did he die?" the answer came back and said, "He died at age 84", which was right, but that he died in a nursing home in Devon, which is wrong. He actually died in Milwaukee, in the United States.

Now, why this can be so misleading is that sometimes ChatCharlie can produce information which you know to be right and accurate, but in amongst that, it drops in things which are horribly wrong. That's why this is, at the moment, so dangerous or certainly so unreliable.

And when I asked Alexa this morning, I said, "Alexa, who was John Lee, the man they couldn't hang?" She tried to convince me that he was the guy that invented the pencil sharpener in America. So it is quite nice, as a human being, to know that between Alexa and ChatCharlie, they are still struggling to get an awful lot right and to know a lot more than I do.

Despite those limitations, this represents a major breakthrough. And these major breakthroughs don't come round very long, but this is like Google on steroids, and it's like Google on steroids, or even better, for two main reasons.

The first is it works, and I'll show you some of the results shortly. They are just absolutely astonishing. And the second reason is it can build on previous inputs. So you can ask it something, and then you can say, "Based on your answer to that, what would you say about this?"

And this is when it begins to get scary because it feels like you're dealing with a human being rather than a machine. I think it's this that is the breakthrough and this that makes it different to everything else that's been before it, and certainly far much better than Google.

I understand that when Google first saw ChatCharlie, they issued a red alert and went on to high status as though to say, "Look, if we don't respond to this, this could wipe us out". So that is why it's so important.

Now, just a bit of housekeeping before I take you to the five areas, and it's simply this. I'm not going to do a live demo of ChatCharlie today, and the reason for that is that I've been playing around with it for the last two or three weeks, and most of the time it's reliable, but like all these things, when you want it to be completely reliable, that's when you have problems. So what I'll be doing is showing you screen grabs of uses that I made of ChatCharlie last week.

Now, some of the answers are quite extensive, and you won't have time to read the answers in detail, but the good news is that we will send you a link after this, which will be a link to our web page, where you can see all the prompts that I've put in and all the answers in their full format and you can read them in your own time. So that's one thing.

The other thing is I'm going to tell you about a Part 2 to this webinar, which is going to take place later this month, in which we'll look at these red flags. So there's going to be information on that, which is really important, at the end of this webinar.

This image, by the way, was created by DALL-E. And to get this image, I put into DALL-E, "Please, produce an image which illustrates housekeeping and do it in the style of van Gogh". And this is what it came up with.

But it's interesting because I had to ask it three times to produce an image that was gender-neutral or gender-balanced, and it was only after the third attempt that it gave me this. First two attempts were just women only. So already with AI, there are, of course, D&I . . . got a lot of acronyms coming around here . . . issues that we need to be thinking about.

So here are my five. I'm sure there's another five. If you've got five of your own, they could be the same as this. They could be different. So what I'm going to do is to show you how I used ChatCharlie recently in this area of recruitment in the hope that you might pick up some ideas of how it might be useful.

So what I did, I asked ChatCharlie, "Draw up a job description for a knowledge worker at Legal-Island". And in doing this, I felt a little bit unfair because "knowledge worker" at Legal-Island is a term that we haven't really used before the end of 2021. So I knew, or at least I thought, it would put ChatCharlie on the back foot. But none of it. It came up with this straight away and it came up with information that was immediately useful.

Now, I'm not saying it was perfect, and this is the important thing I think to understand. But I do know how, in HR, we very often work on what's gone before, be it templates or previous, in this case, job descriptions and so on. So this is a way of producing a very good first draft of something that you can then work from and tweak and can then sign off as good-to-go.

And you'll see there it's set out a whole list of things that would be relevant to the post. I must admit, it's scarily accurate. There are a few things that I would have to change, but it was very, very useful.

The next thing I did, and this is how you can build on previous conversations, if you like, with ChatCharlie, is that I said, "Draft a set of interview questions that might be used for candidates for such a role". And it came out with, I think, nine altogether. These are seven for you. Again, you may not have time to read all of these seven, but you can access them in your own time when we send you the link.

And what I did after that was I said, "What sort of questions should not be asked? Please provide a few examples". And immediately, it went to the protected characteristics and it was highlighting race, gender, age, and questions surrounding that. It also advised me not to be overly personal or invasive with questions and so on. But it also, at the end, cautioned and said, "Please take legal advice before relying on this".

So, again, this is something that's useful, but if it's as serious as a job interview and questions like that, then you would be wise to take legal advice before relying on those questions. So that's one area.

Another is in learning and development. And what I asked it to do in this area for HR was, "Devise a learning and development plan for someone in the role above". So the role above was the knowledge worker at Legal-Island.

And it said as follows. "A learning and development plan for a knowledge worker at Legal-Island could include the following components", and it set those out nicely. Then it put out something like nine responses. And again, you'll be able to see those on the website.

Now, in learning and development, I know there's an awful lot of classic books around, be it business development or other areas relevant to HR or self-development and so on. And one of the things that I regularly do is to try and get summaries of these classic books.

And so what I did with ChatCharlie last week, I said, "Please summarise Peter Drucker's most famous book in no more than eight bullet points". So what it was able to do there was to work out which was the most famous, and it's clearly "The Effective Executive". I'm sure a lot of people have heard of that. And it summarised it quite brilliantly and almost instantaneously in eight key points.

Now, you can have a bit of fun with this, and you have to be careful it doesn't take you over, because I felt last week at certain points it was doing just that.

So I said to it, "Do the same summary, but do it in the style of William Shakespeare". And this is what it came up with, language very reminiscent of that period of time but still very relevant to the points that it had been asked to make.

And when I just say, "Don't get carried away with it", I think I did last week because I asked it to deliver the famous Lincoln Gettysburg Address but in the style of Donald Trump. And I'm just going to stop just to give my voice a break and give you a break from my voice and allow you to read those two paragraphs, because I think they're quite good fun.

So onwards to HR policies and procedures. The next thing that I did was to ask it to devise a gifts policy for my company in Ireland, which has 500 employees. And it produced this. Within just a moment, just a few seconds, it was able to do a number of things. It picked up on Ireland, so it was later on talking about Irish employment law. It dropped in the proper currency, so it went from dollars to euros. And it did finish by advising with the usual legal caveat, "Please consult your company's legal team before modifying and relying on this policy".

So I think the lawyers in the group would say, "Look, you can't rely on this", and I would agree with them. But if you want to send something to a legal team and you don't know where to start and you have nothing to use as an earlier form of something, then it would seem to me to be a lot better to send this to them and say, "Would you check it through once you've made adjustments?" rather than send it to them and sort of say, "Would you devise something?" So I see a lot of time being saved in this sort of area.

Time for a quick poll again. The question that we have for you now is, "Do you work in an organisation that has an AI use policy for the workplace?" You've got options of "yes", "no", or "not sure". But if you could just attend to that, that would be great. I'll just give you a few seconds.

This answer doesn't surprise me. I was expecting an answer in the 90s for this, but I'm sure it's something that when we look at the red flags at the end of this month, our legal team will be advising on in terms of an AI policy that's suitable for today and advancements like ChatCharlie that we have seen.

I was going to say well done to the 1%, but we don't even have 1%. But we have 7%, so maybe there were a few in the "not sure" category that do have something in there already.

So this does suggest red flag warning here because this is an area that does need attention, the AI use policy. And as I said, we are looking at the red flags here on 29 March.

Now, people who know me will know, I hope, that I'm not known for being over dramatic, but I do want to say here that I think this is probably one of the most important webinars, if not that we've ever done, certainly that we've done in a very long time.

So I would say to everybody, please, would you put that in your diary? Make sure you are available to attend that and await news on it and how to register from us by email, because it's such an important webinar to attend.

Item number two is report writing. And this is how I played around with it last week. First of all, I started by asking it, "Check this sentence for syntax errors. 'There are no elephants pink'." And it replied, "The sentence 'There are no elephants pink' is not correct in terms of standard English syntax. It could be corrected to 'There are no pink elephants' or 'Elephants are not pink'." So good first start.

I then put into it this paragraph with a good few spelling mistakes, syntax mistakes, and I said, "Please improve on this paragraph". So I gave it quite a wide remit. And what it came back with was something which I think was a significant improvement. It had replaced the spelling mistakes, it had made adjustments for syntax, it had changed passive construction into active construction, and generally it had improved the whole paragraph.

So there's an obvious use for this here for report writing. But another red flag here is we have to understand that if you put in a report that this gobbles up, this could end up in the public domain. So you have to use this with caution knowing that that is the case for now.

Future iterations of this, it may be possible for organisations to buy their own private version of ChatCharlie. We're all trying to work out what they're going to do with it, and maybe they are as well, but for now that's the big flag that goes with this.

I just want to take a quick detour now outside of the HR department and just flag up where I think it's going to be relevant in other areas of an organisation.

The obvious one is the marketing department. As some people may know, we're currently experimenting with a four-day week here at Legal-Island. We're about two, three months into the six-month trial.

And a couple of weeks ago, we wanted just to put out a press release on this to let people know about it. I went to ChatCharlie and I basically said, "Please start a press release for Legal-Island informing everybody that we are on a trial of the four-day week. And please finish it by mentioning other companies that have successfully and permanently adopted the four-day week. And please do this in no more than 500 words".

It came back with something which was pretty much good to go. So something that would've taken maybe two hours to fiddle around with and to get right took really 10 seconds and would've taken a couple of minutes just to improve on and to tweak before sending it out.

Software development is another obvious area for this. As I said, I'm not a techie, but I do read quite a bit in this space. And a lot of people are saying this is not going to make software developers redundant, but it is going to change their role to some degree.

Back to number one, the number one that I have for you here is to suggest, "Could this be everybody's personal PA?" So, in a way, it's a way of saying that this could be relevant to everybody. If you have colleagues who say, "Oh, it's not relevant to me, nothing to do with me", you could go back to them and say, "Well, it's relevant to everybody. Take a look at this".

There are two ways that I've been playing around with this, first of all. And one is this morning I dropped in a request to ChatCharlie and I said, "Would you please tell me about the train times from Dublin to Belfast for today?" And it got back to me and reminded me that it's only up to date as of to the end of 2021.

But it then said, "But I think you will find these three web links useful", and it linked to Enterprise service, it entered a link to Translink, and I think it gave me a link to a bus service as well. So even when it's out of date or working off dated data, it still can be useful. It still can save time.

The other thing that I've been experimenting with is drafting emails. I've read a lot about this, and I was a little bit dubious about it, because I've tried it before with speech-recognition technology and it's basically taken me more time to do something through speech recognition than it has actually saved me time.

But then I experimented with a Google Chrome attachment, and I will send you details of this in a follow-up email, in which you can actually attach this to ChatCharlie and you can then speak to it rather than to have to type in everything. And that changed things for me because I'm not a particularly fast typer.

I had a go a few days ago and I just said to ChatCharlie, "Please, would you draft an email to Lindsay inviting her to meet me online next Wednesday at 12:30? Please ask her, if this is not convenient, to give me an alternative date or time". It took me 20 seconds to say that. That was typed out in a few more seconds by ChatCharlie, I cut and pasted it, and sent it away.

Now, I think probably that would've taken me about five minutes to do with my typing as it is. And I do find when I type things, I tend to get a bit distracted and I jump around from tasks, whereas with that, I was just focussing on that one thing.

So I think that is something that can save everybody time. I'm still playing with it, but I would encourage people to drop in the Google Chrome extension so that you can actually talk to it as well as type into it.

So the red flag warning, and it's this. We are doing this webinar on 29 March. It is going to be delivered by the legal team at A&L Goodbody that you'll know is an island of Ireland top law firm. They are working on this, quite possibly as we speak, to help us all understand how to regulate the use of ChatGPT and to basically flag up the problems with it.

I'm using the word "flag" a lot, I've noticed, at the moment. But hey, we like flags where I am, so maybe you can forgive me for that.

We have this event, again, on 14 March. So if you have a colleague that you think should see this webinar, it's going to happen on 14 March at 1:30, and it's basically going to be a repeat of this live. There's information on that on our website if you would to share that with your colleagues.

So just to round out the story, I have to say unfortunately we never did see the Northern Lights. We went out on Monday to our local lough, Lough Neagh, and no sign of it there.

And then we got a text from somebody in Toomebridge who was saying that they could see it clearly there. So we jumped in the car and shot over to Toomebridge, and there was not a critter to be seen in Toomebridge and certainly no Northern Lights. So we're still looking, but we're hopeful for tomorrow because the weather forecast, I'm told, is better.

This is what we missed. It's a photograph I think taken from the South of Lough Neagh. For anyone who is in attendance at this webinar not from the island of Ireland, that's what you're missing. So if ever you fancy a weekend away, jump on a plane to Belfast, Dublin, Galway, Cork, whatever, because it's magical. That's my little bit for the Irish tourist board done. Free of charge, I might add.

Any questions that you have with this, because we're out of time as expected, please drop them to me and I will do my best to show them to our lawyers here and to get back to you if we can.

Any techie questions will go to a tech expert that I know that I'm hoping to line up for the event on the 29th as well.

And just finally, I'd like to say this to everybody, if I may. This is something that we have organised free of charge, and the same on the 29th, because we just think as many people in HR need to see this and understand this as possible because it's really important.

But the red flags are big issues too that everybody understands. So we're doing both of these free of charge in the hope that we can get the message out to as many people as we possibly can.

If you feel that this webinar today has been of value to you, then perhaps we can ask for just one thing in return. If you could do this in the next two minutes, I would be eternally grateful, and it's this. Just grab a bit of paper and write down on the paper in front of you which of the five that I've covered you would see as being most useful to you in your work. And if you could just write that down, that would be great.

At the same time, if you could also just add to that what is in your head right now about ChatGPT and the impact that you think it's going to have on the HR world and write that down, that would be much appreciated.

I just want to give you 60 seconds of your own time just to do that and to finish that off. And then I'll come back to you in just a few moments' time.

If you've done that, what we'd to do is to use what you have in front of you, and we'd like to see that because this is going to help us form the event on the 29th. But I think other people outside of this webinar would like to see it as well, because I think it's really quite important.

So what we'd like you to do is to give us that feedback. Not through the chat, but on our LinkedIn page that I'm going to take you to in a moment, because I'll provide you with the link to this, or rather my friend and colleague Maria will do this. So, Maria, if you could just provide that link to our LinkedIn page.

Then what I'm going to ask you to do is just to click on that, and whatever you have in front of you, put that into the comments box of that page. We've never tried this before and we're curious in terms of how it's going to work.

I think that will be useful to us, to other people, and it may even be a record that we can actually look at and return to. We'll do a print of it, a screen print. So who knows? In 25 years, if we're still in existence here, we may be able to come to this record and this print and take a look and see which of your predictions about ChatGPT came true and which were a little bit wide of the mark.

So I'm going to click on that link too. I'll see you on the outside at that LinkedIn page.

That just leaves me to thank you for attending. I hope you found that useful. Thank you, Maria and Arnold, for being such great support for this webinar for me this morning. And I look forward to seeing you on 29 March. Thank you.

Sponsored by: