1 00:00:00,000 --> 00:00:06,800 Today we are talking about generative, AI tools, such as 2 00:00:06,800 --> 00:00:12,000 chat GPT, what's the impact on the Enterprise? 3 00:00:12,200 --> 00:00:17,700 Our Guest is Nicola Marini bnz, know your CTO of ey. 4 00:00:17,700 --> 00:00:21,300 Tell us about your work. I'm responsible for all the 5 00:00:21,300 --> 00:00:23,300 technology that generates revenue for the firm. 6 00:00:23,400 --> 00:00:26,900 So I spent the last two years developing what we call the ey 7 00:00:26,900 --> 00:00:30,900 fabric, which is effectively platform data and and and with 8 00:00:30,900 --> 00:00:33,900 seven components actually station Intelligence on top of 9 00:00:33,900 --> 00:00:36,300 it, the powers our service offering. 10 00:00:36,300 --> 00:00:39,500 So we have about one and a half million clients that are using 11 00:00:39,500 --> 00:00:41,700 it every day. It's interesting. 12 00:00:41,900 --> 00:00:47,200 We don't think of a technology of a consulting firm as being a 13 00:00:47,208 --> 00:00:50,700 technology organization, but obviously you are, I don't think 14 00:00:50,700 --> 00:00:53,400 that is an organization that can be not the technology 15 00:00:53,400 --> 00:00:57,300 organization of the same time. I think technology is, and look 16 00:00:57,300 --> 00:00:59,200 at what we're going to be talking about today, right? 17 00:00:59,200 --> 00:01:02,600 The college is Raising in our lives that there is no business 18 00:01:02,600 --> 00:01:05,800 that doesn't as in users. And then, you know, if you start 19 00:01:05,800 --> 00:01:08,400 thinking about the Consulting organization, when we provide 20 00:01:08,400 --> 00:01:13,000 services to our clients that are very data-intensive technology 21 00:01:13,000 --> 00:01:16,400 becomes so Central, so fundamental to the quality of 22 00:01:16,400 --> 00:01:19,400 the service that we offer the ability of doing and many and 23 00:01:19,400 --> 00:01:25,700 across many countries, etc, etc, Nicola products like chat GPT 24 00:01:25,700 --> 00:01:29,800 Google is coming out with their own version, Microsoft. 25 00:01:29,900 --> 00:01:35,200 Has adopted this now in bing. It's so popular. 26 00:01:35,300 --> 00:01:40,300 But why as an Enterprise CTO do you care about this? 27 00:01:40,700 --> 00:01:43,200 I think what we're going to see. First of all it's an arms race 28 00:01:43,200 --> 00:01:46,400 which is great right? Because usually when the players 29 00:01:46,400 --> 00:01:49,700 are competing with each other you know we tend to benefit from 30 00:01:49,700 --> 00:01:52,600 it right as consumers in this case, or Enterprises. 31 00:01:53,000 --> 00:01:55,600 So it's very exciting over. We're going to be going next in 32 00:01:55,600 --> 00:01:59,200 terms of the capabilities of the technology, but be more specific 33 00:01:59,200 --> 00:02:01,800 about Jenna. May I and you know I'm looking 34 00:02:01,800 --> 00:02:04,000 at my own Organization, for example right? 35 00:02:04,000 --> 00:02:07,500 Which is a knowledge-intensive organization, the ability to 36 00:02:07,500 --> 00:02:12,500 summarize knowledge in the way. You know some of the language 37 00:02:12,500 --> 00:02:17,800 models are capable of doing is absolutely off or opens up an 38 00:02:17,800 --> 00:02:20,000 endless number of opportunities for organization. 39 00:02:20,000 --> 00:02:22,300 So, for example, you know, how do you systematize the knowledge 40 00:02:22,300 --> 00:02:24,800 within the organization, how you being oncologist in 41 00:02:24,800 --> 00:02:28,000 organizations? How do you give advice right to 42 00:02:28,000 --> 00:02:31,700 clients or how you have? For example, scenarios where you 43 00:02:31,700 --> 00:02:33,500 have a sort of a co-pilot, right? 44 00:02:33,500 --> 00:02:36,500 So in pretty much any, any one of the roles and the jobs that 45 00:02:36,500 --> 00:02:38,300 we have in our organization, right? 46 00:02:38,300 --> 00:02:42,200 We would definitely benefit of having an intelligence agent 47 00:02:42,200 --> 00:02:45,800 intelligent agent, sitting beside us, you know everyday not 48 00:02:45,800 --> 00:02:49,700 replacing us, but supporting our knowledge and our capabilities 49 00:02:49,700 --> 00:02:54,000 trying in everything that we do. So endless possibilities. 50 00:02:54,000 --> 00:02:57,600 And so most and also not only in the Enterprise values in our 51 00:02:57,600 --> 00:03:00,500 lives. I think that would be a In fact 52 00:03:00,800 --> 00:03:06,800 as a practical matter as you think about these generative AI 53 00:03:07,100 --> 00:03:11,000 capabilities, how do you, how do you think about a do break it 54 00:03:11,000 --> 00:03:13,600 down into areas that you think might be useful? 55 00:03:13,600 --> 00:03:19,000 Or is it still just too early of an ex exploration phase or where 56 00:03:19,000 --> 00:03:21,600 are you in your thinking? I think he's going to go very 57 00:03:21,600 --> 00:03:24,300 very fast. What already exploring? 58 00:03:24,300 --> 00:03:26,600 I mean in our specific business, right? 59 00:03:26,600 --> 00:03:29,700 Which is to give advice to clients based. 60 00:03:29,900 --> 00:03:33,300 On our knowledge or regulations that our knowledge of other 61 00:03:33,300 --> 00:03:35,300 interaction with other clients. Etc. 62 00:03:35,600 --> 00:03:39,100 You know the ability of you know tapping into the knowledge in a 63 00:03:39,100 --> 00:03:41,900 way that is multi-dimensional like you know some of these 64 00:03:41,900 --> 00:03:46,200 tools can do is absolutely incredible and and fundamental. 65 00:03:46,200 --> 00:03:50,300 So we have several projects already Eddie why that are 66 00:03:50,300 --> 00:03:52,900 tackling that. But I think, what is most 67 00:03:52,900 --> 00:03:56,700 fascinating about this? Latest development is so we had 68 00:03:56,700 --> 00:03:59,800 chatbots for a long time, really and not know exactly when 69 00:04:00,000 --> 00:04:01,400 Laughter. But you know, obviously some 70 00:04:01,400 --> 00:04:04,300 companies had their child boss for at least you know, eight to 71 00:04:04,300 --> 00:04:07,200 ten years, right? But what is different about? 72 00:04:07,200 --> 00:04:10,100 This one is the fact that you actually can have a conversation 73 00:04:10,100 --> 00:04:12,100 with it. Sometimes this job boards you 74 00:04:12,108 --> 00:04:16,000 know you are doing Custom Image as a cook as a as a consumer. 75 00:04:16,000 --> 00:04:19,399 I you go onto a website because you have one I have some more 76 00:04:19,399 --> 00:04:22,000 information about the product so you are complaining about the 77 00:04:22,000 --> 00:04:24,700 protocol something right? You get into these very 78 00:04:24,700 --> 00:04:28,300 frustrating, you know, interaction with the chat bot 79 00:04:28,300 --> 00:04:30,900 that if you don't. Exactly Hit the question, the 80 00:04:30,900 --> 00:04:35,200 way, the child bought the specs, it then you go in nowhere and 81 00:04:35,200 --> 00:04:38,500 you list is what I do. You get to a point where you 82 00:04:38,508 --> 00:04:41,900 say, please I want to speak to an agent, a real human, right? 83 00:04:42,000 --> 00:04:45,600 And this will these tools now that are coming to. 84 00:04:45,600 --> 00:04:48,800 The market are actually very different because they allow you 85 00:04:48,800 --> 00:04:51,200 to actually have a dialogue because they can maintain the 86 00:04:51,200 --> 00:04:52,700 state of that conversation, right? 87 00:04:52,700 --> 00:04:55,700 So the previous job boards, you ask a question and get an answer 88 00:04:55,700 --> 00:04:58,500 and that's it. There is no memory of that 89 00:04:58,500 --> 00:05:01,400 interaction. That the Book everything but now 90 00:05:01,400 --> 00:05:04,700 you can so you can go. It's almost like I call it but 91 00:05:04,700 --> 00:05:09,000 maybe a little bit abusing the term by sort of a dialectic of 92 00:05:09,000 --> 00:05:09,700 AI. Right? 93 00:05:09,700 --> 00:05:12,300 So where you ask a question, you get an answer and you can 94 00:05:12,300 --> 00:05:15,200 another you ask another question and you get closer to what you 95 00:05:15,200 --> 00:05:19,100 want to get out of it as opposed to a sort of an expert system 96 00:05:19,100 --> 00:05:22,600 where you know, it's question and answer and that's pretty 97 00:05:22,600 --> 00:05:25,700 much it, right? So it's very interesting those I 98 00:05:25,707 --> 00:05:28,600 think that is the feature that to me is most exciting. 99 00:05:29,000 --> 00:05:32,600 Would it be Correct to say that you feel the defining 100 00:05:32,600 --> 00:05:39,200 characteristic is this ability to remain to to retain the state 101 00:05:39,600 --> 00:05:43,100 of the conversation? There are so many other things 102 00:05:43,100 --> 00:05:46,200 that are great, right? The ability of writing in in 103 00:05:47,900 --> 00:05:50,400 multiple languages, right? So, I'm Italian original, right? 104 00:05:50,400 --> 00:05:53,900 So, that's my first language. He writes as good in Italian as 105 00:05:53,900 --> 00:05:56,600 it is in English, right? And you can ask the questions, 106 00:05:56,600 --> 00:06:02,400 you can you can Envision like you know, Meetings where, you 107 00:06:02,400 --> 00:06:05,200 know, you have everyone speaks his own language, really think 108 00:06:05,200 --> 00:06:08,300 it would be an incredible for global companies will be an 109 00:06:08,300 --> 00:06:10,800 incredible achievement. Right today, we still have to 110 00:06:11,300 --> 00:06:14,300 there is still a lot of Lost in Translation that these tools are 111 00:06:14,300 --> 00:06:18,800 so refined and sophisticated now that even the nuances of a 112 00:06:18,800 --> 00:06:23,100 translation can be really, you know, highlighted in the right 113 00:06:23,100 --> 00:06:28,200 way, which is incredible. We have a question that came in 114 00:06:28,200 --> 00:06:32,300 from Twitter. Already, which is an important 115 00:06:32,300 --> 00:06:36,000 question from our Salon con are salons, a regular listener. 116 00:06:36,000 --> 00:06:38,800 And I've always thank Our Salon for his wonderful questions. 117 00:06:39,300 --> 00:06:45,500 And he says this, he says, when we use AI as a co-pilot, we have 118 00:06:45,500 --> 00:06:50,000 to be careful about the data. It is pulling in order to give 119 00:06:50,000 --> 00:06:56,300 us suggestions if we aren't aware of biases or stereotypes 120 00:06:56,400 --> 00:07:01,800 in the data than how what should we Do it's one of the 121 00:07:01,800 --> 00:07:04,300 fundamental questions of dealing with these types of 122 00:07:04,300 --> 00:07:06,300 Technologies. Absolutely. 123 00:07:06,300 --> 00:07:09,200 I had lots of people that inside my own company, right? 124 00:07:09,200 --> 00:07:11,500 The rich are say, I want to try GDP, I want to. 125 00:07:11,500 --> 00:07:14,000 How can we use the with clients? I said in a so it will do 126 00:07:14,000 --> 00:07:17,900 caution in. I think it's where warranted at 127 00:07:17,900 --> 00:07:21,500 this point, right? Meaning that you are the human 128 00:07:21,500 --> 00:07:26,000 in charge, you know what I mean? So it cannot delegate completely 129 00:07:26,400 --> 00:07:30,200 the work to these agents because they're not yet at that, Level 130 00:07:30,200 --> 00:07:34,600 of being able to understand, you know, how ethical code, right 131 00:07:34,600 --> 00:07:37,200 the morality. Then you know what is acceptable 132 00:07:37,200 --> 00:07:40,200 or not acceptable? Unless it has been labeled by 133 00:07:40,200 --> 00:07:42,600 another human by. So the human in the loop 134 00:07:43,100 --> 00:07:45,700 approach is absolutely essential here, right? 135 00:07:45,700 --> 00:07:50,000 So you should so I use it as a sort of a, an aid as opposed to 136 00:07:50,008 --> 00:07:52,100 a delegation of responsibility, right? 137 00:07:52,100 --> 00:07:55,300 That cannot happen. So the human has to be in charge 138 00:07:55,300 --> 00:07:58,400 and hopefully the human. I mean you can argue that even 139 00:07:58,400 --> 00:08:01,800 the human-sized a lot of Show flaws, right intensive, what 140 00:08:01,800 --> 00:08:04,600 kind of answers? You know is capable of giving 141 00:08:04,600 --> 00:08:08,400 but you know what not a pusher at the point yet where we can 142 00:08:08,400 --> 00:08:11,600 delegate a function to you know, a tool like this. 143 00:08:12,200 --> 00:08:17,300 I heard a very interesting comment yesterday on the news 144 00:08:17,300 --> 00:08:24,100 from a journalist who said, Who as a reminder that tools like 145 00:08:24,100 --> 00:08:27,400 Chad GPT or what Google is going to come out with all of these 146 00:08:27,400 --> 00:08:32,400 tools, they're really like advanced auto correct on your 147 00:08:32,400 --> 00:08:34,799 phone. They have they give the 148 00:08:34,900 --> 00:08:39,799 appearance of sentience but they're just it's just like 149 00:08:39,799 --> 00:08:42,299 autocomplete. We could go into it. 150 00:08:42,299 --> 00:08:46,200 Like a very long philosophical. Conversation was you know, what 151 00:08:46,200 --> 00:08:49,500 is sentence was sent into really means but I agree, I think I 152 00:08:49,500 --> 00:08:54,300 mean I would say the outer item is a little bit to trivialised 153 00:08:54,300 --> 00:08:57,400 right as a definition. Definitely is now. 154 00:08:57,400 --> 00:09:00,000 So if you autocorrect is on one end of the spectrum, It's 155 00:09:00,000 --> 00:09:02,000 awesome. And on the other end is 156 00:09:02,000 --> 00:09:04,700 sentences. Like it's it's in the middle, 157 00:09:04,800 --> 00:09:06,800 right? So I think we need to 158 00:09:06,808 --> 00:09:09,200 understand. So sometimes, you know, we tend 159 00:09:09,200 --> 00:09:12,100 to whenever there is a new technology especially this kind 160 00:09:12,100 --> 00:09:15,000 of technology that a little bit more difficult to understand 161 00:09:15,000 --> 00:09:18,900 because they're actually quite complex, we tend to give them 162 00:09:18,900 --> 00:09:22,400 either, you know, kind of transfer some sort of 163 00:09:22,400 --> 00:09:24,700 alienation, right? We transferred to them the 164 00:09:24,700 --> 00:09:27,400 expectations that we have from another human being, right? 165 00:09:27,400 --> 00:09:30,800 So if we do that, that is the wrong approach, right? 166 00:09:30,800 --> 00:09:33,600 This is not a human being. This is the bunch of silicon, 167 00:09:33,600 --> 00:09:35,900 you know? And a lot of data that gives you 168 00:09:35,900 --> 00:09:38,700 some kind of an answer. But if you actually think about 169 00:09:38,700 --> 00:09:43,300 it as a sort of a in a dry in taking, an aeroplane to go from 170 00:09:43,300 --> 00:09:46,900 San Francisco to New York, that is what these tools are. 171 00:09:46,900 --> 00:09:50,400 So it's a machine that helps you do things faster, right? 172 00:09:50,400 --> 00:09:54,900 But if you think about it it's not that different than at the 173 00:09:54,900 --> 00:09:58,900 beginning, truly of the mass adoption of the internet, right? 174 00:09:58,900 --> 00:10:02,500 So in the eye, I cannot do anything in the 90s from 175 00:10:02,500 --> 00:10:03,400 University. Right. 176 00:10:03,400 --> 00:10:07,000 So whenever I had to do something research paper, I had 177 00:10:07,000 --> 00:10:11,700 to go to my library in my city to find the physical book and, 178 00:10:11,700 --> 00:10:14,500 you know, taking forever to get the information out excited. 179 00:10:14,500 --> 00:10:18,300 And then the search engines came up and that was like an 180 00:10:18,500 --> 00:10:21,100 incredible because you can access the whole knowledge of 181 00:10:21,100 --> 00:10:24,900 the human race, you know, either the keystroke. 182 00:10:25,500 --> 00:10:29,600 And so it doesn't mean that we assign to that technology like 183 00:10:29,800 --> 00:10:32,400 Superhuman ability, right? It's something that helps you 184 00:10:32,700 --> 00:10:35,200 search things better and the saying I think we need to look 185 00:10:35,200 --> 00:10:38,900 at these new Tools in that same way show me what's possible. 186 00:10:38,900 --> 00:10:42,800 So show me the search criteria. Show me what the search results 187 00:10:42,800 --> 00:10:46,000 are and then I need to be what the one that makes the decision 188 00:10:46,000 --> 00:10:47,700 on what is relevant and what is not. 189 00:10:48,000 --> 00:10:52,100 I think one of the challenges for folks in the Enterprise or, 190 00:10:52,100 --> 00:10:57,200 or just our personal lives, is that the answers these tools 191 00:10:57,200 --> 00:11:01,100 give comes across as being Authoritative. 192 00:11:01,400 --> 00:11:02,600 Yes. Yes. 193 00:11:02,800 --> 00:11:06,200 But it may be totally incorrect. Absolutely. 194 00:11:07,600 --> 00:11:10,400 And that is what we shouldn't. This is the thing, right? 195 00:11:10,700 --> 00:11:14,200 If we think about it, we can start talking about human 196 00:11:14,200 --> 00:11:18,900 characteristics and we transferred them to describe the 197 00:11:18,900 --> 00:11:24,000 tool we make a mistake, right? We will never call, you know, a 198 00:11:24,000 --> 00:11:27,600 car like, you know, fast legs, you know, maybe that's the wrong 199 00:11:27,600 --> 00:11:29,300 example. But you said, I mean somebody if 200 00:11:29,300 --> 00:11:32,700 we Use them and say, okay, so we need to put them in sort of the 201 00:11:32,700 --> 00:11:36,800 in the box, right? And say like you know, today you 202 00:11:36,800 --> 00:11:40,200 do some of the search or Google do a Google Search, right? 203 00:11:40,200 --> 00:11:43,400 And it comes back like with millions of pages, sometimes, 204 00:11:43,400 --> 00:11:45,900 right? So it's up to us to select. 205 00:11:45,900 --> 00:11:49,500 What is the page that has the highest level of relevance of 206 00:11:49,500 --> 00:11:52,500 what we're looking for, right? We need to look at it the same 207 00:11:52,500 --> 00:11:55,200 way as opposed to. Okay. 208 00:11:55,200 --> 00:11:58,500 That is a you know, it in a super lap. 209 00:11:58,500 --> 00:12:01,300 You superhuman intelligence Since the, you know, knows 210 00:12:01,300 --> 00:12:03,200 everything, right? So that's why I think, you know, 211 00:12:03,200 --> 00:12:05,100 when you think about that Katrina as well. 212 00:12:05,400 --> 00:12:09,800 I think if we start relying on these things too much, while we 213 00:12:09,800 --> 00:12:13,700 lose the fundamental understanding of a domain or the 214 00:12:13,700 --> 00:12:15,500 subject, right? It can become a little bit 215 00:12:15,500 --> 00:12:19,400 dangerous from that perspective. Please subscribe to our YouTube 216 00:12:19,400 --> 00:12:22,000 channel. Hit the Subscribe button at the 217 00:12:22,000 --> 00:12:28,000 top of 60 talk.com, our website and subscribe to our newsletter. 218 00:12:28,300 --> 00:12:33,000 Well, we have another Question from Twitter that's related to 219 00:12:33,000 --> 00:12:35,400 this. And this is from Chris Peterson 220 00:12:35,500 --> 00:12:39,300 who says, in business Communications, whether it's B2B 221 00:12:39,300 --> 00:12:43,200 or b2c. Do we have an ethical duty to 222 00:12:43,200 --> 00:12:47,500 call out what's written or edited by an AI assistant? 223 00:12:47,900 --> 00:12:53,800 He says, if a human reviews, it modifies it and hit, send, does 224 00:12:53,800 --> 00:12:59,200 that change this ethical duty today, when you publish a paper 225 00:12:59,800 --> 00:13:02,400 There is a process to publish a paper, right? 226 00:13:02,400 --> 00:13:05,800 So you need peer reviews, you need to, you know, cite your 227 00:13:05,800 --> 00:13:08,600 sources, you need to do all these type of things, okay? 228 00:13:08,600 --> 00:13:12,200 You cannot just say, I came up with this right without, you 229 00:13:12,200 --> 00:13:14,500 know, that that would be pleasure is MM, right? 230 00:13:14,800 --> 00:13:20,100 So I think there has to be a framework similar to that one in 231 00:13:20,100 --> 00:13:22,700 this space, right? So if I'm like doing, for 232 00:13:22,700 --> 00:13:25,800 example, like, you know, a client, for example, asks me 233 00:13:25,800 --> 00:13:28,900 about whatever some regulation somewhere, right? 234 00:13:28,900 --> 00:13:32,900 I can do some research and I can use charge apt or can use 235 00:13:32,900 --> 00:13:36,500 another tool that I can use, you know, my own traditional, you 236 00:13:36,500 --> 00:13:40,200 Google searches or something, to get the information out my own 237 00:13:40,200 --> 00:13:42,800 internal knowledge and provide that answer, right? 238 00:13:42,800 --> 00:13:45,200 So when I do that in an email, right? 239 00:13:45,200 --> 00:13:48,700 As long as I think, you know, it is fine to do it as a human 240 00:13:48,700 --> 00:13:50,500 being, right? So I think, I don't know why 241 00:13:50,500 --> 00:13:54,200 there shouldn't be any other way, but if I'm producing a 242 00:13:54,208 --> 00:13:57,900 paper just using that right relying on somebody else's work 243 00:13:57,900 --> 00:14:01,700 and I don't, you know, to That and I don't do the citations and 244 00:14:01,700 --> 00:14:04,000 I don't do this. And that then, of course, you 245 00:14:04,000 --> 00:14:05,600 know, I'm getting into plagiarism. 246 00:14:05,800 --> 00:14:08,200 I think it's the same way that a human would be doing in a 247 00:14:08,208 --> 00:14:11,500 research paper like that. So complicated, so I think that 248 00:14:11,500 --> 00:14:16,800 is the need of especially in the IP in the hole, you know, legal 249 00:14:16,800 --> 00:14:20,900 discipline of intellectual property, is probably as well. 250 00:14:21,100 --> 00:14:24,700 That there is a need of some kind of a rethinking about some 251 00:14:24,700 --> 00:14:27,200 of the fundamentals of that because this is different. 252 00:14:27,500 --> 00:14:30,800 Yeah. And clearly There are a whole 253 00:14:30,800 --> 00:14:36,300 set of ethical and legal issues. Absolutely, right. 254 00:14:36,300 --> 00:14:39,400 So and and so that's why I think we need it's funny that 255 00:14:39,400 --> 00:14:41,500 philosophy, right? Probably the day ancient most 256 00:14:41,500 --> 00:14:46,200 ancient academic discipline is still, you know, very much alive 257 00:14:46,200 --> 00:14:48,800 is because we need to ask ourselves so these questions, 258 00:14:48,800 --> 00:14:50,400 right? And we need to have Frameworks 259 00:14:50,400 --> 00:14:53,600 to to address them. Like, you know, it wasn't 260 00:14:53,600 --> 00:14:56,600 considered when the the beginning of intellectual 261 00:14:56,600 --> 00:15:02,600 property came up as a protected. You know, right that the in 262 00:15:02,600 --> 00:15:06,100 those days like when the legislation started to get 263 00:15:06,100 --> 00:15:09,200 introduced to wear similar conversations, I'm expecting, 264 00:15:09,200 --> 00:15:11,000 right? So you're not supposed to take 265 00:15:11,000 --> 00:15:14,400 somebody else's intellectual work and make it yours, right? 266 00:15:14,600 --> 00:15:17,000 And so I think we need to have that dialogue today. 267 00:15:17,300 --> 00:15:19,300 What is going to be much more difficult today? 268 00:15:19,300 --> 00:15:25,100 Is that these tools that the marginal cost of producing or 269 00:15:25,100 --> 00:15:28,100 generating, you know, this type of documentation. 270 00:15:28,100 --> 00:15:31,700 So, whatever we generate is is almost 0. 271 00:15:31,700 --> 00:15:38,700 So we will be inundated by this volume of documents papers, 272 00:15:38,800 --> 00:15:42,800 books, images and stuff and will be really hard to understand 273 00:15:43,100 --> 00:15:45,000 what is really generated by human. 274 00:15:45,000 --> 00:15:47,400 What is not excited. So, you know, maybe the 275 00:15:47,400 --> 00:15:51,600 Christmas asking, like, identifying that content be, you 276 00:15:51,600 --> 00:15:54,600 know, a really good step, first step, I guess. 277 00:15:55,100 --> 00:15:58,400 And this is our Salon. Khan comes back again and he 278 00:15:58,400 --> 00:16:02,500 says when everyone Is quote doing AI? 279 00:16:02,500 --> 00:16:06,100 What sort of governance structures should be placed? 280 00:16:06,500 --> 00:16:09,800 And I love this. Lets you know, let's let's 281 00:16:09,800 --> 00:16:11,700 right. Let's write a thesis on this. 282 00:16:11,700 --> 00:16:14,900 What sort of governance structures should be in place at 283 00:16:14,900 --> 00:16:20,300 an organizational level as a governmental and as a global 284 00:16:20,300 --> 00:16:24,200 level, what are we just simply start with organizational 285 00:16:24,200 --> 00:16:27,500 governance of a iine that's going to become important. 286 00:16:27,500 --> 00:16:29,000 Very important. Any thoughts on that? 287 00:16:29,300 --> 00:16:32,100 So, for example, I was sunk Mentioning the example of a 288 00:16:32,100 --> 00:16:34,000 knowledge company like ours, right? 289 00:16:34,000 --> 00:16:39,800 So first of all, you want to be able to store then knowledge in 290 00:16:39,800 --> 00:16:43,500 a way that is relatively secure in the sense that no secure, I 291 00:16:43,508 --> 00:16:46,600 mean, of course, from attacks from the outside, but that's one 292 00:16:46,600 --> 00:16:48,100 piece. But also secured in the sense 293 00:16:48,100 --> 00:16:52,800 that is the knowledge that the company wants to protect, right? 294 00:16:52,800 --> 00:16:54,300 So the Nola. So if you think about an 295 00:16:54,300 --> 00:16:56,300 ontology within an Enterprise, right? 296 00:16:56,300 --> 00:16:59,900 And you have the definition of some entities or concepts, you 297 00:16:59,900 --> 00:17:04,099 want to make sure That definition is shared by 298 00:17:04,099 --> 00:17:06,700 everybody. It is approved, Etc. 299 00:17:06,800 --> 00:17:08,599 Right? So I think that is the kind of a 300 00:17:08,800 --> 00:17:11,700 need of establishes. A sort of in a, I'm gonna say 301 00:17:11,700 --> 00:17:13,700 this but probably the wrong word, right? 302 00:17:13,700 --> 00:17:17,099 It's all sort of in. It is toriel board, right? 303 00:17:17,099 --> 00:17:21,599 For these type of knowledge, the guess systematized within the 304 00:17:21,599 --> 00:17:23,500 knowledge repository of the Enterprise. 305 00:17:23,500 --> 00:17:27,700 So the concept of you know, what is revenue, for example, for a 306 00:17:28,000 --> 00:17:30,300 in all the company, right? It can be different. 307 00:17:30,400 --> 00:17:32,900 Then what is used in on the street. 308 00:17:33,200 --> 00:17:37,300 And so when people search for the word Revenue, there has to 309 00:17:37,300 --> 00:17:40,300 be an approved definition of it behind the scenes. 310 00:17:40,300 --> 00:17:44,400 So to me that is highly curated content so you have to have an 311 00:17:44,400 --> 00:17:48,300 organization I believe that looks after that make sure that 312 00:17:48,300 --> 00:17:51,000 nothing you know we've asked Revenue only gives you another 313 00:17:51,000 --> 00:17:52,300 answer, that's not good. Okay. 314 00:17:52,300 --> 00:17:57,000 So that type of, you know, potential Divergence has to be 315 00:17:57,700 --> 00:18:00,800 understood as to be prevented and has to be Much. 316 00:18:01,000 --> 00:18:05,100 So that to me is the key with other areas where they kind of 317 00:18:05,100 --> 00:18:11,400 it, you know, being true to the book definition of things is not 318 00:18:11,400 --> 00:18:13,300 that critical. I think, you know, a little bit 319 00:18:13,300 --> 00:18:16,200 more dialogue could be, we welcome, right? 320 00:18:16,200 --> 00:18:20,100 And I think we need to lacks the tool itself, learn and evolve 321 00:18:20,100 --> 00:18:22,000 Etc. So too much controls and too 322 00:18:22,000 --> 00:18:24,100 much governance, and I think is going to be a good thing. 323 00:18:24,700 --> 00:18:27,700 So I guess the answer is it depends right? 324 00:18:27,700 --> 00:18:30,800 Technical Consulting answer, but it depends in the Is that, you 325 00:18:30,800 --> 00:18:32,900 know, what is the type of knowledge that you want to 326 00:18:33,200 --> 00:18:36,500 maintain govern and protect as opposed to the other ones that 327 00:18:36,500 --> 00:18:39,700 you want a little bit more free flowing and and more 328 00:18:39,700 --> 00:18:43,700 interactive. That makes a lot of sense and 329 00:18:44,500 --> 00:18:48,500 looking at other important domains inside an organization 330 00:18:48,500 --> 00:18:53,200 that require governance. Just, this is a natural kind of 331 00:18:53,200 --> 00:18:57,000 a pro to use that as a model. But then be open to the fact 332 00:18:57,000 --> 00:19:01,600 that, as you said generative a I will Change and evolve over 333 00:19:01,600 --> 00:19:04,100 time. And so we had we can't be to 334 00:19:04,100 --> 00:19:05,800 lock down about it. You cannot? 335 00:19:05,800 --> 00:19:07,200 You cannot constrain it too much. 336 00:19:07,200 --> 00:19:09,900 I think because otherwise you lose the value of it, right? 337 00:19:09,900 --> 00:19:11,900 It's just curate and knowledge. And you know, if there's a 338 00:19:11,908 --> 00:19:15,700 knowledge it's it's it's real view, more knowledge, right? 339 00:19:15,700 --> 00:19:18,300 So you don't, you don't, you're not able to anticipate what's 340 00:19:18,300 --> 00:19:20,000 coming. We have another really 341 00:19:20,000 --> 00:19:25,900 interesting question, short question, hard question from 342 00:19:25,900 --> 00:19:30,200 Anil Vos and he says, what kind of jobs? 343 00:19:30,300 --> 00:19:35,000 May become redundant by generative, AI tools. 344 00:19:35,400 --> 00:19:38,100 I don't think this technology yet, okay. 345 00:19:38,100 --> 00:19:40,300 I don't know about a next five to ten years, okay. 346 00:19:40,300 --> 00:19:44,500 That's different. But right now, I don't see the 347 00:19:44,600 --> 00:19:49,400 again, the delegation of the function completely towards 348 00:19:49,400 --> 00:19:52,600 another, to even if in the Creator's economy, right? 349 00:19:52,600 --> 00:19:56,400 Even if you create art like digital art in your life, that's 350 00:19:56,400 --> 00:19:59,100 your job. I don't think he can be 351 00:19:59,100 --> 00:20:01,900 completely created by the machine, right? 352 00:20:01,900 --> 00:20:04,900 We're seeing, you know, a I asked Etc but at the same time I 353 00:20:04,908 --> 00:20:08,000 think if you are an artist and if you are the Creator, that is 354 00:20:08,000 --> 00:20:12,200 a lot of additional value that you can add on top of what the 355 00:20:12,200 --> 00:20:14,000 machine creates by itself, right. 356 00:20:14,000 --> 00:20:19,300 And also, in the Enterprise, I see more jobs created than less 357 00:20:20,100 --> 00:20:23,100 the jobs disappearing, right? Because I see a lot of 358 00:20:23,100 --> 00:20:26,200 opportunity for example, for you know, people that can 359 00:20:26,200 --> 00:20:29,400 systematized. In a different way, right? 360 00:20:29,400 --> 00:20:34,300 Or can, you know again? Like I am very big on this 361 00:20:34,500 --> 00:20:38,200 concept of ontology is because I think if you can store the 362 00:20:38,200 --> 00:20:41,600 Enterprise knowledge, you know, they're its value that 363 00:20:41,600 --> 00:20:43,700 translates into shareholder value, right? 364 00:20:43,700 --> 00:20:48,300 And it's also like, in allows you to protect the IP and the 365 00:20:48,300 --> 00:20:49,900 capabilities of an organization, right? 366 00:20:49,900 --> 00:20:54,100 So think about our organization 360,000 people, eight hours a 367 00:20:54,100 --> 00:20:57,500 day, is 2.4 million hours of work every day issue. 368 00:20:57,600 --> 00:21:01,500 Can capture the knowledge structure it in a machine, and 369 00:21:01,500 --> 00:21:05,700 then having give access to it to our clients, at our own people, 370 00:21:06,100 --> 00:21:09,500 that makes our the, our company even more valuable as what it 371 00:21:09,500 --> 00:21:11,800 is. So I see more opportunities than 372 00:21:11,800 --> 00:21:14,200 less. If that makes sense, you have 373 00:21:14,200 --> 00:21:19,000 been alluding to and talking about this Knowledge Management 374 00:21:19,000 --> 00:21:24,500 function. How does generative a I change 375 00:21:24,500 --> 00:21:27,500 Knowledge Management relative to this? 376 00:21:27,700 --> 00:21:30,900 Torkoal tools. If I look at my company, right? 377 00:21:30,900 --> 00:21:35,700 And not going outside of our business that is to me is the 378 00:21:35,800 --> 00:21:37,700 the low-hanging fruit. Okay? 379 00:21:37,700 --> 00:21:43,000 That is so the fun day we had able to access all the knowledge 380 00:21:43,500 --> 00:21:46,900 and the fingertip in a way that is human friendly. 381 00:21:47,100 --> 00:21:49,300 So today, right? If I have to say for example in 382 00:21:49,300 --> 00:21:53,200 the in our tax business, if I had to help a client you know 383 00:21:53,200 --> 00:21:57,800 understand the taxation regulation around 10 count It's 384 00:21:57,800 --> 00:22:00,900 usually I had to read like, you know, millions of pages of stuff 385 00:22:00,900 --> 00:22:03,200 Ryan and has to be an army of people doing it. 386 00:22:03,500 --> 00:22:06,500 If I have a tool that's not that, you know, with level of 387 00:22:06,500 --> 00:22:09,600 confidence summarizes. So that information I can spend 388 00:22:09,600 --> 00:22:13,300 more time talking about client with clients about their options 389 00:22:13,300 --> 00:22:17,200 than just, you know, reciting what the regulations of the 390 00:22:17,200 --> 00:22:19,300 right. So people will be happier will 391 00:22:19,300 --> 00:22:21,300 have access. We have course, I've everything 392 00:22:21,300 --> 00:22:22,800 has to be cured as I said, right? 393 00:22:22,800 --> 00:22:25,700 Especially in this type of businesses, as to be curated and 394 00:22:25,700 --> 00:22:28,800 management government class. I can The wild west, right? 395 00:22:28,800 --> 00:22:32,100 Because you don't wanna, you know, then the machine thinks 396 00:22:32,100 --> 00:22:34,800 that, you know, we're looking at the legislation of Switzerland. 397 00:22:34,800 --> 00:22:37,400 And in reality, we need to talk about Austria, right? 398 00:22:37,400 --> 00:22:40,700 So, that's, that is a potential issue, right? 399 00:22:41,000 --> 00:22:44,100 But when you have done that, I think, you know, the research 400 00:22:44,400 --> 00:22:48,000 and the work they will be doing is much more, will be more like 401 00:22:48,400 --> 00:22:51,200 the number of hours that we spend doing more valuable work 402 00:22:51,800 --> 00:22:53,600 will be. I think hiring people will be 403 00:22:53,600 --> 00:22:56,600 satisfied and the club, the quality of the work that we'll 404 00:22:56,600 --> 00:23:01,000 be doing. - 1. How will these kinds of tools 405 00:23:01,000 --> 00:23:07,800 change work inside and Enterprise in every business 406 00:23:07,800 --> 00:23:10,700 application, right? The one that we build that we 407 00:23:10,700 --> 00:23:12,400 don't have it yet, but, you know, we're going in that 408 00:23:12,400 --> 00:23:14,300 direction. I still have like I will have my 409 00:23:14,600 --> 00:23:19,300 little, you know, some window or the frame on this side of the 410 00:23:19,300 --> 00:23:22,900 main screen that allows me to ask questions on the fly in real 411 00:23:22,900 --> 00:23:26,600 time about what I'm doing. And so think about like someone 412 00:23:26,600 --> 00:23:30,000 gets out of school. I am 23, 24 years old, out of 413 00:23:30,000 --> 00:23:33,000 University to joined our company, right? 414 00:23:33,000 --> 00:23:35,600 Instead of, you know, said getting them to do like, you 415 00:23:35,608 --> 00:23:39,300 know, weeks and weeks of training and asking making sure 416 00:23:39,300 --> 00:23:42,100 that they get to the right person as a mentor excited that 417 00:23:42,100 --> 00:23:45,100 they can do that, of course because the human relationship 418 00:23:45,100 --> 00:23:48,300 is so important. But at the same time if they can 419 00:23:48,300 --> 00:23:51,200 access their knowledge, institutional knowledge, 420 00:23:51,200 --> 00:23:54,500 codified in the system, I mean there will be there will be much 421 00:23:54,500 --> 00:23:57,000 easier for them to work and to grow, right? 422 00:23:57,000 --> 00:24:00,500 So I think think that to me, that's what the copilot concept, 423 00:24:00,500 --> 00:24:03,800 right? It's not like it's the function 424 00:24:03,800 --> 00:24:08,500 will be automated or replaced but I have this agent that I can 425 00:24:08,500 --> 00:24:10,500 talk to and ask questions, we fantastic. 426 00:24:10,500 --> 00:24:12,500 I wish I had that beginning of my career. 427 00:24:13,100 --> 00:24:19,400 I to a fair amount of writing and I use chat GPT and I just 428 00:24:20,100 --> 00:24:26,700 subscribed or was put on the, the user list for Microsoft's 429 00:24:26,700 --> 00:24:29,400 Bing and I can't wait. For Google to come out with 430 00:24:29,400 --> 00:24:32,200 their product. And I actually subscribe to 431 00:24:32,200 --> 00:24:37,500 several of these products. And as a co-pilot I have to say 432 00:24:37,500 --> 00:24:41,300 it's a phenomenal, I mean, the body of knowledge, the 433 00:24:41,308 --> 00:24:44,100 efficiency that it helps. And just coming up with new 434 00:24:44,100 --> 00:24:49,400 ideas, it's very fast much faster than than I could do on 435 00:24:49,400 --> 00:24:52,900 my own absolutely and the new ideas I mean they're not really 436 00:24:53,400 --> 00:24:54,900 the machine, this is the thing right? 437 00:24:54,900 --> 00:24:57,400 We shouldn't think about. So Alan a thing. 438 00:24:57,600 --> 00:25:02,700 Us in the sense that we give these to, you know, too much 439 00:25:03,400 --> 00:25:07,600 responsibility in terms of, you know, the level of intelligence 440 00:25:07,600 --> 00:25:10,500 because ultimately there is always like a today. 441 00:25:10,500 --> 00:25:13,600 At least, there is always a human behind the scenes that has 442 00:25:13,600 --> 00:25:17,200 created the content. I structured in a certain way as 443 00:25:17,200 --> 00:25:20,800 trained the model in a certain way so it's still human 444 00:25:20,800 --> 00:25:23,100 knowledge, right? But it's but the nice thing 445 00:25:23,100 --> 00:25:25,600 about it is that you can have access to it immediate. 446 00:25:25,600 --> 00:25:30,900 So you know, I was making the Earlier of the the old school, 447 00:25:30,900 --> 00:25:34,800 you know, Heart, You Know, cover tackle library, right? 448 00:25:34,800 --> 00:25:38,100 You have to find the book, you know in the library to be able 449 00:25:38,100 --> 00:25:40,100 to talk about the specific domain here. 450 00:25:40,100 --> 00:25:43,900 You can we can have access to an incredible amount of data like 451 00:25:43,900 --> 00:25:46,700 very very quickly. We have another really 452 00:25:46,700 --> 00:25:49,800 interesting question from Anil Vos. 453 00:25:50,200 --> 00:25:56,300 And he says, if the data set that trains, the generative, I 454 00:25:56,300 --> 00:26:01,900 say I have As an intrinsic bias, how can we get rid of this bias 455 00:26:01,900 --> 00:26:06,700 or overcome this by us? So that it is neutral. 456 00:26:06,900 --> 00:26:09,100 And I think he's talking in two things. 457 00:26:09,100 --> 00:26:12,700 One from the point of view of the software developer, and 458 00:26:12,700 --> 00:26:17,000 number two, from the point of view, as us as consumers as 459 00:26:17,000 --> 00:26:20,800 users, the people that manage did, they send that drains the 460 00:26:20,800 --> 00:26:23,800 moment, right? That's what fundamentally comes 461 00:26:23,800 --> 00:26:26,900 down to, right? So if you feed it in the model 462 00:26:26,900 --> 00:26:30,400 racist, You know type of Concepts, write the model will 463 00:26:30,500 --> 00:26:33,800 Jason that so that is 0 bias, you know, of any kind. 464 00:26:34,500 --> 00:26:38,600 So it's a fundamentally it's a human responsibility, right? 465 00:26:38,600 --> 00:26:43,200 So in Enterprise has, as I said earlier, we need to govern that 466 00:26:43,700 --> 00:26:47,700 we need to have, you know, a sort of a, first of all we need 467 00:26:47,700 --> 00:26:50,700 to have values, right? The Enterprise level values, the 468 00:26:50,708 --> 00:26:54,000 things that you wouldn't do and not do because they will help 469 00:26:54,000 --> 00:26:56,800 guide. You know what is the you know 470 00:26:56,800 --> 00:27:00,700 what is the Onset of bias and how people should should should 471 00:27:00,700 --> 00:27:03,400 should manage that part, right? And then at the same time, I 472 00:27:03,408 --> 00:27:06,500 think, as I said earlier, during the has to be in an authorial 473 00:27:06,500 --> 00:27:13,500 Board of some sort that oversees and takes the responsibility or 474 00:27:13,900 --> 00:27:17,200 food for what goes into the machine ultimately, right? 475 00:27:17,200 --> 00:27:20,900 So, I think it wouldn't be there is risk for sure, right? 476 00:27:20,900 --> 00:27:22,900 And it's something that we need to take very seriously. 477 00:27:23,400 --> 00:27:26,300 This is again from our Salon, Khan who says, when it comes to 478 00:27:26,300 --> 00:27:32,700 life and death Situations. And the AI is wrong, who should 479 00:27:32,700 --> 00:27:38,600 we Sue AI? Vendor the data owners, the 480 00:27:38,700 --> 00:27:41,900 algorithm creators, who do we Sue? 481 00:27:41,900 --> 00:27:44,800 When there's a, when something really goes wrong, I think we 482 00:27:44,800 --> 00:27:48,000 should not get into position when it's a question of life and 483 00:27:48,000 --> 00:27:53,500 death left to a machine. That to me, is the number one. 484 00:27:53,700 --> 00:27:57,300 So if we had to make, if you have to get to a position where 485 00:27:57,300 --> 00:28:00,300 the machine has to make a decision and it has, you know, 486 00:28:00,300 --> 00:28:05,300 the power of, you know, driving an event of this level. 487 00:28:06,100 --> 00:28:08,300 I think there is something wrong in the way of approaching it. 488 00:28:08,800 --> 00:28:12,200 These Technologies are not at that level yet, right? 489 00:28:12,200 --> 00:28:16,800 So we do and we don't want that, I think, yet until it can be 490 00:28:16,800 --> 00:28:20,500 proven in a different way. So if let's say it's a medical 491 00:28:20,700 --> 00:28:24,300 Your diagnosis that has to be provided. 492 00:28:24,400 --> 00:28:27,700 The expectation is that there is always a physician in charge 493 00:28:27,900 --> 00:28:30,000 with the help of the tools, right? 494 00:28:30,500 --> 00:28:33,700 But fundamentally it's down to the human. 495 00:28:33,700 --> 00:28:38,600 So I don't I mean everybody really even you can think about, 496 00:28:38,600 --> 00:28:43,100 you know, just stretching out a little bit from just the Genera 497 00:28:43,100 --> 00:28:48,000 to the eyepiece of it even for we're not ready yet to do full, 498 00:28:48,000 --> 00:28:50,600 you know, full a full self-driving cars. 499 00:28:51,500 --> 00:28:54,300 Right foot for the number of reasons, but one of the reasons 500 00:28:54,300 --> 00:28:56,900 is not just, the college is about their kind of a who's 501 00:28:56,900 --> 00:28:58,500 liable. Who's accountable? 502 00:28:58,500 --> 00:29:01,300 What kind of ethical decisions they need to be taken excited? 503 00:29:01,300 --> 00:29:04,200 So if you're not already there which is in a way, the more 504 00:29:04,200 --> 00:29:08,400 mechanical way of doing it, we shouldn't be ready to add to 505 00:29:08,500 --> 00:29:09,900 delegate. Again, you know, the 506 00:29:09,900 --> 00:29:13,400 responsibility for these type of decisions to adjust the machine. 507 00:29:13,500 --> 00:29:17,400 Yet we have another great question from this time. 508 00:29:17,400 --> 00:29:23,800 From Elizabeth Shaw on Twitter and she says So why do tools 509 00:29:23,800 --> 00:29:30,300 such as Google barred or chat GPT hallucinate even when it's 510 00:29:30,300 --> 00:29:36,400 using correct Source data and what are the implications there 511 00:29:36,400 --> 00:29:41,100 for for using these kinds of tools for Knowledge Management 512 00:29:41,100 --> 00:29:44,700 inside an organization? Like see why it is better 513 00:29:44,700 --> 00:29:47,900 described because at the end you know it's reinforcement learning 514 00:29:47,900 --> 00:29:51,400 you learn and you know that the answer that gives You know, the 515 00:29:51,400 --> 00:29:55,000 best results are the ones that the win, right? 516 00:29:55,000 --> 00:30:00,300 So there is a sort of a, you know, a selection mechanism of 517 00:30:00,300 --> 00:30:03,200 the answer. So if the answers that are 518 00:30:03,200 --> 00:30:06,300 awarded, are those that are going, you know, on a tangent. 519 00:30:06,500 --> 00:30:08,200 Probably, you can get in this scenario. 520 00:30:08,200 --> 00:30:11,500 This is at the end of the day, it's a statistical tool, right? 521 00:30:11,500 --> 00:30:14,300 So I mean, we call it in artificial intelligence, 522 00:30:14,300 --> 00:30:17,500 whatever, but effectively, it's pathetic, supplication of 523 00:30:17,500 --> 00:30:20,900 Statistics, right? To a very complex problem. 524 00:30:21,000 --> 00:30:22,300 Violent. But distinct statistic. 525 00:30:22,900 --> 00:30:26,400 And so there are up absolutely situations where statistics gets 526 00:30:26,400 --> 00:30:28,600 you into the outlier. Right. 527 00:30:28,600 --> 00:30:30,000 Absolutely. So that's possible. 528 00:30:30,200 --> 00:30:32,300 Even I don't know, I'm not a mathematician but I'm sure that 529 00:30:32,300 --> 00:30:37,000 a lot of wasted that this can be described mathematically in 530 00:30:37,000 --> 00:30:39,800 terms of what needs to happen for the Enterprise. 531 00:30:39,800 --> 00:30:42,800 As I said earlier, right? We should that should not be 532 00:30:42,800 --> 00:30:45,300 allowed, right? Because you don't want to have 533 00:30:45,300 --> 00:30:49,300 you know, a tool speculating on what tax legislation might be. 534 00:30:49,400 --> 00:30:52,800 If I look at our own business All right, so we need to give 535 00:30:52,800 --> 00:30:55,500 specific answers to clients. That's why this kind of a 536 00:30:55,500 --> 00:30:59,900 curated approach to it, you know, for the time being is 537 00:30:59,900 --> 00:31:02,400 absolutely critical. So will be, it's an investment 538 00:31:02,400 --> 00:31:05,900 that every company needs to make both from the bias, the ethical 539 00:31:05,900 --> 00:31:09,800 side, but also, for the content of the, of the answers and the 540 00:31:09,800 --> 00:31:12,200 questions, Etc. So there has to be some sort of 541 00:31:12,208 --> 00:31:15,500 the labeling that happens on the tool. 542 00:31:15,500 --> 00:31:17,900 When you give an answer in your train, it right to make sure 543 00:31:17,900 --> 00:31:20,700 that this is the right, that the answer that is proper, right? 544 00:31:21,100 --> 00:31:25,800 Can I ask to what extent are you thinking through these kinds of 545 00:31:26,000 --> 00:31:29,700 processes at a why we are experimenting? 546 00:31:29,700 --> 00:31:33,300 So we're trying to understand, you know, how much work needs to 547 00:31:33,300 --> 00:31:34,100 be done. Etc. 548 00:31:34,100 --> 00:31:37,800 So we are in the process of defying, all of that. 549 00:31:38,600 --> 00:31:41,700 I think that's why, you know, the previous questions was, you 550 00:31:41,700 --> 00:31:44,400 know, are we gonna replace people? 551 00:31:44,700 --> 00:31:46,600 I don't say that. I think, you know, the more we 552 00:31:46,600 --> 00:31:48,900 use this to us, the more we need to add it up. 553 00:31:48,900 --> 00:31:51,800 Human expertise to it. So it's going to be The opposite 554 00:31:51,800 --> 00:31:55,400 but I think if we put at least for the next you know few years 555 00:31:55,400 --> 00:31:58,900 until these tools mature and there is more knowledge about 556 00:31:58,900 --> 00:32:03,000 it, you know there is definitely the need for for investing in 557 00:32:03,000 --> 00:32:06,500 the Enterprise in human capabilities around that Chris 558 00:32:06,500 --> 00:32:11,200 Peterson on Twitter comes back. And he's asking if you have any 559 00:32:11,200 --> 00:32:15,600 thoughts on cybersecurity issues, created a need for side 560 00:32:15,600 --> 00:32:21,400 for the Cyber insurance industry and do such firms and These have 561 00:32:21,400 --> 00:32:26,400 a role in shaping and governing the use of AI in the Enterprise 562 00:32:26,400 --> 00:32:30,000 regarding liability. I think in the future probably. 563 00:32:30,000 --> 00:32:36,300 Yes, I don't again, I don't see any Enterprise function, 564 00:32:36,400 --> 00:32:41,500 completely delegated to Ai. And so if it's always a sort of 565 00:32:41,500 --> 00:32:45,000 a center model where there is the human and the Machine 566 00:32:45,000 --> 00:32:49,000 together doing something on behalf of the Enterprise, I 567 00:32:49,000 --> 00:32:52,700 think fundamentally it's the same level of liability or 568 00:32:52,700 --> 00:32:57,000 responsibility that you have today in the future, okay? 569 00:32:57,000 --> 00:33:03,100 So for example, let's say, you know, credit decision or other 570 00:33:03,100 --> 00:33:06,000 type of Thanksgiving without going into the most dramatic, 571 00:33:06,100 --> 00:33:08,200 you know, healthcare related decisions. 572 00:33:10,300 --> 00:33:14,300 I say, I don't see why not. That could be some level of 573 00:33:14,600 --> 00:33:17,500 assurance but I have it in my understanding so far. 574 00:33:17,600 --> 00:33:20,000 Okay and it's for sure, not complete. 575 00:33:21,200 --> 00:33:24,600 I haven't seen anybody yet. That is planning to completely 576 00:33:25,300 --> 00:33:27,600 you know run the piece of a business like that. 577 00:33:28,000 --> 00:33:31,600 You know in a schematic fashion what are some of the challenges 578 00:33:31,600 --> 00:33:35,300 that you see to adoption of these kinds of tools in the 579 00:33:35,300 --> 00:33:39,300 Enterprise one is over hope. And then, you know, for most of 580 00:33:39,300 --> 00:33:44,500 my career, AI starting from school, basically, and I've seen 581 00:33:44,500 --> 00:33:47,900 these herbs and flows of excitement, right? 582 00:33:47,900 --> 00:33:51,400 So I started, I started really 90s. 7 and he was like, you 583 00:33:51,400 --> 00:33:53,800 know, the winter of the I nothing really happened for 584 00:33:53,800 --> 00:33:55,500 years, right? Even if there was a lot of 585 00:33:55,508 --> 00:33:59,400 potential in those days, it was all about expert systems. 586 00:33:59,400 --> 00:34:01,600 You know, that kind of knowledge because we didn't really have 587 00:34:01,600 --> 00:34:04,500 the technology in terms of the processing power. 588 00:34:04,500 --> 00:34:08,000 You know, the memory and all those kind of things then he 589 00:34:08,000 --> 00:34:11,600 started getting into, you know, some games, you know they were 590 00:34:11,600 --> 00:34:14,100 win like you know the Chess World Championships, blah blah 591 00:34:14,100 --> 00:34:15,100 blah. So I'm not going to make the 592 00:34:15,100 --> 00:34:18,699 whole history but it was but you know every single time that 593 00:34:18,699 --> 00:34:22,100 there was a major breakthrough like One the big blue that I DM 594 00:34:22,100 --> 00:34:26,800 DT 90s you know beating Castro it was a huge you know you know 595 00:34:26,800 --> 00:34:30,400 the interest and everything then it went down again right from 596 00:34:30,400 --> 00:34:34,500 few years then deep learning came up and we started dealing 597 00:34:34,500 --> 00:34:38,300 much more with you know image recognition which I think is a 598 00:34:38,300 --> 00:34:40,300 great application of artificial intelligence. 599 00:34:40,300 --> 00:34:46,000 So then another big you know Spike of interest that actually 600 00:34:46,000 --> 00:34:48,100 is leading to this new language models right there are 601 00:34:48,100 --> 00:34:53,900 similarities and so Sometimes, you know, we tend to again to 602 00:34:53,900 --> 00:34:57,900 humanize these Technologies to the point that we the our 603 00:34:57,900 --> 00:35:01,800 expectations of what they can do becomes the same level of 604 00:35:01,800 --> 00:35:03,500 expectation. We will have from a human but 605 00:35:03,500 --> 00:35:06,700 it's not that. So I think that if I was an 606 00:35:06,700 --> 00:35:11,200 Enterprise outside of mine because I will spend time trying 607 00:35:11,200 --> 00:35:15,900 to demystify in a way what these things can do. 608 00:35:15,900 --> 00:35:19,600 But at the same time, showing the value that they can provide 609 00:35:19,600 --> 00:35:24,300 because there is a The value of their it's just unfortunately 610 00:35:24,300 --> 00:35:27,200 what happens is that people get overhyped and then they get all 611 00:35:27,200 --> 00:35:32,000 will be disappointed and so you know in to me in a year from now 612 00:35:32,000 --> 00:35:34,600 and we're going to say the tragedy was a great failure. 613 00:35:34,700 --> 00:35:37,700 Even if is actually an incredible Innovation, I don't 614 00:35:37,700 --> 00:35:40,600 think so, right? But it could happen, it could 615 00:35:40,600 --> 00:35:44,600 happen because people, you know, think that these things can do 616 00:35:44,600 --> 00:35:47,300 things that are, you know, comparable to what a human being 617 00:35:47,300 --> 00:35:49,200 can do. They can't right now. 618 00:35:49,700 --> 00:35:52,900 What It is vice. Do you have for folks in the 619 00:35:52,900 --> 00:35:56,300 Enterprise, who are Business, Leaders, and Technology leaders, 620 00:35:56,800 --> 00:36:03,000 who are looking at these kinds of Technologies and trying to 621 00:36:03,000 --> 00:36:06,400 figure out what to do the bottom line is that this is going to go 622 00:36:06,400 --> 00:36:08,900 really fast again, right? I was that saying, in the 623 00:36:08,900 --> 00:36:14,300 beginning, this is going to, it's an arms race between super 624 00:36:14,700 --> 00:36:18,900 powerful companies and they have a lot of money to invest, right? 625 00:36:18,900 --> 00:36:21,800 Then you start getting into to, you know, government except so, 626 00:36:22,000 --> 00:36:23,900 I think this is going to be exponential. 627 00:36:23,900 --> 00:36:27,500 The speed of these Innovation will be exponential. 628 00:36:27,800 --> 00:36:33,300 I think we're just saying the beginning of it and so I think 629 00:36:33,300 --> 00:36:40,000 that two things to think about so one is There is no right now, 630 00:36:40,600 --> 00:36:45,800 you need to be knowing what it is and being able to understand 631 00:36:45,800 --> 00:36:49,800 how to apply to your business. So that is no postponing it a 632 00:36:49,808 --> 00:36:52,100 say, oh, you know, I'm not going to be an early adopter. 633 00:36:52,500 --> 00:36:56,500 I'm going to be very fast. Follower very first, follower is 634 00:36:56,500 --> 00:36:59,400 Superfast, follow it very fast. Photo could be six months. 635 00:36:59,400 --> 00:37:03,600 So I think for large Enterprises that can afford to do that. 636 00:37:03,600 --> 00:37:06,100 I think that is the need of experimenting. 637 00:37:06,500 --> 00:37:10,400 Hiring the right Talent or training the right people to get 638 00:37:10,400 --> 00:37:13,400 to a good handle on where this industry is going, right? 639 00:37:13,400 --> 00:37:17,300 This is its you have to be keeping an eye on this. 640 00:37:17,400 --> 00:37:19,700 That's, that's the bottom line. And then the other thing is 641 00:37:19,700 --> 00:37:21,500 that, I think, oh, lots of questions. 642 00:37:21,500 --> 00:37:24,800 Very, very good questions about the ethical implications, the 643 00:37:24,800 --> 00:37:27,100 bias and all of that. It's really important because it 644 00:37:27,100 --> 00:37:30,000 has a massive brand impact, right? 645 00:37:30,000 --> 00:37:34,000 So there is also a risk management side of around these 646 00:37:34,000 --> 00:37:36,400 technologies that has to be understood very well sir. 647 00:37:36,600 --> 00:37:41,500 If I, you know, I would be a CEO of an organization. 648 00:37:42,100 --> 00:37:45,700 I would push my technology team to have the right skill sets to 649 00:37:45,700 --> 00:37:47,700 understand what it is and where it is going. 650 00:37:47,800 --> 00:37:50,500 And at the same time, I would put in place in a sound risk 651 00:37:50,500 --> 00:37:54,200 management, thinking around it as well to what extent do you 652 00:37:54,200 --> 00:37:59,400 think that the developers of these tools and the folks were 653 00:37:59,400 --> 00:38:04,600 gathering the data sets, need to have that ethical understanding. 654 00:38:04,700 --> 00:38:09,600 And the reason I ask is because Historically technology was kind 655 00:38:09,600 --> 00:38:12,300 of separate from the ethical, or the application of that 656 00:38:12,300 --> 00:38:14,700 technology. From an ethical perspective is 657 00:38:14,700 --> 00:38:16,100 absolutely, do you think room, right? 658 00:38:16,100 --> 00:38:18,500 And you know, think about what's happening, you know we social 659 00:38:18,500 --> 00:38:21,200 media right? So it's how important it is the 660 00:38:21,600 --> 00:38:23,700 content moderation in social media. 661 00:38:24,200 --> 00:38:26,700 So you know, this is not exactly the same of course. 662 00:38:26,700 --> 00:38:30,700 But you know if you translate it into this space in there is 663 00:38:30,800 --> 00:38:34,100 opportunity for abusing these tools everywhere, right? 664 00:38:34,100 --> 00:38:38,100 And they as I said, the brand impact, It is massive. 665 00:38:38,100 --> 00:38:41,100 So there is the need of, you know, setting guidelines. 666 00:38:41,100 --> 00:38:44,100 So for example, internally that is why we have specific 667 00:38:44,100 --> 00:38:46,900 guidelines on how to use the tool what we can do and not do 668 00:38:46,900 --> 00:38:50,100 with it, right for now maybe they're a little bit restricted 669 00:38:50,200 --> 00:38:53,800 but we want to be on the on the safe side of things and then 670 00:38:53,800 --> 00:38:57,900 we're going to open them up a little bit more as time goes by. 671 00:38:57,900 --> 00:39:00,300 But you know, there has to be, but fundament. 672 00:39:00,300 --> 00:39:03,000 Everything starts with the true understanding of what is 673 00:39:03,700 --> 00:39:07,000 Available impossible, right? And if you don't understand that 674 00:39:07,000 --> 00:39:10,200 in detail, it's difficult to put, you know, the boundaries. 675 00:39:10,200 --> 00:39:13,300 And so today we're really at this phase where we're 676 00:39:13,300 --> 00:39:16,000 understanding that. What's reasonable, what's 677 00:39:16,000 --> 00:39:18,100 practical? What can we do? 678 00:39:18,500 --> 00:39:22,900 Yes, and this shouldn't be underestimated in terms of the 679 00:39:22,900 --> 00:39:26,300 ability of driving Innovation, but it shouldn't be either, 680 00:39:26,400 --> 00:39:30,200 overestimated intense of being a sentient being or anything like 681 00:39:30,200 --> 00:39:33,300 that because that, you know, I think it's worth far. 682 00:39:33,500 --> 00:39:36,800 From that, you know, I have to say the interesting thing about 683 00:39:36,800 --> 00:39:39,700 this question to me, is the fact that people have this 684 00:39:40,100 --> 00:39:46,800 conversation because at times these chats million human 685 00:39:46,800 --> 00:39:49,500 conversations, so what? Yeah, it's incredible. 686 00:39:49,500 --> 00:39:52,500 I think what the heck done? It's absolutely one of the, in 687 00:39:52,500 --> 00:39:56,400 my mind, one of the most incredible inventions of the 688 00:39:56,400 --> 00:39:58,400 last definitely. The century, right? 689 00:39:58,400 --> 00:40:02,500 And there is the being the iPhone 5 in mobile phone, social 690 00:40:02,500 --> 00:40:04,000 media, all that kind of stuff. Great. 691 00:40:04,000 --> 00:40:07,600 I think is super important but this one, it goes towards 692 00:40:07,600 --> 00:40:11,500 another level of innovation because it impacts our, you 693 00:40:11,500 --> 00:40:16,300 know, it's really like, you know, it's almost like looks 694 00:40:16,300 --> 00:40:18,400 like reasoning right over human, right? 695 00:40:18,400 --> 00:40:22,000 So it's absolutely incredible. So I'm, you know, I'm super 696 00:40:22,000 --> 00:40:25,400 excited about it. And why don't we finish up with 697 00:40:25,400 --> 00:40:29,500 another question from our Salon? Khan who says on Twitter, what 698 00:40:29,500 --> 00:40:33,300 do you hope future iterations of a I can do. 699 00:40:33,700 --> 00:40:36,200 I think the future of Aviation, I think it has to be more 700 00:40:36,200 --> 00:40:37,500 transparent. Right? 701 00:40:37,500 --> 00:40:42,400 So that you guys have asked fantastic questions about by us, 702 00:40:42,400 --> 00:40:44,800 ethics and all of that. So that is a problem, right? 703 00:40:44,800 --> 00:40:46,900 So if we go deep this, the count, this technology are 704 00:40:46,900 --> 00:40:48,700 powerful. If they go in the wrong hands, 705 00:40:48,700 --> 00:40:52,900 in terms of, you know, being able to generate content as 706 00:40:53,300 --> 00:40:56,600 chests Behavior. So public, you know, 707 00:40:56,600 --> 00:40:59,700 understanding of things, it's really dangerous because we're 708 00:40:59,900 --> 00:41:04,000 basically adding another layer of intermediation between And 709 00:41:04,000 --> 00:41:09,700 the sources of knowledge, right? So if that is done with, not the 710 00:41:09,700 --> 00:41:12,900 right ethical, I guess, you know, it can influence politics, 711 00:41:12,900 --> 00:41:15,400 it can influence the election. We can do a lot of different 712 00:41:15,400 --> 00:41:18,700 things that we don't want, so I think that transparency and 713 00:41:18,700 --> 00:41:22,000 ability to monitor what it does. And often strained, I think is 714 00:41:22,000 --> 00:41:26,300 really important, but then I cannot wait for the future 715 00:41:26,300 --> 00:41:29,700 because you know, like there is so much more in think about the 716 00:41:29,700 --> 00:41:35,600 breakthroughs that we can have, when When the sum of the human 717 00:41:35,600 --> 00:41:38,500 knowledge is so accessible, right? 718 00:41:38,500 --> 00:41:42,000 And in such a smart way. I think, you know, if we take 719 00:41:42,000 --> 00:41:45,000 here so I saw something that you know Monday we're gonna you 720 00:41:45,000 --> 00:41:48,000 know, really Implement nuclear fusion, that's not the thing. 721 00:41:48,000 --> 00:41:52,300 But in terms of stimulating a different level of thinking and 722 00:41:52,500 --> 00:41:55,800 summarized in this knowledge, I mean it's incredible. 723 00:41:55,800 --> 00:41:59,200 I'm so excited about it. We have one last question that 724 00:41:59,200 --> 00:42:02,100 snuck in under the wire and so how about if we finish out with 725 00:42:02,100 --> 00:42:08,600 something from Melvin aguiar who says, how comparable is our 726 00:42:08,600 --> 00:42:12,400 products like chat GPT to low code? 727 00:42:12,400 --> 00:42:15,200 No code. Will they coexist or eliminate 728 00:42:16,000 --> 00:42:18,100 one? Eliminate the other. 729 00:42:19,000 --> 00:42:23,300 And I think he's referring to the fact that these tools are 730 00:42:23,300 --> 00:42:25,500 really optimized to help write code. 731 00:42:25,800 --> 00:42:28,800 Yes. I think right now, it's not that 732 00:42:28,800 --> 00:42:31,300 big yet, but at definitely that is Direction. 733 00:42:31,300 --> 00:42:33,900 So you would ask, you know how much code would Do you have to 734 00:42:33,900 --> 00:42:37,400 write in in five years? When you can actually talk to a 735 00:42:38,200 --> 00:42:42,200 tool and say, can you build me this application with these 736 00:42:42,200 --> 00:42:46,200 characteristics? These inputs outputs data, Etc. 737 00:42:46,200 --> 00:42:50,100 So I think the direction, I think the what I say like you 738 00:42:50,100 --> 00:42:53,800 know the writing is on the wall. In terms of, you know, software 739 00:42:53,800 --> 00:42:56,300 engineering will have to be changed completely. 740 00:42:56,300 --> 00:42:59,500 It doesn't mean that they're not going to be, there's not going 741 00:42:59,500 --> 00:43:01,600 to be software Engineers. Absolutely not. 742 00:43:01,700 --> 00:43:03,900 But I think the role of the software You would have to 743 00:43:03,900 --> 00:43:06,900 change a little bit more and get closer to the business because 744 00:43:06,900 --> 00:43:09,500 you still need a human. They can summarize. 745 00:43:09,500 --> 00:43:12,500 Those needs and requirements into and ask the machine. 746 00:43:12,900 --> 00:43:15,300 What do you need? You know, armies of people to 747 00:43:15,300 --> 00:43:18,700 code you know routines and different lengths, programming 748 00:43:18,700 --> 00:43:22,600 languages, maybe not. And with that unfortunately 749 00:43:22,600 --> 00:43:25,500 we're out of time. It's been a fascinating 750 00:43:25,500 --> 00:43:28,700 conversation. I want to say a huge thank you 751 00:43:28,700 --> 00:43:32,600 to Nicola Marini beings. You know Nicola thank you for 752 00:43:32,600 --> 00:43:33,900 being here with us. Us today. 753 00:43:33,900 --> 00:43:36,600 Thank you for having me Michael and everybody. 754 00:43:36,600 --> 00:43:39,400 Thank you for watching. Especially those folks who asked 755 00:43:39,400 --> 00:43:42,400 such great questions. I always encourage you to watch 756 00:43:42,400 --> 00:43:44,600 and ask questions. You guys are an amazing 757 00:43:44,800 --> 00:43:49,100 audience, but before you go, please subscribe to our YouTube 758 00:43:49,100 --> 00:43:51,800 channel. Hit the Subscribe button at the 759 00:43:51,800 --> 00:43:57,700 top of 60 talk.com, our website, and subscribe to our newsletter 760 00:43:58,400 --> 00:44:01,700 and tune in for our live shows and we will see you again next 761 00:44:01,700 --> 00:44:03,000 time. Have a great day everybody.