1 00:00:00,280 --> 00:00:02,400 Hi, and welcome to the Neil Ashton Podcast. 2 00:00:03,080 --> 00:00:05,920 In each episode, we explained some of the fascinating ways 3 00:00:05,920 --> 00:00:09,080 that science and engineering are changing the world around us. 4 00:00:09,800 --> 00:00:12,760 We talked to leading engineers from elite level sports like 5 00:00:12,840 --> 00:00:16,880 cycling in Formula One to some of the world's top academics to 6 00:00:16,880 --> 00:00:20,960 understand how fluid dynamics, machine learning, supercomputing 7 00:00:21,360 --> 00:00:23,000 are bringing in a new era discovery. 8 00:00:23,960 --> 00:00:27,080 We also hear some of their life stories, their career advice, 9 00:00:27,640 --> 00:00:30,080 the lessons they've learned on the way that I hope will be 10 00:00:30,080 --> 00:00:33,800 helpful to you too. So sit back and enjoy this 11 00:00:33,800 --> 00:00:41,000 episode. Hi, and welcome back to the Neil 12 00:00:41,000 --> 00:00:45,160 Ashton Podcast. So today I wanted to give 5 tips 13 00:00:45,320 --> 00:00:49,320 for CAE engineers in the era of AI. 14 00:00:50,120 --> 00:00:54,640 Five things that I think will be useful for you from a career 15 00:00:54,640 --> 00:00:58,480 point of view and from hopefully making the most of what I think 16 00:00:58,480 --> 00:01:01,240 is quite an exciting new technology. 17 00:01:02,120 --> 00:01:05,640 You may disagree with some of these things, and if you do or 18 00:01:05,640 --> 00:01:10,240 have other tips that you think people should should adopt, then 19 00:01:10,520 --> 00:01:13,560 you know, feel free to leave a comment if it's on YouTube or 20 00:01:13,560 --> 00:01:15,920 LinkedIn or wherever you're listening to this. 21 00:01:16,600 --> 00:01:21,880 OK, so let's get started. First of all, I'd say an an open 22 00:01:21,880 --> 00:01:26,800 mind would be tip one. It's very easy to come from a 23 00:01:27,040 --> 00:01:31,040 negative judgmental viewpoint when it comes to AI and it's 24 00:01:31,040 --> 00:01:34,080 natural. We as humans often, you know, 25 00:01:34,480 --> 00:01:38,440 see things that are pushed at us and sometimes we have a 26 00:01:38,440 --> 00:01:42,200 temptation to, you know, be skeptical of some of this. 27 00:01:42,800 --> 00:01:48,200 But I think that is probably the the worst attitude you can have. 28 00:01:48,320 --> 00:01:52,640 I think having an open mind and being willing to look and listen 29 00:01:53,920 --> 00:01:58,000 and read is important. At the end you may still make 30 00:01:58,000 --> 00:01:59,680 the same judgement as you did at the beginning. 31 00:02:00,000 --> 00:02:05,000 But I think most people who actually look into this do end 32 00:02:05,000 --> 00:02:09,199 up changing their mind and and they understand better where 33 00:02:09,560 --> 00:02:14,800 some of maybe the marketing or hype is over egged and but where 34 00:02:14,800 --> 00:02:19,240 there are actual benefits. It is very tempting, of course, 35 00:02:19,240 --> 00:02:21,920 if you have 30 years experience, you know, like some people have 36 00:02:21,920 --> 00:02:29,640 in in CFD or FEA, to almost be insulted by some young, you 37 00:02:29,640 --> 00:02:32,680 know, 20 year olds doing ML research and suggesting that 38 00:02:32,680 --> 00:02:34,600 their method, you know, can be better. 39 00:02:35,880 --> 00:02:41,440 But sometimes that's from a good intent place, they're excited 40 00:02:41,440 --> 00:02:45,000 and maybe they don't know or they're not from the CE domain. 41 00:02:45,000 --> 00:02:49,520 And, and so having an open mind, and this is true on sort of both 42 00:02:49,520 --> 00:02:53,080 sides on the CEA side and the ML community is important. 43 00:02:55,080 --> 00:02:58,920 The second one, which you sort of need the first one to get to 44 00:02:58,920 --> 00:03:02,040 the second one is educating yourself. 45 00:03:03,720 --> 00:03:06,280 You know, a lot of companies call this continual professional 46 00:03:06,280 --> 00:03:09,640 development. And I think this is very 47 00:03:09,640 --> 00:03:14,800 important in the era of AI. The problem and I which I think 48 00:03:14,800 --> 00:03:21,880 creates sometimes this push back is because the, the ML community 49 00:03:21,880 --> 00:03:24,520 and the CAE community come from very different backgrounds. 50 00:03:24,520 --> 00:03:27,360 They often have studied different courses at university. 51 00:03:27,360 --> 00:03:30,600 They, they have a different preference in terms of software, 52 00:03:30,600 --> 00:03:33,280 in terms of programming. They, they're, they're two 53 00:03:33,280 --> 00:03:38,040 different communities and the challenge for CAE to look at AI 54 00:03:38,040 --> 00:03:41,120 is that it all seems quite foreign and complex. 55 00:03:42,080 --> 00:03:45,280 And This is why I think educating yourself is important. 56 00:03:45,640 --> 00:03:49,120 Going straight into a journal paper describing an ML 57 00:03:49,120 --> 00:03:51,280 architecture may be a bit overwhelming. 58 00:03:51,280 --> 00:03:54,040 It certainly was for me at the beginning, but there's a lot of 59 00:03:54,040 --> 00:03:58,080 content now that helps you. So Coursera is one of those 60 00:03:58,080 --> 00:04:02,520 platforms where they have some great courses by lots and lots 61 00:04:02,520 --> 00:04:05,480 of people from different, you know, walks of life and 62 00:04:05,480 --> 00:04:08,680 different backgrounds that can really help you. 63 00:04:08,680 --> 00:04:11,680 Your company may already have a subscription or if you don't, 64 00:04:11,720 --> 00:04:14,840 you know, like I did it in the past, I think paying it yourself 65 00:04:14,840 --> 00:04:18,680 is a worthwhile investment. But beyond these sort of 66 00:04:18,680 --> 00:04:22,480 certifications like ABS does them, other tech companies do 67 00:04:22,480 --> 00:04:26,120 them. A lot of conferences, especially 68 00:04:26,120 --> 00:04:29,520 with AI topics, have a real push around transparency. 69 00:04:29,880 --> 00:04:32,520 And so I found that many of those conferences actually 70 00:04:32,520 --> 00:04:36,480 published their entire talks online, which is fantastic 71 00:04:36,480 --> 00:04:39,920 because you can actually watch on YouTube or whatever platform 72 00:04:39,920 --> 00:04:43,040 you prefer Many of these talks, you can speed them up. 73 00:04:43,040 --> 00:04:47,640 You know, if you'd want to get through quickly, you can also 74 00:04:47,720 --> 00:04:50,800 use AI. Remember to summarize papers. 75 00:04:50,800 --> 00:04:56,080 This is something I do a lot. So for example, if you have a, 76 00:04:56,320 --> 00:04:59,600 you know, a paper like attention is all you need or on 77 00:04:59,600 --> 00:05:03,480 Transformers or maybe something specifically on CAE related AI. 78 00:05:04,160 --> 00:05:07,120 Now, most of these AI large language models can do a pretty 79 00:05:07,120 --> 00:05:11,520 good job of explaining things. And if you say, if you upload 80 00:05:11,520 --> 00:05:14,200 the paper and say, please summarize the this to a lay 81 00:05:14,200 --> 00:05:19,400 audience or explain to me why how this bit works. 82 00:05:19,400 --> 00:05:22,480 Or can you can you give me more examples of this? 83 00:05:22,720 --> 00:05:28,000 It's amazing how these AI can actually help teach you to learn 84 00:05:28,480 --> 00:05:33,120 AI, but you can't get away from sometimes speaking to people. 85 00:05:33,120 --> 00:05:36,480 And that's one reason I would definitely try and broaden 86 00:05:36,480 --> 00:05:41,120 yourself and go maybe to even AI Pacific conferences like Nurips 87 00:05:41,360 --> 00:05:44,280 and others. Come and go to some of the 88 00:05:44,280 --> 00:05:46,560 workshops, listen in. You might not understand 89 00:05:46,560 --> 00:05:50,400 everything, but you'll slowly get and build the network. 90 00:05:50,400 --> 00:05:52,800 You'll start the people one-on-one and ask stupid 91 00:05:52,800 --> 00:05:55,080 questions. People who know me, I often I'm 92 00:05:55,080 --> 00:05:57,680 in meetings and I'll say, I'm sorry to ask a stupid question, 93 00:05:57,680 --> 00:06:00,760 but and often times it maybe not. 94 00:06:01,040 --> 00:06:02,400 It's a completely stupid question. 95 00:06:02,400 --> 00:06:06,080 And regardless, they usually explain something in a easy to 96 00:06:06,080 --> 00:06:09,120 understand manner that it would be difficult for me to get. 97 00:06:09,520 --> 00:06:12,720 And sometimes all you needed is a one or two key concepts. 98 00:06:12,720 --> 00:06:15,720 And once you get the concept, you're like, OK, I see it now 99 00:06:16,080 --> 00:06:20,120 where at the beginning it it, yeah, seems a bit foreign. 100 00:06:20,600 --> 00:06:23,480 And the final ways podcasts, obviously I'm biased, don't make 101 00:06:23,480 --> 00:06:25,560 my own, but there's a lot of people creating them now, 102 00:06:25,560 --> 00:06:29,600 interviewing people and those again could be a a great thing. 103 00:06:30,040 --> 00:06:32,400 Finally, of course, depending on, you know, where you are in 104 00:06:32,400 --> 00:06:35,040 your career, going back to university or taking the 105 00:06:35,040 --> 00:06:37,840 university course could also be a worthwhile investment, 106 00:06:37,840 --> 00:06:40,720 particularly if you're thinking of a, a job change. 107 00:06:40,720 --> 00:06:44,520 You know, if you actually want to go into AI and say the sort 108 00:06:44,520 --> 00:06:48,520 of two roles, you can either be the sort of CAE domain specific 109 00:06:48,520 --> 00:06:51,760 person, or you might want to be more hardcore into the AI 110 00:06:51,760 --> 00:06:54,400 itself. And probably for that you would 111 00:06:55,120 --> 00:06:56,840 value from some of these courses. 112 00:06:57,040 --> 00:07:00,720 Although many universities now do offer these online courses as 113 00:07:00,720 --> 00:07:02,160 well, like I think Stanford does them. 114 00:07:02,920 --> 00:07:05,120 So you know, there's a lot of options now. 115 00:07:05,120 --> 00:07:07,720 And I don't think there's any excuse not to learn. 116 00:07:08,160 --> 00:07:10,280 You know, it's not like 40 years ago where you have to go to a 117 00:07:10,280 --> 00:07:12,440 library where you're in a physical thing. 118 00:07:12,440 --> 00:07:14,880 The Internet gives you so many opportunities now. 119 00:07:16,160 --> 00:07:19,640 So if you have an open mind, which is 1 and you've, you know, 120 00:07:19,680 --> 00:07:22,800 educated yourself, which is number 2, and of course they're 121 00:07:22,800 --> 00:07:28,000 continual processes, then let's talk about the AI physics. 122 00:07:28,240 --> 00:07:32,840 And this is probably the most common use case or one of the 123 00:07:32,840 --> 00:07:36,040 most common use cases for AI in the context of of CAE and 124 00:07:36,040 --> 00:07:39,560 engineering AI physics, you know, referring to the use of AI 125 00:07:39,560 --> 00:07:44,640 to create typically some sort of surrogate model where you take 126 00:07:44,640 --> 00:07:47,480 data to train a model. And then once the model's 127 00:07:47,480 --> 00:07:50,000 trained that inference, you can give it a new condition, a 128 00:07:50,000 --> 00:07:52,120 geometry, a binary condition, and they'll go and predict it, 129 00:07:52,360 --> 00:07:58,160 typically in close to real time. There are other use cases people 130 00:07:58,160 --> 00:08:01,720 sometimes look at developing better models, you know, 131 00:08:01,720 --> 00:08:05,640 transition model, the turbulence model using AI. 132 00:08:05,640 --> 00:08:08,520 But I'd say probably the predominant one that you may 133 00:08:08,520 --> 00:08:11,640 come across is more of the surrogate modelling. 134 00:08:11,640 --> 00:08:13,600 And yes, some people say, oh, we've been doing that for 135 00:08:13,600 --> 00:08:17,720 decades, we've produced all the models, but typically those did 136 00:08:17,720 --> 00:08:22,760 not have the flexibility like modern day ones in terms of the 137 00:08:23,360 --> 00:08:27,480 non parametric ability as you can just bring any arbitrary 138 00:08:27,480 --> 00:08:30,520 geometry or flow condition in, if you've trained across it, you 139 00:08:30,520 --> 00:08:33,600 can get the results out. So things are different today 140 00:08:34,120 --> 00:08:36,840 than what they used to be. And that common argument of, 141 00:08:36,840 --> 00:08:39,240 well, we've just been doing ML for 30 years and they've just 142 00:08:39,240 --> 00:08:43,799 changed it to something else is probably going back to #1 and #2 143 00:08:43,960 --> 00:08:46,320 and open minded educating yourself to realize, no, no, 144 00:08:46,480 --> 00:08:50,240 things have changed. Although of course they're based 145 00:08:50,240 --> 00:08:52,360 on foundations from from earlier. 146 00:08:52,640 --> 00:08:54,760 So how do you prepare yourself for the air physics? 147 00:08:54,760 --> 00:08:58,840 Before we talk about it in more detail, I said the first one is, 148 00:08:59,280 --> 00:09:04,480 is data Today, the biggest difference between your probably 149 00:09:04,720 --> 00:09:08,920 common use of AI with Google Gemini or ChatGPT or whatever is 150 00:09:08,920 --> 00:09:11,080 they're already trained. You're essentially just doing a 151 00:09:11,080 --> 00:09:13,080 text prompt. You're doing inference on a pre 152 00:09:13,080 --> 00:09:15,480 trained model. The big thing with AI in the 153 00:09:15,480 --> 00:09:19,200 context of CAE and CFD and FEA is you're probably going to have 154 00:09:19,200 --> 00:09:22,880 to train the model yourself. That's most likely at least in 155 00:09:22,880 --> 00:09:27,760 the short term. So you need data and if you have 156 00:09:27,760 --> 00:09:30,720 existing data, it's what formats it in. 157 00:09:30,720 --> 00:09:33,160 Where is it at? Can you get it to the model? 158 00:09:33,160 --> 00:09:35,040 Have you scrutinized it? How? 159 00:09:35,760 --> 00:09:38,400 How do you describe it? The model needs to know what it 160 00:09:38,400 --> 00:09:41,960 was generated with. Not all data is identical, and 161 00:09:41,960 --> 00:09:43,320 that is often the biggest challenge. 162 00:09:43,840 --> 00:09:46,640 Did you run two different geometries, but one, you change 163 00:09:46,640 --> 00:09:49,720 the CFD settings or did you change the material properties 164 00:09:49,720 --> 00:09:53,120 for that crash test? So labeling the data, 165 00:09:53,120 --> 00:09:57,520 classifying the data in a format that a model could read is 166 00:09:57,520 --> 00:10:00,160 important. So for example, you know, coming 167 00:10:00,160 --> 00:10:03,000 up with some sort of schema, some Jason format where you can 168 00:10:03,000 --> 00:10:07,480 say, OK, this data was created on this day using these settings 169 00:10:07,760 --> 00:10:10,720 by this person. You may even add security into 170 00:10:10,720 --> 00:10:13,440 it. I want this to be able to be 171 00:10:13,440 --> 00:10:16,000 trained by this model, not to be trained, you know, if you have 172 00:10:16,000 --> 00:10:19,360 different sorts of data that you want the model to know about, 173 00:10:19,680 --> 00:10:21,920 you should try and describe as well as possible. 174 00:10:22,360 --> 00:10:25,640 The other one, and this is probably more translating now to 175 00:10:25,640 --> 00:10:30,280 new data, is the data that you would traditionally keep or 176 00:10:30,280 --> 00:10:33,440 destroy should be reconsidered in the era of AI. 177 00:10:34,040 --> 00:10:36,280 The example I would give is some CFD. 178 00:10:36,840 --> 00:10:41,240 You may say all I need out of it is the drag and the lift and 179 00:10:41,240 --> 00:10:44,400 some pictures of the flow. And after a certain point, why 180 00:10:44,400 --> 00:10:47,600 do I need the full 3D volume that's 50 gigabytes. 181 00:10:47,600 --> 00:10:51,520 I'm just going to delete that. But by deleting it, you've 182 00:10:51,520 --> 00:10:54,400 probably lost a lot of information that would be needed 183 00:10:54,400 --> 00:10:57,520 by the AI model to go and train the 3D volume solution. 184 00:10:58,240 --> 00:11:01,200 It needs it to be able to learn 3D field. 185 00:11:01,960 --> 00:11:04,840 And if you deleted that, well, you are going to have to 186 00:11:04,840 --> 00:11:08,800 regenerate it. But then how do you regenerate 187 00:11:08,800 --> 00:11:10,920 it with that version of the software that was done and the, 188 00:11:11,160 --> 00:11:13,880 of course, the cost. So traditionally this was a 189 00:11:13,880 --> 00:11:17,400 balance of storage cost versus need. 190 00:11:17,880 --> 00:11:22,880 I would argue now, because data is the key part of AI, it's the 191 00:11:22,880 --> 00:11:26,200 biggest cost of AI when it comes to CAE, you need to keep the 192 00:11:26,200 --> 00:11:27,920 data. So I would argue it's best 193 00:11:27,920 --> 00:11:31,360 investing in paying more for storage to have the data. 194 00:11:31,360 --> 00:11:35,920 Even if you're not today trading AI models, you will do, I'm sure 195 00:11:35,920 --> 00:11:39,960 at some point. So generating data, thinking 196 00:11:39,960 --> 00:11:43,800 about how you keep it and maybe outputting more information than 197 00:11:43,800 --> 00:11:48,600 you think you need because the AI model may need it. 198 00:11:48,600 --> 00:11:51,280 So if normally you would just save the pressure and the 199 00:11:51,280 --> 00:11:54,000 velocity, you might think, well, what about all the other 200 00:11:54,200 --> 00:11:58,920 variables that I might need beyond just what I'm 201 00:11:58,920 --> 00:12:02,120 traditionally getting out? You know, instead of being just 202 00:12:02,120 --> 00:12:04,360 the stream wise velocity, maybe I need all the components of the 203 00:12:04,360 --> 00:12:07,240 velocity, little things like that that you would say, well, I 204 00:12:07,240 --> 00:12:09,280 don't need it because I don't need to explore it. 205 00:12:09,760 --> 00:12:11,920 You might need to train a model later. 206 00:12:12,760 --> 00:12:15,800 Another angle to look at is monetizing this. 207 00:12:16,280 --> 00:12:19,160 So if you're a business or engineering company or even 208 00:12:19,160 --> 00:12:23,680 individual, you may find that your data is very valuable and 209 00:12:23,680 --> 00:12:26,920 you may consider is the way of making a value at that. 210 00:12:26,920 --> 00:12:30,440 Can you change your business around a little bit to offer 211 00:12:30,440 --> 00:12:33,000 this data, whether it's computational or experimental, 212 00:12:33,720 --> 00:12:35,400 to help people train? If you're a company that 213 00:12:35,400 --> 00:12:39,040 operates a wind tunnel, maybe you start to think about using 214 00:12:39,040 --> 00:12:42,880 that to generate data or to monetize your data. 215 00:12:43,880 --> 00:12:46,280 So there's a lot of opportunities in the a, in the 216 00:12:46,400 --> 00:12:50,160 era of AI physics, which is true for AI, for science, that data 217 00:12:50,160 --> 00:12:52,720 is key. And if you have data, it could 218 00:12:52,720 --> 00:12:54,280 be valuable. And you really need to have a 219 00:12:54,280 --> 00:12:56,400 good strategy of how to deal with this. 220 00:12:57,480 --> 00:13:01,560 OK, So you have an open mind, you've educated yourself, you're 221 00:13:01,560 --> 00:13:07,240 preparing for the AI physics. Now what about doing it? 222 00:13:07,400 --> 00:13:10,240 And I think the argument or the thing I would like to discuss 223 00:13:10,240 --> 00:13:13,720 now is really about the builders buy it's a common question I get 224 00:13:13,720 --> 00:13:17,720 when I speak to many people is around, you know, is or an off 225 00:13:17,720 --> 00:13:19,480 the shelf solution. Should I be building this 226 00:13:19,480 --> 00:13:22,960 myself? Well, obviously linked to #2 you 227 00:13:22,960 --> 00:13:27,600 need to educate yourself. And I should say sorry on #3 228 00:13:27,600 --> 00:13:30,440 actually, one thing I forgot to say was staff. 229 00:13:31,560 --> 00:13:34,560 You can prepare yourself from a technology point of view in 230 00:13:34,560 --> 00:13:38,280 terms of getting the data ready, processes ready, but you need to 231 00:13:38,280 --> 00:13:39,760 prepare yourself in terms of hiring. 232 00:13:40,040 --> 00:13:43,040 Do you have somebody in your business who is it an expert? 233 00:13:43,200 --> 00:13:46,840 If you're a manager, if you're an individual, the the preparing 234 00:13:46,840 --> 00:13:49,120 is to prepare yourself, you know, by doing the education. 235 00:13:49,600 --> 00:13:53,120 But if you're a company, you know, don't expect that people 236 00:13:53,120 --> 00:13:55,600 who have been doing CFD for for 20 years are going to be the 237 00:13:55,600 --> 00:13:58,000 best people to do AI and help you develop. 238 00:13:58,000 --> 00:14:01,920 AI is probably going to be, you know, a bunch of graduates who 239 00:14:01,920 --> 00:14:06,040 have AI native and are very familiar with those programming 240 00:14:06,040 --> 00:14:08,800 styles with that, maybe they've studied at university. 241 00:14:10,480 --> 00:14:14,560 There's just a reality that staffing is important. 242 00:14:14,560 --> 00:14:18,000 And you may also need to look beyond your typical recruiting 243 00:14:18,000 --> 00:14:22,040 grounds to get those staff. Those people may not come out of 244 00:14:22,040 --> 00:14:25,120 the universities that you would normally go to, to require to 245 00:14:25,120 --> 00:14:27,400 hire engineers. And I think this is really 246 00:14:27,400 --> 00:14:30,760 interesting blend. And it's a career opportunity 247 00:14:30,760 --> 00:14:34,240 that AI is needed for engineering, but also engineers 248 00:14:34,240 --> 00:14:36,800 are needing it are needed at AI companies. 249 00:14:37,920 --> 00:14:41,680 So OK, let's go to #4 So you've don't open mind, educate 250 00:14:41,680 --> 00:14:43,880 yourself and you're sort of starting to prepare yourself 251 00:14:43,880 --> 00:14:47,160 from a data staffing. What about the actual doing it 252 00:14:47,160 --> 00:14:49,560 now? Well, this is the bill versus 253 00:14:49,560 --> 00:14:51,680 buy. And I would put the analogy to 254 00:14:52,840 --> 00:14:55,360 something I'd probably know best, which is CFD. 255 00:14:56,200 --> 00:14:59,840 And if you looked in the 70s and the 80s, there weren't really 256 00:14:59,840 --> 00:15:04,080 commercial solutions on the market that were the de facto 257 00:15:04,080 --> 00:15:07,880 solutions. In fact, before the 1970s, 258 00:15:08,120 --> 00:15:10,040 people, there was no commercial solutions. 259 00:15:10,280 --> 00:15:12,360 And so you had to develop the code to yourself. 260 00:15:12,520 --> 00:15:15,440 And that was a big thing of NASA, but not just NASA, but 261 00:15:15,440 --> 00:15:19,800 almost all aerospace companies at car companies to some point 262 00:15:20,480 --> 00:15:24,120 manufacturing, they would develop their own codes out of 263 00:15:24,120 --> 00:15:26,400 necessity if they needed to. There wasn't, you know, there 264 00:15:26,400 --> 00:15:27,920 wasn't a commercial solution available. 265 00:15:28,840 --> 00:15:32,880 It was only in the 70s and the 80s and certainly going into the 266 00:15:32,880 --> 00:15:37,040 90s and the 2000s where commercial codes became far more 267 00:15:37,040 --> 00:15:41,600 mature. Interestingly, if you look today 268 00:15:41,600 --> 00:15:46,000 and this is maybe contentious opinion, lots of these companies 269 00:15:46,000 --> 00:15:50,120 that did traditionally only use internal codes are starting to 270 00:15:50,120 --> 00:15:54,000 use commercial codes and that shift is accelerating because 271 00:15:54,000 --> 00:15:56,280 frankly, the commercial codes are getting so good. 272 00:15:56,680 --> 00:15:59,840 They've hired so many people, they've, you know, acquired 273 00:15:59,840 --> 00:16:04,840 start-ups that is your code really as good as theirs. 274 00:16:05,640 --> 00:16:08,640 I think it's a interesting debate. 275 00:16:09,680 --> 00:16:13,600 So debate of do you see it as core to your mission to have 276 00:16:13,600 --> 00:16:16,880 your own code or actually is your main business to go and 277 00:16:17,040 --> 00:16:19,600 build something and the code itself should be done by 278 00:16:19,600 --> 00:16:21,840 somebody else. Most people are starting to 279 00:16:21,840 --> 00:16:25,480 shift towards the latter, that actually it's better just to buy 280 00:16:25,480 --> 00:16:27,960 in the code and have less people developing their own code. 281 00:16:27,960 --> 00:16:31,800 Now again, I'm not saying that is the right approach, I'm just 282 00:16:31,800 --> 00:16:33,640 saying that is what I see happening. 283 00:16:34,320 --> 00:16:39,760 So why am I saying that? If you look at AI physics, it's 284 00:16:39,840 --> 00:16:42,840 probably the fact that we are in the equivalent of the 70s or 285 00:16:42,840 --> 00:16:45,840 80s. We're in this new phase where 286 00:16:46,880 --> 00:16:49,920 there is not the same maturity of commercial solutions. 287 00:16:49,920 --> 00:16:52,200 It's largely start-ups at the moment. 288 00:16:52,800 --> 00:16:57,400 And you could argue that actually, could I build this 289 00:16:57,400 --> 00:16:59,360 myself? Does that give me a strategic 290 00:16:59,360 --> 00:17:01,960 advantage like it was a strategic advantage building 291 00:17:01,960 --> 00:17:06,480 your own CFD code? I would say I can understand the 292 00:17:06,480 --> 00:17:10,560 arguments for both. I can today I can see why maybe 293 00:17:10,839 --> 00:17:13,640 you think if you hire some, you know, smart engineers and build 294 00:17:13,640 --> 00:17:18,040 an open source frameworks that you, you could do that, but you 295 00:17:18,040 --> 00:17:23,760 have to be prepared to keep up. And that's my advice or warning 296 00:17:23,760 --> 00:17:30,120 that even if today there is a split on, oh, it's not so 297 00:17:30,120 --> 00:17:34,320 obvious whether to build or buy, you should be flexible and 298 00:17:34,320 --> 00:17:37,400 prepared that maybe in two years, maybe in five years, 299 00:17:37,680 --> 00:17:41,840 maybe at the extreme 10 years, it's I think it highly like that 300 00:17:42,000 --> 00:17:45,680 the commercial solutions will be as good or better than what you 301 00:17:45,680 --> 00:17:48,440 could do yourself. And so just as people are now 302 00:17:48,640 --> 00:17:51,480 having to debate whether to bring in commercial solutions 303 00:17:51,480 --> 00:17:54,560 from a non AI point of view, I would say you need to be 304 00:17:54,560 --> 00:17:58,200 flexible enough to be able to adopt when there is a good off 305 00:17:58,200 --> 00:18:01,360 the shelf solution that's maybe better than what you could do 306 00:18:01,760 --> 00:18:03,880 internally. And that's part of the sort of 307 00:18:03,880 --> 00:18:08,960 prepare yourself and planning. Luckily today most solutions are 308 00:18:09,240 --> 00:18:13,840 sort of API driven and therefore if done in the right way, you 309 00:18:13,840 --> 00:18:16,800 can sort of integrate things together a bit like you can use 310 00:18:16,800 --> 00:18:20,360 the API of open AI to call that model and call the different 311 00:18:20,360 --> 00:18:23,360 model. I suspect it'll be the case. 312 00:18:23,360 --> 00:18:29,640 So I would 100% advocate people coding and building stuff 313 00:18:29,640 --> 00:18:34,080 themselves, whether they go to production with it or they use a 314 00:18:34,080 --> 00:18:37,160 start up or or buy something. It's very much an individual 315 00:18:37,160 --> 00:18:39,480 choice. But I do have a strong feeling 316 00:18:40,000 --> 00:18:44,880 that given the investment it is likely that they'll be so many 317 00:18:44,880 --> 00:18:47,040 better and good commercial solutions. 318 00:18:47,040 --> 00:18:51,880 Given how big this AL market is for for CAE, that you may need 319 00:18:51,880 --> 00:18:55,560 to be flexible on that choice and reassess it and don't get 320 00:18:55,560 --> 00:18:59,200 locked in to thinking I can't take a commercial because I've, 321 00:18:59,600 --> 00:19:01,120 you know, gone down building myself. 322 00:19:01,120 --> 00:19:02,960 I need to. There's nothing wrong with 323 00:19:03,200 --> 00:19:09,560 having a mixture. And finally, .5, which is 324 00:19:09,800 --> 00:19:13,520 probably the newest, I would argue in the wave of AI for CAE 325 00:19:13,560 --> 00:19:17,160 is agentic AI and a bit like AI physics. 326 00:19:17,160 --> 00:19:19,640 Sometimes people can get people's backs up with the 327 00:19:19,640 --> 00:19:22,600 marketing and the way that it's bullshit this that AI can 328 00:19:22,720 --> 00:19:26,800 automate and do everything. Essentially the agentic AI is to 329 00:19:26,800 --> 00:19:30,360 sort of next wave of where LLMS were. 330 00:19:30,920 --> 00:19:35,120 So in the sense what are the frustrations of using a 331 00:19:35,560 --> 00:19:38,480 traditional LLM is it can't go and do things for you. 332 00:19:38,760 --> 00:19:43,160 It's very much a, you know, look at this document and summarize 333 00:19:43,160 --> 00:19:45,800 it for me. But what you would really like 334 00:19:45,800 --> 00:19:49,080 to do, and to be fair, even now, some of the off the shelf that 335 00:19:49,080 --> 00:19:52,720 can do it, you'd say, I would like you to go and do this for 336 00:19:52,720 --> 00:19:54,840 me. Go and research something, go go 337 00:19:54,840 --> 00:19:57,160 on the Internet, go and search for this, then call this, then 338 00:19:57,160 --> 00:19:59,240 do this. So what does that mean in the 339 00:19:59,240 --> 00:20:05,360 context of, of of CAE? That means at it's very simplest 340 00:20:06,080 --> 00:20:09,960 things like a text prompt to set up simulations and run 341 00:20:09,960 --> 00:20:14,000 simulations. So rather than you clicking 342 00:20:14,000 --> 00:20:16,200 buttons and and going and setting up a simulation and 343 00:20:16,200 --> 00:20:19,080 manually running a bash script. It's through a text prompt. 344 00:20:19,080 --> 00:20:22,080 You should be able to instruct an agent. 345 00:20:22,080 --> 00:20:24,040 And usually there's some sort of master agent that's then 346 00:20:24,040 --> 00:20:26,320 controlling other agents and sending that on. 347 00:20:26,320 --> 00:20:29,000 And I will do a dedicated episode on this because I think 348 00:20:29,000 --> 00:20:34,480 it's such a fascinating topic that essentially allows you then 349 00:20:34,480 --> 00:20:39,800 to quote, UN quote, have a sort of AI engineer where that agent 350 00:20:39,800 --> 00:20:42,480 would be able to go and call other agents. 351 00:20:42,960 --> 00:20:45,160 And again, the reason I'll do a dedicated episode on this is 352 00:20:45,160 --> 00:20:47,320 there's been some good papers out there that I'd like to 353 00:20:47,320 --> 00:20:51,560 discuss and and talk about. But at a very high level, it 354 00:20:51,640 --> 00:20:55,240 essentially would be the ability through a text prompt for one 355 00:20:55,240 --> 00:20:58,200 agent to call another agent, which is perhaps a surrogate 356 00:20:58,200 --> 00:21:03,920 model to go run a simulation. But in importantly, another 357 00:21:03,920 --> 00:21:09,040 agent would then perhaps do an analysis of that, but in a fully 358 00:21:09,040 --> 00:21:11,800 automated way. And then finally, another agent 359 00:21:12,160 --> 00:21:15,680 may decide to go and then do some optimization of the 360 00:21:15,680 --> 00:21:18,840 geometry and pass the information back to your 361 00:21:19,120 --> 00:21:23,880 surrogate model agent. Then do more analysis, let's say 362 00:21:24,200 --> 00:21:27,360 image analysis through an LLM and they may write you APDF. 363 00:21:27,920 --> 00:21:31,280 And the important bit that's all been kicked off by one prompt 364 00:21:31,280 --> 00:21:34,680 from yourself. So it's agents that are 365 00:21:34,680 --> 00:21:37,560 essentially calling each other. And there's a lot of complexity, 366 00:21:37,560 --> 00:21:41,720 of course, with this, but it really gets much more than just 367 00:21:41,720 --> 00:21:44,040 the surrogate. The surrogate still relies you 368 00:21:44,120 --> 00:21:48,400 as a human to run the simulation, a bit like ACFD 369 00:21:48,400 --> 00:21:52,040 simulation or FEA simulation. The agentic side really starts 370 00:21:52,040 --> 00:21:55,800 to get more, I would say, into the the vision of AI, where it's 371 00:21:55,800 --> 00:21:59,160 more fully automated. And if you extract this to its 372 00:21:59,160 --> 00:22:02,800 maximum, you can imagine many Asians acting almost like their 373 00:22:02,800 --> 00:22:07,040 own engineering company. That's obviously quite far into 374 00:22:07,040 --> 00:22:09,760 the future. But I would start preparing for 375 00:22:09,760 --> 00:22:11,080 that. I would start reading up on 376 00:22:11,080 --> 00:22:13,360 that. This is very much where the the 377 00:22:13,520 --> 00:22:17,560 AI researchers at the moment go and read papers, type in into 378 00:22:17,560 --> 00:22:20,880 Google agentic AI, go into YouTube, watch videos, look for 379 00:22:20,880 --> 00:22:24,360 start-ups, speak to them. This is definitely a new wave 380 00:22:24,360 --> 00:22:27,840 that's coming that AI physics is a crucial part of it. 381 00:22:28,640 --> 00:22:32,880 But the agentic AI is arguably more potentially transformative 382 00:22:32,920 --> 00:22:35,640 and more important to to be aware of. 383 00:22:35,640 --> 00:22:38,880 But there's even, I'd say less solutions on the market now, 384 00:22:38,880 --> 00:22:41,760 which is why it's really good to stay ahead of the curve. 385 00:22:42,480 --> 00:22:47,520 So those are just 5 tips being open minded, educate yourself, 386 00:22:47,760 --> 00:22:50,360 prepare for this sort of AI physics revolution. 387 00:22:51,160 --> 00:22:54,520 Get involved in the AI visit tries to foul build by, you 388 00:22:54,520 --> 00:22:57,760 know, kick the tires. And then prepare yourself for 389 00:22:57,760 --> 00:23:02,720 the agentic AI move, which could really blow away and be quite 390 00:23:02,720 --> 00:23:06,920 transformative if things live up to what people hope for. 391 00:23:07,640 --> 00:23:09,280 So I hope this has been interesting. 392 00:23:09,280 --> 00:23:11,280 I tried to keep it a bit short and snappy. 393 00:23:11,640 --> 00:23:15,400 I hope you've learned something. Agree with at least some of the 394 00:23:15,400 --> 00:23:18,240 stuff I said. And what I'm going to do is over 395 00:23:18,240 --> 00:23:21,520 the course of the the next season, we'll, we'll dive into a 396 00:23:21,520 --> 00:23:27,320 couple of these topics a little bit more and hopefully do #2 397 00:23:27,440 --> 00:23:30,600 educate a little bit. So with that, thanks very much 398 00:23:30,600 --> 00:23:33,960 for listening and hope to see you in the next episode.