The Red Review -Fireside chat on AI with Darrell Woodward
Following on from our well received webinars on the new procurement act and social value, we are pleased to be able to deliver a game-changing series of webinars moving forward.
We have upgraded our usual monthly drop-in Zoom session for June to a guest fireside chat.
At recent sessions, and on training courses, attendees have asked that we get a real AI expert along. We asked our friend and AI thought leader Darrell Woodward to join us for a chat.
In the conversation, we covered subjects including what AI is generally, where we are currently with AI in the world of winning work, what the next 3 to 5 years might look like, what it means for bidding people and what the opportunities are.
Transcript
[00:00.9]
Welcome to the Red Review with me, Jeremy Brim. The Red Review is brought to you by Growth Ignition, the transformation and capability development business, all in the work winning space and the bid toolkit, its product set in bid process and training videos. Good afternoon everybody.
[00:18.1]
Welcome along. So this is one of our monthly drop in calls that we've converted to webinars on I'll stop sharing my screen and move to just video. So we offer these sessions to anybody we've ever worked with, and our wider audience and community.
[00:37.7]
Anybody's welcome to join. Each month, we now run a webinar session like this one, with an expert on a particular topic. And then every other month we have just an open drop in call, where anybody's welcome to join and ask a question, have a conversation about what they're finding in bidding.
[00:58.0]
So the idea is anyone we train anywhere in the world, after we train them are very welcome to re engage with us and tell us how they're getting on, struggles, they might have things they'd like to tune up, questions they've got etc. In those drop in calls.
[01:13.2]
But we found people started to ask about interesting subjects. And so this is I think third or fourth with, we've done a session on the new Procurement act, with my colleague Gemma, who's an expert in that. We've done a session on social value with Carrie Anne Hewlin who's one of the world's experts in delivery of social value.
[01:35.4]
And so, but one of the subjects that comes up more often than not, both in our drop in calls, but also where we're training people in writing bids and capture etcetera, is the onset of AI. And so welcome Daryl. Great to have you along, sir.
[01:51.1]
Yeah, thanks for me. So can you just give us a bit of an intro to yourself Daryl, in the first instance and a bit of background and then we can get into the meat of the subject, as it were. Yeah, sure. Well, firstly, I would never describe myself as an expert in AI by the way, just because, you know, it's a very big and broad and deep subject and I believe where lots of people go in their minds, that it's an assembly of boilerplate.
[02:20.9]
And I tend to see that there's a way of yes, creating pre written content, but then using that within proposal automation software to then ask some sensible questions, about what kind of proposal you want to write.
[02:42.2]
So things like, well, who is the client and what industry are they in? What kind of, you know, what kind of problems does a client in those industries tend to face and what sort of solutions are they looking for?
[02:59.9]
How do my clients, products and services, support that? And then of course the very important you know, can we prove it and demonstrate it and if you can break that up into small enough chunks and ask enough questions then you can start to reassemble that content back into a really compelling story that's, that's, that's specific to that particular client and their opportunity and it only gets you to a, to a first draught.
[03:34.5]
I'm never going to claim that it, that it kind of writes the story for you. But you can start to develop smaller and smaller chunks of content and create more and more options for your, your particular stories and your proposals.
[03:50.8]
And then about 18 months ago of course along came across, along came this thing called Generative AI which when you think about it is really doing that. But it's doing it at almost the per word level where it's, you know, we're asking it to, we can ask it to tell a story about something and it will work on each word and come back with with some sort of output.
[04:14.7]
So I saw immediately that there was, you know this was going to be something that has an effect on our on our industry as I'm sure many of us did. It's sort of something that writes immediately looks like something that should be of interest to us as bid people.
[04:35.2]
And so that's what sparked my initial interest in Generative AI. Excellent. So that lands us nicely in the subject. Thank you Daryl. So yeah, although you wouldn't paint yourself as an expert, you know a damn sight more about it than me.
[04:50.3]
And so I've been avoiding the subject to be honest with you. We've actually shaped our business to be dead in 5 to 10 years time because AI and automation will come for it fundamentally. I've got my first client that has asked to buy our IP so they can feed it into an AI machine that will deliver the training in any language, in video form etc.
[05:15.1]
So as, as much as I will always say in person training is absolutely best and obviously we do loads of work around that in how we set that up for success and how we provide ongoing learning and things off into the future. You know, like many different types of industries and work game's going to change in it.
[05:34.5]
And so I've kind of been putting it off, but I was. I got to a place where I thought we need to have a chat with Daryl about this. People keep asking me about it and I've seen you do some really good stuff on the subject, so. And I particularly, I recognise, Darryl, you work with lots of vendors. It's all part of the job.
[05:50.1]
We're not here to slag any vendors off or get into the details like that at all, actually. And in fact, there are some majestic platforms out there and lots of them. Them. But I wanted to talk to somebody that wasn't one of those, so that we got, you know, an honest appraisal of, of where we are and where things are going.
[06:08.8]
So. But before we go down that rabbit hole too far, what does AI mean? In general con, not just bidding, but just in the world, what's AI mean in the first instance? Yeah. So, I definitely want to get too technical.
[06:24.8]
One, because everyone would switch off. And secondly, because I'm not an AI engineer, so there's always going to be limits. But, I think sort of for normal users of AI, I think it's important to, see it for what it really is.
[06:46.3]
And it's a machine that can recognise patterns and then make predictions. And so if we look at sort of generative text AI like chat, GPT and all of those others, then what they're really doing is going out there looking for patterns in the way that humans, have written and currently do write different types of documents and ways of speaking, in different languages as well.
[07:22.8]
And so they're taking those patterns and then when we prompt them for something, what they're really doing is just making a prediction about what is the most likely next word that we would see based on all of the information that the AI knows and the context that you provide it.
[07:43.1]
So, if I can Borrow you, Jeremy, I'm gonna. I'm gonna make you into my AI for a moment. And I'm gonna. I'm gonna give you a prompt. I'm gonna ask you to complete the prompt if. If, you're up for it. Okay.
[07:59.1]
So. So the prompt is, The glass is half. Oh, right. Full. Yeah. So. So you're making a guess about what the next word is. Right? And it's likely it's. It's full or empty, because what's in your mind.
[08:17.5]
But that full or empty answer that you gave is going to depend on, probably whether or not you're see yourself as an optimist or a pessimist in the classic definition of that phrase. Or, if you were thinking as an AI, you might have been thinking about whether you think I am an optimist or a pessimist.
[08:34.0]
Based on any of our previous interactions, I slightly expected you to come back with one of the smart allergy answers to, that. But, you didn't. But good guess. You're making a guess, but you have no context what that guess is.
[08:54.6]
So you may, Purely a guess based on your own biases and ideas. So then if I say, Jeremy, I'm a pessimist, the glass is half empty.
[09:11.7]
Empty right. Now I've given you some context that you can then start to actually guess what then. And now you're more confident you've got the right word this time, right? Because, because I gave you some context and some, and some. Something to work on other than just a phrase to complete where you were, where you were guessing.
[09:30.2]
And so at the simplest level, that's what generative AI is really doing. It's sort of making a, a guess as to what is likely to be the next word based on all of the knowledge that it's essentially consumed. And the way to then think about that in terms of how we interact with it, it really think of it as a really highly intelligent but completely inexperienced and totally untrained intern.
[09:57.0]
And so if we were to, if we were to deal with such an intern, we would try and give them, we try and give them some information, wouldn't we? We'd tell them in some detail what we were looking for them to do for any task. We certainly wouldn't take an intern and say, look, here's an rfp.
[10:13.6]
Can you go answer that for me? Because whilst they might be really smart and they might have some ideas, it's very unlikely they're going to return the answer that we want. So we tell them, we tell them that, okay, we put them on a course probably and show them like, this is how you, you write RFP answers.
[10:34.2]
We'd show them maybe some previous, RFPs that we'd answer to say that this is the style that we answer in. This is what's sort of expected of you. We give them a framework, a structure. We tell them to kind of like think in a particular way. We coach and help them along.
[10:51.7]
And when we're working with AIs, we kind of need to do the same thing. We need to provide them with Some, some, some context in the same way as I sort of told you, I'm a pessimist, not by the way, I'm an optimist. But, and so then what you start to get is the more detail we give them about, what we actually want them to output, then the more likely it is that our output is going to be right.
[11:19.5]
And that's why sometimes when you see people say, well, I asked AI for, you know, X, and it gave me a completely nonsensical answer. And, and very often kind of think, well, but you didn't give it any real information, you just asked it to guess what you wanted. And that's not really how it works.
[11:36.2]
So, so then going back to our glasses half full, if I was to then say, you know, I'm a physicist, then an AI might then start to be able to say with your physicist, then the glass is half full of liquid and half full of gas. Or if I said, well, I'm a chemist, then an AI could come back and say, well, if you're a chemist, then the glass is half full of H2O and half full of a mixture of O2N2 and various other gases.
[12:05.8]
So depending on the context to give it, the answer is going to change. And I think it's really important to consider that when we're working with AIs. Interesting. Okay. So, yeah, the last 18 months or so has been quite interesting specifically in our world of winning work, hasn't it?
[12:26.9]
With some of the existing content management platforms, bid management platforms, bringing forward their AI widgets or capability, and then completely new startups, coming forward with, with propositions to the market with very large equity backing, in some cases very large sponsorship, advertising, you know, stuff, you know, the, the, a professional association in our space seems to be overrun with the stuff.
[12:57.8]
You can't get away from it. Slightly ironic when most of their members are bid writers who, you know, well, let's see, may or may not exist in the future. So we'll get to that debate, I guess. So, so where are we currently with AI in our world of winning work and particularly bid writing at the moment, do you think?
[13:19.0]
No, I think, I think you're right. So the first thing to recognise is, I mean we are generally in the very early days of AI anyway. I mean it burst onto the scene, what, 18 months ago in terms of, certainly the public consciousness, of what it can do and it's, it's developing at an incredible rate.
[13:40.9]
And so we're, you know, it seems like almost every week there's some new announcement about something different that an AI can do that we've, you know, we've not seen before. And I think there's been a sort of a tendency to sort of certainly initially I think where people were responding to say well you know, well I would never be able to do something that I, you know, something that's important to me.
[14:02.2]
And then I always think that never be able to do is dangerous phrase because we don't know what it's going to be capable of in the future. But I think we're in the very early days as you say, we're trying to find use cases for it in general in our profession.
[14:23.6]
The really obvious one really is we're responding to RFPs, we're writing answers to, to questions. Then it's very obvious that a machine that can write seems like something that is going to threaten what we do, that's going to contribute directly to that final output.
[14:49.1]
And I think whilst that is slightly true, I think there's also huge potential for it to help us in all the other things in, in research that we do in managing our time and resources in our planning work or all the kind of back at back office stuff that we do.
[15:11.9]
There's plenty of scope for that as well. But the real question, the real underlying question is is it going to replace us all as writers as in our, you know, in our fundamental job of responding to, in trying to win work?
[15:30.8]
And I, I think the answer that really is going to be kind of a bit of yes and no, which I'm fully aware is a Weasley answer out of that. But the reason I think that is because certainly at the moment AI is a, machine that can produce words and it can do them very quick, very quickly and it can do them in a kind of, in a, in a sensible order.
[15:56.4]
In a way that looks like, you know, that looks like writing. And But is that really replacing us, as writers? And so if you kind of look back in time, you know, writers were people who had a quill and they would form each letter very carefully.
[16:19.9]
I mean look at the handwriting for, for the past it's people had amazing calligraphy that we don't have now. And then they got real, you know, then, then along comes the printing cross and, and which is a type of automation but it doesn't, doesn't replace the writers.
[16:37.9]
It Just replaces the ability to form letters individually in a more efficient way. The idea behind what we're actually going to write, what's going to go into a book, for example, is still a very human concept. And then more recently we've had word processors which do the same things we're allowed to.
[17:00.2]
Now we're really forming words because we can take entire sections and copy and paste them and move them around and we can have the word processor check our spelling and grammar and all of all of that stuff. But it doesn't actually change the fact that the writer is still the one that's in charge of the narrative.
[17:20.3]
And I think now the same is true as we look at generative AI is that yes, it's automating words and it's automating them really quickly but the story is still ours.
[17:35.7]
It's still us as writers who are going to define what that story is. We're just getting a little bit of help actually. Just physically forming the words on the page. Yeah, absolutely. So it's like, you know, whenever people ask me about this stuff more seriously, I do refer them back to Obama's final speech as President of which a large part of it was about automation and AI and how governments should set up the population for success in its adoption.
[18:07.2]
Because it's, it's going to be. And it is an industrial revolution. But, and people think that's a big scary term but actually like any industrial revolution people were just displaced into other places or different type of work. And it's interesting, you know Martin, my mate Martin Smith's asked a question about Co pilot from Microsoft that we'll come to in a minute.
[18:29.0]
But it's interesting that they've called it Copilot. It's a sidearm, it's seen as a tool to help us. But there's got to be, if you are going to spend money on an AI platform there's got to be a business case, there has to be an efficiency. And so you know it is going to take away the need for so much capacity.
[18:49.3]
We will need less people writing bids and particularly of course filling in SQs or PQQs and all of that kind of stuff. I mean it's, it's quite interesting. At a certain conference in October last year, apparently Crown Commercial Services announced that they have procured an AI platform to be the first reader of bids that they receive.
[19:14.0]
So their robot is going to read your stuff. It doesn't take a rocket scientist to figure out that in time we're going to create robots that talk to their robot best. And you tend to find with CCS that their thinking scatters down through the rest of the procurement world.
[19:32.7]
In fact, there's very much the New Procurement act, the mood music behind that is either use ccs or behave like ccs do, which is terrifying to most of us because they're obviously shit at it, but, at procurement. But, you know, the behaviours aren't great, etc.
[19:49.0]
But it's, it's interesting the journey that they are on and how there are subsets of activity in Cabinet Office around use of AI, adoption of AI to automate those sorts of tasks. So, you know, it's coming, it's about how we set ourselves up for success, with all of that, isn't it?
[20:08.8]
And I certainly think the bit there's. It links to my favourite space of capture, in that it's always been my belief that bidding people are the best armed, that they have the right competencies, they're the right people in terms of behaviours to manage and facilitate capture campaigns of big pursuits, because we understand what's going to be required in the end game of a proposal for a negotiation, ideally or the fallback position of, you know, writing the client's ITT for them and then responding to that, in a way that wins.
[20:45.9]
So I think we bidding people have the right competency to run capture campaigns. So I think that's the. There's two ways to go career wise. It's either learn how to, you know, use this sidearm and be a, you know, a prompt engineer in effect in this space, or it's gravitate more towards capture because the robots can't, you know, replace relationships and trust and they can't co solution specification with client yet.
[21:14.9]
So I'm going to disagree a little bit with your assessment, that it means, you know, there'll be fewer people writing bids or fewer people in our profession, I suppose, generally. Because I would sort of agree that, you know, yeah, there's definitely, there'll be definitely less need to have as many people, writing, forming the words because, you know, AI is going to be able to do that much more efficiently than we can.
[21:45.4]
But I don't think, I don't think it means that there'll be, you know, that the bidding profession is, is going away in, in any sense. Because what, what you end up with is that everybody's going to have AIs writing their RFPs. And then logically, how do you differentiate yourselves?
[22:01.5]
Because it's like it's the same. It's effectively going to be the same mind behind it all, writing, you know, writing those responses. So we need to differentiate ourselves. How we're going to do that is having good people, people who are good storytellers, using AI to tell those stories.
[22:18.3]
And like you said, you know, having elevating the profession into more of the kind of building relationships, really digging into the research behind what's going on and how to win the bid, and into that kind of human stuff that AI isn't going to do.
[22:37.4]
So, yeah, I think it will. It's going to transform, I mean pretty much all professions really, certainly, knowledge professions. But, I think in a really positive way it'll have us doing stuff that's fun rather than typing on a keyboard all the time.
[22:55.3]
Yeah, well, stuff that adds value. Absolutely. And feels valued. I think it's a great. Yeah, sorry, I slightly miscommunicated. You're right. I agree with you and I think it is a great opportunity. So for me, there's something I say in our bid writing course that's quite jarring for bid writers, which is that, if your company has the best solution that delivers the most value for the client, let's say you can save the client $10 million more than anyone else because your base solution in its DNA is more valuable, it's better, it's faster, it's more efficient, whatever it is, then it almost doesn't matter what you write in your bid, as long as it's compliant, you're probably going to win.
[23:41.2]
We need to be excellent in how we write and how we put our propositions to clients because the scoring on bids is getting ever closer. You know, one percentage between us and all that kind of stuff. But fundamentally it's solution that actually wins and it's solution that we're writing about.
[23:57.5]
It's the value that your solution creates for clients that we're writing about. So there's two. From an AI perspective, there's two building blocks then I think, that need to be, great or the best. So one is your, your approach to delivery.
[24:16.2]
So I. E. That manifests itself as your quality management system. So your, if you've got ISO9001, you've got a quality management system that basically says how the business goes to work. And so you could point the AI at that to articulate how the business does stuff in whichever way clients ask about it.
[24:37.3]
So I think that's an enabler for, our robots to talk to the client's robots, et cetera. But the other thing which I, I think you've got a view on D is, is the bid content library. So I've, I've had, Which, you know, is the content that articulates how we, we go to work and how we deliver value for clients.
[24:56.0]
I've had a number of clients say to me, actually, that they were going to invest in a new AI platform, but have realised the sheer scale of the task in transferring their content library into the platform. And actually this takes us back to your core discipline of, you know, content systems and their adoption, and how that's done.
[25:17.3]
Well, the same problem still exists that you've got to run quite a big programme. I can see my friend Catherine Potter's on here, who's a bit of an expert in content libraries. You know, that's, that's a big job, isn't it? So, yeah, is there.
[25:33.1]
So in terms of the. The next question was about the next three to five years, what it might look like. But those, those two building blocks are a key part of it, I think, aren't they? Yeah, exactly. And I mean, I think it's always been true with our knowledge libraries is that, if they're full of rubbish, then people are just going to pull rubbish out of them and they get terrible RFPs.
[25:57.9]
And so managing them is really important. And then when I talk to people about automating their proposals, it's like, well, you can't, you don't really don't want to automate your proposals if your knowledge library is full of. Is full of nonsense because it's just going to write bad RFPs even faster.
[26:14.8]
And, generative AI is exactly, is exactly the same, but to an even greater degree. If you're telling it, our knowledge library is where, you go and get all of the facts about our, our business, but it's actually full of, terrible content, then it's going to automate that even faster.
[26:35.3]
And so, yeah, it becomes quite fundamental then that you, that you get your, your library right? Yeah, absolutely. And Catherine's just asked a question actually around. Is this an opportunity for proposals people to push knowledge management and library investment as a key building block of implementing AI?
[26:55.4]
And I would say it is, it's one. One of the aspects. She would say that, obviously, wouldn't she? But it's. I'd say it as well, but many things. Yeah. Phone these two up. But you know, yes, fundamentally rather than writing responses for every bid yourself live, you're going to use an AI to hone those.
[27:16.4]
But you're going to have needed to have built a decent library and keep that up to speed and you know, adopt all the best practises around, perhaps create a community of practise so you get the experts in your business, you know, helping you storyboard and write the best hero content for your responses and then how you leverage that and how you review that.
[27:37.6]
So absolutely there's a whole, a core building block is, you know, crap in, crap out, isn't it? You've got to have a decent content library. But I think the focus on having the most valuable solution for the client is going to become fundamental.
[27:55.0]
Because it's going to, it's going to expose that more than anything. So what else did we have to talk about? So what, what do you think the next three to five years looks like? Daryl? I know it's difficult isn't it, because things move so fast and the market surprises us with new innovation but can you already see a road map?
[28:13.9]
And particularly actually Martin asked a question about Co Pilot. So Co Pilot's bundled in with 365 and you can protect or inference your data within that. It's not like using Chat gtp. So one thing we didn't say actually is I have had clients say to me that they've been using Chat GTP to help them edit content.
[28:35.2]
Don't do that. Would be some, I think fairly robust advice because that's open source. Anyone can it learns from your content then and I. E. Your competitors could in effect get hold of it. But so it needs. Marty Makes a great point that it ring fences your content but given that most corporates have office it's suite, is that a direction in terms of the adoption use of that or is there, is there a use case for other bidding specific platforms?
[29:04.8]
I don't know how close you are to that in. Yeah. So I mean I think, I think this is this, this, this was, this is the most difficult question at the moment I think because if you think the work that AI in the form that we have it now has only really been around for 18 months and then when you start asking questions like what do the next three to five years look like?
[29:29.6]
That's twice as far away again, as where we've got already. And 18 months ago nobody I think would have seen where we already are now. So it's it's a really big, it's a really big time frame.
[29:46.1]
It's changing so quickly that I think it is, it's very difficult to see what the next three to five years might look like. And I think my, my biggest prediction will be that all predictions will be wrong, including that one.
[30:03.7]
But that said I'm going to make three potentially wrong predictions and see where they go. So I think the first one is we'll see AI models get bigger and smaller.
[30:20.6]
So we've already seen a lot of them getting bigger. Each new release has got more capacity, more ability to reason. They start adding in things like being able to read documents and look at images and we're seeing incredible stuff with video at the moment that's that's coming out.
[30:42.4]
So all the time it's just getting bigger, more capable and more powerful. But at the same time I think we'll start to see them getting effectively smaller. And what I mean by smaller is like you know the other day when Apple announced their version of AI so we're all going to have AI on our phones.
[31:08.7]
And it's a bit like anything, anything else. A bit like sort of the microchip industry, where they used to be huge and you know now you've got more computing power in your microwave oven than landed the module on the moon. So I think we'll start sort of see the same thing.
[31:27.5]
They'll get smaller and and as a consequence then we'll see the AI will just get embedded in everything and it will be everywhere. I think it won't take very long before we don't even know that we're using it or not. Because it will just be part of everything that we use.
[31:48.1]
Which incidentally is why I find that that kind of trend that we're starting to see towards with suppliers asking us disclose where we've used AI or we're going to, in the product delivery. I think that's just going to become harder and harder to, to answer. It'll be like well you know, I mean everywhere.
[32:06.2]
And it had an impact in, in lots of things but it's going to become more and more difficult to actually pick out exactly where AI had an influence or, or didn't. See that because that's interesting.
[32:21.3]
We'd had a question from Sarah on that actually that we're starting to see AI disclosure forms coming out, tenders, you know, obviously for a few years now we've seen particularly CCs asking if you've had a bidding consultant working on your bid and who. So they could see if there's any cross pollination.
[32:38.6]
Clearly AI is the next step in that, isn't it? But that, like you say, as it gets smaller using, using your term, that's going to become harder and harder actually, because it's just part of the DNA of how we're all going to work, isn't it? Yeah, because things like, well, there's elements of AI in spelling and grammar checking.
[32:56.7]
So, you know, do we have disclose. Do we have to disclose that? We have to disclose if we use AI to help us, schedule some of the meetings where we, you know, we brainstormed our ideas for our rfp. Do we, you know, in terms of our, delivery, do we have to disclose that, you know, that we're going to be using satellite navigation in our vehicles when we, you know, when we perform some of these services?
[33:22.1]
And that's got elements of AI in it. I mean, it's going to become so hard to understand. Firstly, it's going to have to be, well, you know, what do you mean by AI? Because it's a really broad term. And then secondly, like, where does it really, where does it really matter and how are we going to pick that out?
[33:37.3]
And I just think that's going to become impossible to differentiate over time and it'll just. Yeah, I'll be asking those questions with yes, probably and everywhere. Yeah, but they're not asking the right question just thinking about it, are they?
[33:53.5]
Because actually if, if you were using Autogen or something as your AI bidding platform, it doesn't matter. For example, sorry. And many other, you know, Lupio etc, all fantastic, I'm sure. But the point is it's ring fenced, it can only see your bid content library.
[34:10.3]
So it doesn't matter to ccs, it's none of their business. And also. But it's not the question they really want to ask. What they want to ask is, you know, is somebody responsible, responsible for what you submitted as a human accountable?
[34:27.2]
Because it sort of doesn't matter that AI was used in the background as long as there was somebody who kind of sat through all of that, all of that material and output that was generated and said, yeah, I'm happy to put that into a proposal on behalf of my organisation to say what they're going to do and I will stand by it.
[34:44.1]
And I think that's really the important thing is, you know, is, is understanding what? You know, are we just going to turn around at the end and say, well, it was an automated response and we didn't mean it, or are we actually, do we actually mean to stand by it? Yeah, it's interesting.
[35:00.0]
So Matt has just put in an actual quote from a CCS tender, which is interesting that AI tools can be used to improve the efficiency of your bid writing process. However, they may introduce increased risk of misleading statements via hallucination.
[35:17.2]
Have you used AI or machine learning tools, including large language models, to assist in any part of your tender submission? What's that mean? Yeah, that's actually a slightly more sensible question. And what, but what they're really asking there is about, you know, what, did you check it?
[35:34.1]
Did you verify what the. Did you use AI and therefore did you know? And then did you actually check and verify that what it's. What it's written for you is correct? And that's the important bit, isn't it? It's like, you know, you're going to turn around and go, no, that's just a hallucination.
[35:50.8]
It's interesting though because CCS are the only client or framework provider I can think of where you probably couldn't use AI as it stands currently to write your responses because they give you so little character count that you have to update the text that you, you can't just use normal English language.
[36:09.1]
You have to invent a mechanism, a sort of coding to get across your points. Because you know, they asked us in one tender, for instance, tell us everything that you're going to do between REBA stages A to C or one to three. And it, and, and it was in 400 words, 2,000 characters.
[36:27.1]
That's, that's most of what you do in a construction project. They received like 500 clarifications just about that. And so yeah, your ability to use AI for that is, I would say limited. But anyway, so there's a slight irony in them asking that.
[36:42.7]
But no, I think the question is about have you. Should be about have you used open source AI, I. E. Other people's stuff or, and, or have you checked it? Is really the question, isn't it? But there you go.
[36:58.6]
Okay, good stuff. So other questions. Sorry, I said I was going to give you three predictions. Oh yeah, sorry, carry on. The third one is I think is about vendor consolidation. So as you said, there's like, there's been a, a burst of AI, new startups coming onto the scene.
[37:21.6]
There's obviously You know, all of the kind of traditional RFP vendors have been racing to introduce it into, into their products as well. And then as I think Martin pointed out you've got AI being added in on the underlying tools that we're using like Microsoft Office, like you know, the operating system for example as well where it's sort of being introduced.
[37:51.1]
So you've got kind of like three levels of pure AI, brand new tech, existing technology with AI then integrated into it and then the fact that we're just going to have AI available on the desktop to help us in daily tasks.
[38:09.0]
And I think in terms of a three to five year prediction that's going to shake itself out and we'll see a lot more consolidation of that. I think some people will like any other bubble, some people will survive and and some won't.
[38:27.2]
And it's not necessarily going to be you know, the biggest that that survive that. It'll be the ones that can best understand a use case and execute it. So definitely, definitely see that over time we're going to you know, we're going to see that there'll be, there'll be winners and losers and And I think that will happen sooner rather than later.
[38:52.2]
Yeah, it's interesting isn't it? It's challenging in knowing when to jump really or how to approach it. So what do you think this is going to mean? That we've sort of touched on it a bit. But what's this going to mean for bidding people and what are the opportunities for them? I guess.
[39:08.4]
Yeah. So I mean I think for me there are kind of three primary benefits for bidding people really. One is increased efficiency, the ability to the ability to certainly to answer some of the questions completely autonomously potentially.
[39:31.6]
And definitely to help us with answering sums. Help us with doing things like research and planning some of our back office bid management activities as we as we work through the process. So definitely going to make us more efficient and save us time.
[39:53.0]
I think there's a big argument for improving quality both in terms of
[40:03.6]
being able to, to critique the way that we've answered a question and look at well did we actually answer the question? Does it make sense? Is it a bit ambiguous? You know and when I, this is something I do with with with AI all the time is ask it to have a look at what I've done and be, be be critical and then I ignore it Obviously, but, but, but it's, it's a, it's weird getting critiqued by a machine.
[40:30.7]
It's, it feels much less personal. And if somebody else was to tell me that I'd written something really badly, but it helps me then to stimulate me to write something that's better. And then with things like doing more research and so on, we can actually introduce more, better answers, with more and better data and information behind it.
[40:55.3]
So I think that's definitely an opportunity, as well. So I think faster, better, and then another one is on the creativity. And I think this is often overlooked, but the ability to use AI to say, you know, give me ideas, And one way that it's really effective is that, because it's just so fast in generating things, then don't ask it for an idea, ask it for 10 or 20 and then take all of those and then extract the ones that you think are kind of look about right or combine a couple or disregard them entirely.
[41:38.4]
I mean a lot of the time I ask AI for ideas about things. I don't use any of its output, but it inspires me to think of something else for myself that I probably wouldn't have thought of. So I think we can do, we can do more.
[41:55.1]
You know, we can, we can work faster and we can do something that's better. But the real question is then what we're going to do with that. Because the danger for me is if, if we have more time, let's not fill that with just doing more and definitely not just bidding more.
[42:13.9]
Because we all know that bidding more is, you know, in most cases is going to be a recipe for just losing more too. And also this is a profession that historically has had a real problem with overwork and late night working and working over weekends.
[42:34.8]
And we need to get some more balance in that. So definitely I think it'd be good if we can work faster and more efficiently but then not use that time just to do more, to use that time to do things better or in more interesting and creative ways.
[42:54.7]
Well, I think those bid functions, grabbing capture, getting ahead of opportunities, use the time to get ahead of opportunities, and even grab it. I spoke to my first head of bids in a long time who's been asked by the business to run their key account management programme, which is quite interesting.
[43:13.2]
You don't get that very often. So get, get up the funnel, get ahead of this stuff because 80% of any B2B. B2G. Business's margin comes from 20% of its clients on average anyway, so act accordingly. You know, add another client to that list, look after those clients, land and expand.
[43:28.8]
You'll smash your numbers to bits and need to do a lot less of this stuff anyway, is our proposition often with clients. So why, why not? Interesting. So how do, how do we engage with it and particularly with suppliers? Daryl, what's, what's the.
[43:44.2]
Obviously they should phone you up, and get some advice and help on how to do these things. But in general terms, how do we get in the game with AI? Yeah, you sort of stole my line there. I was going to say, wouldn't it be ideal if they were all brilliant at capture and were able to get in touch with us and give us those answers?
[44:06.4]
So yeah, I think the first, the first thing is, understand what your needs are. I mean this is sort of reverse bidding, right? We're looking at ourselves now as the, as the client. So really deeply understand what your needs are, where your pain points are in the process, and what your, you know, what you think your goals would be from introducing some technology, I mean any technology actually.
[44:37.2]
Not just, not just AI, but definitely, I mean, I think don't just go for AI for the sake of it. Don't do it because it's cool and trendy and everyone else seems to be doing it. We really need to define exactly what it is that you want to, achieve from it on a personal level as well, is defining some of what it is that you, as a, as the human, what would you want to continue to do?
[45:08.8]
What, what elements of the of the bidding process you want to keep under, under your control and, and part of your remit. And so, you know, if you enjoy writing and that's what you know and you're good at it and that's what you want to, to keep, then by all means look at AI solutions, but perhaps in sort of some of the helping, you plan, helping you research rather than, you know, employing an AI to take away the bit of your job that you find fun.
[45:38.4]
So definitely understand your needs. Certainly, I mean you need to research the solutions. As I said, there seems, you know, there's a lot out there. And it's, it's changing every day as well. So I will, I will give one tip on that, which is pretty much every vendor right now, both traditional RFP software vendors that are moving into AI and the startups nearly everyone is, is offering some sort of guide to AI and how to use it within, within this space.
[46:18.1]
And they are I think worth reading to get a feel for the philosophy behind each of those, those vendors, and how they see AI being applied to the proposal process. So there's loads and loads of those guides out there.
[46:37.6]
I'm trying to, trying really hard actually to put together a sort of a consolidated list of them all and so to help people find them. But it's actually a really difficult task to do. So moving fee set about some stage.
[46:54.2]
Yeah, very good. Okay. So, yeah, similarly I'd add to that in terms of sort of opportunities and how to control this a bit in your business. So I would grab it in your business before somebody else does. You don't want your sales people using AI, probably chat GTP to write responses to tenders and cut you out the process.
[47:17.7]
Clearly we've got a vested interest in controlling this but also doing the best for it. And then, you know, I've been involved in large CRM platform deployments into businesses in the past. Your Salesforce and Dynamics and Oracle and stuff like that.
[47:35.4]
And don't let it lead the project. They are an enabler. They are not great at understanding generally without casting aspersions in case there's any IT people on. But you know, do you, as you sort of alluded to Daryl, you need to understand why you're doing it first and what you're trying to achieve, what the objectives are and then find the right technology to suit that and implement it in such a way to meet those goals.
[48:02.3]
It's a transformation. It's not just a technology implementation. Those are not the same thing. And so there's a whole comm strategy around that. Catherine's made a really good point around training. You're going to have to take anyone that touches the platform on the journey with you, train them in how to use it, but also why, what's the benefit to them, what's the benefit to the business, how to safely use it, all of that stuff.
[48:27.4]
So just, just bear in mind, like any technology implicate implementation, that the technology is just an enabler to a transformation. We're trying to get a business benefit, higher win rate, lower cost of sale, ultimately, et cetera. How are we going to go on that journey?
[48:43.6]
And this AI might be one component in a bigger piece that's going on. But you will also bear in mind of course us as the work winning function are just one part of business. And AI will be being considered and obviously copilot and the baked in stuff will be being considered by IT functions and the business generally probably if you're a document, you know, engaged kind of business.
[49:10.0]
So particularly if you work for a law firm. Obviously lawyers are big into writing stuff down. You know they're going to be looking at this. It affects them as professionals as much as anyone else. And so there may be transformations implement you technology implementations happening in your business already that you're not party to.
[49:29.3]
So. Or that you should be because this might perform part of that or vice versa. So look at their interrelationship with what else the business is, is looking at. But look at, look at it as an opportunity, become an expert in it. You know, socialise yourself with people like Darwell and experts.
[49:45.2]
Read those guides. And try and pick a path through this over the next few years. And try and keep your finger on the pulse. I should say because he's on the call I think still my mate Martin. Keep an eye out. Bid solutions. Have some interesting stuff coming in, in this space that I don't think I'm allowed to say what yet but it's the you know, in terms of where to find your information on things.
[50:09.0]
Keep an eye out on BQ and, and what follow bid solutions on LinkedIn and all that good stuff. And likewise Daryl. Watch this space. Martin said smiley face. I know he's up to something. Sneaky bastard. So he's a very clever man. So where do people find you Daryl?
[50:27.1]
How do they reach you mate? Well probably probably main, main place is going to be on, on LinkedIn. And I'm, I'm quite, I'm. I'm kind of relatively. I'm noisy I suppose.
[50:43.3]
On, on there I I kind of. I think out loud in, in some ways I suppose in terms of, particularly in terms of technology in bidding and AI as well. So you know please do connect with me on, on LinkedIn also have a specialist group on LinkedIn for generative AI and proposals.
[51:08.1]
So there's a place for us all to go within the proposal industry specifically talking about gen AI and what it, what it means for us and, and there's resources in there. I put together a list of available training courses.
[51:27.2]
Generally they're quite, they're quite generic but they're how to use AI or how AI works or what it is. And there's, there's dozens of those on there that you can, you can go and explore. Many of those are free as well so please do use those as well.
[51:44.1]
And then I guess, I don't know. Jeremy, you're going to share my email or. Yeah, well we'll we'll send out an email to everybody that's attended and those that registered with, with your contact details and a link to that LinkedIn group etc and as well as a link to the recording of this.
[52:01.8]
So this will go out as a podcast episode audio only on the major platforms as part of the Red Review podcast. And it will go out on our YouTube channel. So we, we'll stand up a thought on our website a little blog post that's got the links to that and the connections with you Daryl.
[52:18.5]
And we'll go from there. So thank you very much chap. I really appreciate it. I hope people have found that valuable. As I say it, it will go out as an episode on, on YouTube and all of that kind of stuff. I need to think about what's next. We'll have a standard drop in session next month.
[52:36.5]
We have podcast episodes coming with Rin Bennett from the PI community, and some other bits and bobs going on. So keep an eye out. You'll, you'll get emails and all the rest of it. We'll, I'm sure you'll, you'll be sick of the sight of us to be honest. But hopefully you found it useful.
[52:54.2]
And we'll see you at the next time. Thank you very much. Thanks Darryl. Thanks Jeremy. Thank you. Bye bye everybody. Have a good rest of the day.
