WEBVTT - Meta's Zuckerberg on Largest Open-Source AI Model Ever 0:00:00.600 --> 0:00:10.959 Bloomberg Audio Studios, podcasts, radio news. This is Bloomberg BusinessWeek 0:00:11.280 --> 0:00:14.560 inside from the reporters and editors who bring you America's 0:00:14.560 --> 0:00:19.200 most trusted business magazine, plus global business, finance and tech news. 0:00:19.360 --> 0:00:23.479 The Bloomberg Business Week Podcast with Carol Messer and Tim 0:00:23.480 --> 0:00:25.800 Stenebeck from Bloomberg Radio. 0:00:27.480 --> 0:00:30.480 Well Mark Zuckerberg inspired his latest salvo in the AI wars. 0:00:30.760 --> 0:00:34.000 Just today, Meta Platforms unveiled the largest open source AI 0:00:34.040 --> 0:00:36.760 model ever. It's called Lama three point one. The company 0:00:36.920 --> 0:00:39.680 is also updating its meta AI service, built as a 0:00:39.800 --> 0:00:42.240 rival too chat GPT, and in. 0:00:42.080 --> 0:00:44.760 An exclusive interview with Bloomberg's Emily Chang, the Meta CEO 0:00:44.800 --> 0:00:47.159 and founder opened up about his strategy and why he 0:00:47.200 --> 0:00:50.800 believes it's superior to competitors such as Open ai and Google. 0:00:50.840 --> 0:00:54.040 Bloomberg Original's host and executive producer Emily Chang joining us 0:00:54.080 --> 0:00:56.680 now from our San Francisco bureau. 0:00:56.800 --> 0:00:58.959 Emily spoke to Mark Zuckerberg for the circuit. 0:00:59.040 --> 0:01:01.440 It is a Bloomberg Original series and it's hosted by her, 0:01:01.920 --> 0:01:05.240 where she goes behind the scenes with founders, executives, and innovators. 0:01:05.280 --> 0:01:07.320 Sometimes gets to wear pink high heels, but I'm not 0:01:07.360 --> 0:01:08.160 going to talk about. 0:01:08.000 --> 0:01:10.360 That right now. Good to have you back with us. 0:01:10.680 --> 0:01:14.000 Mark Zuckerberg, Man, when we think about big megacap tech companies, 0:01:14.720 --> 0:01:17.000 that is certainly one of them in a big way. 0:01:17.840 --> 0:01:20.160 AI sounds like it was top of mind for you guys. 0:01:21.240 --> 0:01:23.640 Well, AI is their big news today, and Mark has 0:01:23.640 --> 0:01:26.720 been really excited about this new model, the biggest open 0:01:26.760 --> 0:01:30.440 source model ever. He says, it takes it to another level, 0:01:30.520 --> 0:01:32.720 like you can think about it like a teacher, and 0:01:32.760 --> 0:01:37.160 it will teach all the other models, smaller models below it. 0:01:37.319 --> 0:01:40.319 And so you know, getting to this milestone was a 0:01:40.319 --> 0:01:43.200 really big deal for them. They've obviously invested billions of 0:01:43.200 --> 0:01:45.280 dollars to get here and to try to catch up 0:01:45.959 --> 0:01:49.160 to open Ai and Google. So the interesting thing is 0:01:49.160 --> 0:01:51.040 they're really coming from behind. But what he thinks is 0:01:51.040 --> 0:01:54.560 with by the next update, they're going to completely have 0:01:54.680 --> 0:01:57.640 closed the gap. So you know, it was a really 0:01:57.920 --> 0:02:00.000 rare opportunity to sit down with him and talk about 0:02:00.040 --> 0:02:03.280 out his vision. And you know, I've met him, I've 0:02:03.280 --> 0:02:07.520 interviewed him over the years. I'd also say, you know, 0:02:07.560 --> 0:02:09.520 he seems a lot more comfortable in his own skin 0:02:09.600 --> 0:02:14.200 and just like answering questions more more directly and and 0:02:14.400 --> 0:02:17.280 that was really refreshing. I thought, I thought we got 0:02:17.280 --> 0:02:19.919 to see a more more real, more human. 0:02:19.960 --> 0:02:20.480 Side of him. 0:02:20.560 --> 0:02:22.359 Okay, Emily, I want to talk about Lama. I want 0:02:22.360 --> 0:02:24.320 to talk about Ai, but really what I want to 0:02:24.320 --> 0:02:28.000 talk about. Yeah, so think it's so fun. I mean 0:02:28.240 --> 0:02:31.720 it's like it's it's I think it's easier than regular surfing. 0:02:31.760 --> 0:02:33.680 I don't know if how you felt about it. You 0:02:34.040 --> 0:02:35.880 they talked in the you talked in the program. Got 0:02:35.880 --> 0:02:37.960 how you were from Hawaii? You got you got up 0:02:37.960 --> 0:02:40.040 there and you wakesurf. We're looking at pictures right now. 0:02:41.080 --> 0:02:44.400 You're looking at me wiping out. No, come go wiping 0:02:44.440 --> 0:02:49.120 out again, wiping out again. Yeah, I think we're going 0:02:49.200 --> 0:02:50.960 to show you the climax here. 0:02:51.200 --> 0:02:55.079 So look it was I'm from Hawaii, so I have 0:02:55.160 --> 0:02:57.560 surfed before, but I'm not a surfer. Wakesurfing is a 0:02:57.560 --> 0:02:58.440 totally different thing. 0:02:58.520 --> 0:02:59.080 It was hard. 0:02:59.400 --> 0:03:02.200 But actually Mark and Priscilla joined us as well. Was 0:03:02.200 --> 0:03:05.160 a good teacher and he was super patient with me. 0:03:05.280 --> 0:03:06.959 Which you know, what did. 0:03:06.800 --> 0:03:09.320 You what did you throw to What did you throw 0:03:09.360 --> 0:03:09.639 to him? 0:03:10.040 --> 0:03:12.160 Was it a beer through a lacroix? It was a 0:03:12.240 --> 0:03:15.919 la croix. We were we were non alcoholic on this journey. 0:03:15.919 --> 0:03:16.840 But there, look, I did it. 0:03:17.040 --> 0:03:19.680 Yeah, that's what I'm talking about. That's what I'm talking about. 0:03:20.840 --> 0:03:24.440 And you know it was it was brief, but hey, 0:03:24.840 --> 0:03:28.080 it happened. And and I think the fact that he 0:03:28.160 --> 0:03:30.720 was willing to have us at his home and to 0:03:30.800 --> 0:03:33.760 show us this side of himself, you know, shows a 0:03:33.760 --> 0:03:36.160 lot about his evolution as a leader and sort of 0:03:36.160 --> 0:03:38.200 where he's at and just sort of being you know, 0:03:38.400 --> 0:03:41.720 confident and comfortable in his own skin. But we also 0:03:41.720 --> 0:03:43.920 talked about how all these different side quests, you know, 0:03:43.920 --> 0:03:45.480 as you know, it's not just wake surfing, but he's 0:03:45.520 --> 0:03:48.360 doing Mma and jiu jitsu, and he learned Mandarin, and 0:03:48.360 --> 0:03:52.120 he's growing you know, his own wagu beef cows and 0:03:52.120 --> 0:03:54.680 phoebe the macnuts in Kawhi, but you know, all of 0:03:54.720 --> 0:03:57.560 these other pursuits. He just has a ton of energy, 0:03:57.760 --> 0:04:00.320 at least that's what Priscilla says, and do all of 0:04:00.320 --> 0:04:03.200 these other things enables him to focus. But the thing 0:04:03.240 --> 0:04:06.840 he loves most doing still, he says, is building and 0:04:06.880 --> 0:04:08.880 that means the next generation of AI. 0:04:09.160 --> 0:04:10.800 Well, before we get to that, because we do have 0:04:10.840 --> 0:04:14.800 a clip from the show in your conversation Emily, Why 0:04:14.840 --> 0:04:17.679 do you think he was so open to showing everything? 0:04:17.680 --> 0:04:20.080 As you said, You've had many conversations over the years, 0:04:20.080 --> 0:04:22.400 this was a different side you felt like, or that 0:04:22.440 --> 0:04:24.240 he was much more comfortable in his skin. Why do 0:04:24.279 --> 0:04:26.400 you think he's ready to kind of reveal that. Is 0:04:26.520 --> 0:04:29.040 that just being older and kind of living life a 0:04:29.080 --> 0:04:29.640 little bit more. 0:04:30.760 --> 0:04:32.960 Yeah, I think it's just a matter of time. You know, 0:04:33.040 --> 0:04:35.640 he talked about how when he started the company twenty 0:04:35.760 --> 0:04:38.160 years ago, I mean it was twenty years ago that 0:04:38.240 --> 0:04:41.520 he was sort of thrust into the spotlight. He just said, 0:04:41.800 --> 0:04:44.120 I didn't know anything about running a company, and I 0:04:44.160 --> 0:04:46.240 would totally mess up in interviews, and then when I 0:04:46.279 --> 0:04:47.920 messed up, I would get all this negative feedback, and 0:04:47.920 --> 0:04:49.599 then I would claim up and I just wouldn't want 0:04:50.000 --> 0:04:50.920 to talk to anyone. 0:04:51.560 --> 0:04:53.960 And over time he's just. 0:04:53.960 --> 0:04:57.800 Gotten a lot more comfortable and confident saying what he thinks. 0:04:57.880 --> 0:05:00.200 And he said, if people don't like it, that's fine. 0:05:00.000 --> 0:05:02.840 And if people like it, great, Like you judge for yourself. 0:05:02.839 --> 0:05:07.960 This is who I am. And you know, one of 0:05:08.000 --> 0:05:10.240 the other interesting things is, you know we've talked about 0:05:10.360 --> 0:05:13.240 Jeff Bezos has has left Amazon, Larry and SERGEI have 0:05:13.320 --> 0:05:16.039 left Google's Google there you know, are these other big 0:05:16.040 --> 0:05:18.120 tech founders who have taken a step back. I didn't 0:05:18.120 --> 0:05:20.280 get the sense that that was happening at all. Like 0:05:20.320 --> 0:05:23.360 Mark Zuckerberg is not going anywhere. You know, he just 0:05:23.400 --> 0:05:27.000 turned forty. He seems to have more energy than ever, 0:05:27.560 --> 0:05:30.799 and he's really excited about the next ten to fifteen 0:05:30.880 --> 0:05:34.239 years building out this AI vision, along with the metaverse 0:05:34.240 --> 0:05:35.599 and augmented and virtual reality. 0:05:35.640 --> 0:05:37.200 All of that is still a. 0:05:37.160 --> 0:05:38.920 Thing, but he says in the next ten to fifteen 0:05:38.960 --> 0:05:42.800 years then the next platform will emerge, and building is 0:05:42.839 --> 0:05:44.120 what makes him the most happy. 0:05:44.240 --> 0:05:47.479 Well, speaking of that AI vision, here's Mark Zuckerberg from 0:05:47.520 --> 0:05:50.800 your interview on the Circuit on controlling his AI destiny. 0:05:50.880 --> 0:05:51.760 Check it out. 0:05:52.160 --> 0:05:55.480 Some people see you as an unlikely champion of open 0:05:55.560 --> 0:05:57.280 source today you're laughing. 0:05:57.279 --> 0:05:58.120 Well, I don't know why. 0:05:58.360 --> 0:06:01.160 I mean, I actually think, well, I. 0:06:01.120 --> 0:06:03.279 Get it, but yeah, you understand the word unlikely. 0:06:03.480 --> 0:06:08.200 Yeah, we've actually Meta have been pretty big proponents of 0:06:08.240 --> 0:06:10.920 open source for a while and it's been a really 0:06:10.960 --> 0:06:13.800 good formula for us. So dating back to how we 0:06:13.839 --> 0:06:16.479 design servers, how we design our data centers, we have 0:06:16.520 --> 0:06:19.600 this whole open compute project that kind of standardized a 0:06:19.640 --> 0:06:22.960 lot of the infrastructure for the industry that's going to 0:06:22.960 --> 0:06:23.839 happen with AI too. 0:06:24.240 --> 0:06:26.880 You're really putting a stake in the ground by open 0:06:26.880 --> 0:06:30.039 sourcing your AI in this attempt to build the AI 0:06:30.200 --> 0:06:34.240 rails for the future. How much of this is a 0:06:34.240 --> 0:06:38.080 strategic way to control or own the next technology wave. 0:06:38.640 --> 0:06:40.320 A lot of how we've grown up over the last 0:06:40.400 --> 0:06:44.920 ten or fifteen years was building our apps through phone 0:06:45.000 --> 0:06:49.520 platforms that our competitors controlled. So it's somewhat soul crushing 0:06:49.560 --> 0:06:52.279 to like go build something that you think is going 0:06:52.320 --> 0:06:54.279 to be good and then just get told by Apple 0:06:54.320 --> 0:06:56.040 that you can't ship it because they want to put 0:06:56.080 --> 0:06:57.440 us in a box because they view. 0:06:57.400 --> 0:06:58.200 Us as competitive. 0:06:58.480 --> 0:07:01.200 You know, we're a big enough company now that one 0:07:01.240 --> 0:07:03.919 of the things that I've resolved is that for the 0:07:03.960 --> 0:07:07.479 next generation of technology, I want us to build and 0:07:07.800 --> 0:07:10.120 have more control over the next set of platforms that 0:07:10.160 --> 0:07:11.560 we're going to build. So I think AI is a 0:07:11.560 --> 0:07:15.400 critical one, and I think augmented and virtual reality is 0:07:15.440 --> 0:07:18.400 another critical one. It's kind of the way that we 0:07:18.440 --> 0:07:20.640 can control our own destiny on this and make sure 0:07:20.680 --> 0:07:23.760 that we have access to leading AI is by building 0:07:23.840 --> 0:07:26.760 it and having it become an industry standard, but it 0:07:26.840 --> 0:07:29.960 actually gets stronger by being able to share it and 0:07:29.960 --> 0:07:31.400 have the ecosystem around it. 0:07:31.920 --> 0:07:35.280 You're continuing to improve meta AI across all of your products, 0:07:35.280 --> 0:07:38.120 but also as a standalone chatbot. Why should we use 0:07:38.160 --> 0:07:39.840 meta over chat GPT? 0:07:40.880 --> 0:07:43.400 Well, there's a bunch of things where it's better, you know. 0:07:43.440 --> 0:07:46.040 One of the things that we're rolling out soon is 0:07:46.360 --> 0:07:49.280 the ability to just like imagine stuff. You're typing something 0:07:49.280 --> 0:07:50.960 in real time. I do this with my daughters all 0:07:50.960 --> 0:07:56.280 the time, and as you're kind of typing and entering 0:07:56.800 --> 0:08:00.080 the query, it's just generating the images as you enter. 0:08:00.080 --> 0:08:00.680 The key strokes. 0:08:00.680 --> 0:08:04.680 It's just really cool. My goal for the medai launch, 0:08:04.760 --> 0:08:07.960 which I mean it's only really a few months old 0:08:08.120 --> 0:08:10.360 at this point, was to by the end of the 0:08:10.440 --> 0:08:13.600 year have met aai'd be the most used AI assistant 0:08:13.640 --> 0:08:15.960 in the world. And I think we're basically on track 0:08:16.000 --> 0:08:17.400 for that. I mean, there's hundreds of millions of people 0:08:17.400 --> 0:08:19.280 who are at the end of this year, yeah, and 0:08:19.280 --> 0:08:20.560 I think we're going to be there before the end 0:08:20.600 --> 0:08:20.960 of the year. 0:08:21.720 --> 0:08:26.120 That's Mark Zuckerberg with Emily Chang on the Circuit Bloomberg 0:08:26.120 --> 0:08:29.160 original series that goes behind the scenes with founders, executives, 0:08:29.200 --> 0:08:31.880 and innovators. Emily, I have to tell you the Mark 0:08:31.960 --> 0:08:35.120 Zuckerberg that we just heard in that clip, he sounds 0:08:35.160 --> 0:08:38.880 different than the Mark Zuckerberg we've heard on stage at 0:08:39.480 --> 0:08:42.080 the Developers conference that we've heard in other interviews. He 0:08:42.160 --> 0:08:43.880 sounded so much more relaxed. 0:08:45.120 --> 0:08:46.280 I completely agree. 0:08:46.320 --> 0:08:48.520 I mean, that question that you saw me ask about 0:08:48.559 --> 0:08:51.360 being the unlikely Champion was basically the first or maybe 0:08:51.360 --> 0:08:52.720 the second question in the interview. 0:08:52.800 --> 0:08:55.400 And just the fact that he kind of laughed. 0:08:55.080 --> 0:08:57.680 And pushed back a little but then acknowledged that he understood. 0:08:57.720 --> 0:09:00.240 I just thought it was such a great moment. You know, 0:09:00.320 --> 0:09:04.680 he was a lot more direct in this interview than 0:09:04.920 --> 0:09:07.280 I've seen him before. I mean, and look, this is 0:09:07.280 --> 0:09:10.640 still his strategy is highly scrutinized. He's still one of 0:09:10.640 --> 0:09:12.800 the most scrutinized people, and Meta is one of the 0:09:12.840 --> 0:09:15.760 most scrutiny scrutinized companies in the world, and they deserve 0:09:15.800 --> 0:09:20.000 that scrutiny because they are hugely, hugely influential. It just 0:09:20.040 --> 0:09:25.920 seems like he's really grown into this role. And uh, 0:09:25.960 --> 0:09:29.000 you know, I'm glad that he is. He's now showing 0:09:29.000 --> 0:09:30.640 a bit of that side of himself to us so 0:09:30.679 --> 0:09:33.840 that we can better understand how he thinks and how 0:09:33.840 --> 0:09:36.400 he leads, and then we can judge for ourselves well. 0:09:36.400 --> 0:09:40.200 Speaking of influence and showing himself on a day where 0:09:40.200 --> 0:09:44.440 we've been focused on Kamala Harris her first campaign rally, 0:09:44.679 --> 0:09:46.160 we were just covering it moments ago. 0:09:46.320 --> 0:09:48.360 You did talk politics with Mark Zuckerberg. 0:09:49.480 --> 0:09:51.240 We did, and. 0:09:52.840 --> 0:09:55.880 He said more than I expected that he would say. 0:09:55.920 --> 0:09:59.080 He you know, you know, first obviously talked about Facebook's 0:09:59.120 --> 0:10:00.800 role in elections and and one of the things that 0:10:00.800 --> 0:10:04.680 they're doing at Facebook is trying to recommend less political 0:10:04.679 --> 0:10:06.840 content in feeds. He said, the number one thing I 0:10:06.880 --> 0:10:08.319 hear from people is that they don't want to see 0:10:08.320 --> 0:10:10.920 political content, and we're gonna we're listening to that. So 0:10:10.960 --> 0:10:13.040 I think Facebook will be playing less of a role 0:10:13.440 --> 0:10:15.760 in the election going forward. But of course he went 0:10:15.800 --> 0:10:20.160 on to talk about Donald Trump and the appeal that 0:10:20.160 --> 0:10:23.760 Trump has, and he did not endorse either candidate, by 0:10:23.800 --> 0:10:28.160 the way, but he did say that he found former 0:10:28.160 --> 0:10:31.720 President Trump's response to the assassination, that that famous picture 0:10:31.760 --> 0:10:32.600 of him pumping his. 0:10:32.520 --> 0:10:35.840 Fist that we all now know so well. He thought 0:10:35.920 --> 0:10:37.359 that was pretty inspiring. 0:10:37.440 --> 0:10:40.880 In fact, he used the word badass to describe Trump's 0:10:40.880 --> 0:10:43.920 reaction to the assassination attempt, and he was like, I 0:10:43.920 --> 0:10:45.840 can see why people like the guy. It's a crazy 0:10:45.880 --> 0:10:49.600 time in America. And the role that I want to 0:10:49.600 --> 0:10:53.839 play is help letting people express themselves but in a 0:10:54.679 --> 0:10:58.360 safe and healthy way, and we're going to try to 0:10:58.360 --> 0:10:59.960 evolve the platform to make that happen. 0:11:00.200 --> 0:11:04.120 Well, incredible stuff. Another incredible episode and interview, Emily, Thanks 0:11:04.160 --> 0:11:06.599 so much, so glad we could catch up with you. You 0:11:06.320 --> 0:11:08.480 can watch The Circuit with Emily Chang tonight at six 0:11:08.600 --> 0:11:11.280 thirty pm Wall Street Time on Bloomberg Television. Also streaming 0:11:11.320 --> 0:11:14.319 at eight thirty pm Wall Street Time Bloomberg Originals. Emily, 0:11:14.360 --> 0:11:17.679 of course, is the host and executive producer of the 0:11:17.720 --> 0:11:20.160 Circuit from Bloomberg Originals. 0:11:20.840 --> 0:11:24.360 You're listening to the Bloomberg Business Week podcast. Catch us 0:11:24.400 --> 0:11:27.640 Live weekday afternoons from two to five pm Eastern Listen 0:11:27.679 --> 0:11:29.839 on Apple car Play and and Broud Auto with a 0:11:29.840 --> 0:11:34.000 Bloomberg Business act or want us Live on YouTube. 0:11:34.640 --> 0:11:36.320 Cheryl, how many times do you keep heard that AI 0:11:36.440 --> 0:11:36.960 is going to make. 0:11:36.880 --> 0:11:37.560 Their lives easier? 0:11:37.800 --> 0:11:38.960 What in the last sixty seconds. 0:11:39.080 --> 0:11:41.560 Yeah, right, it's going to do the busy work for 0:11:41.640 --> 0:11:43.720 us so we can get to the stuff that really matters. 0:11:44.320 --> 0:11:46.560 Even though I was actually typing this in Google Docs 0:11:46.600 --> 0:11:48.320 and like the automation was trying to fill it in, 0:11:48.760 --> 0:11:50.360 that's kind of what I'm actually trying to say. So 0:11:50.400 --> 0:11:51.439 it's a little impressive. 0:11:51.520 --> 0:11:53.880 I like it when it finishes your sentences or my sentences. 0:11:53.960 --> 0:11:55.880 Okay, So there's this new study out from the freelancing 0:11:55.920 --> 0:11:59.160 platform upwork that actually found that full time employees say 0:11:59.200 --> 0:12:02.360 that AI has added to their workload and then it's 0:12:02.400 --> 0:12:03.880 even contributing to burnout. 0:12:03.920 --> 0:12:05.079 Wait, added to your work? 0:12:05.120 --> 0:12:06.840 Yeah, it's not making things easier, all right. 0:12:06.760 --> 0:12:07.360 So what's going on? 0:12:07.440 --> 0:12:10.600 Kelly Monahan is managing director at the Epwork Research Institute. 0:12:10.640 --> 0:12:12.920 They focus on the future of work, which includes a 0:12:12.920 --> 0:12:15.360 lot of conversations I'm assuming about AI, She joins us 0:12:15.360 --> 0:12:16.000 from Chicago. 0:12:16.080 --> 0:12:17.680 Kelly, what the heck is going on? 0:12:17.760 --> 0:12:19.840 I thought this is supposed to make our life easier 0:12:20.000 --> 0:12:22.000 for us up to do the things that really matter. 0:12:23.920 --> 0:12:26.080 Absolutely. I think what we're seeing right now is the 0:12:26.080 --> 0:12:29.200 classic productivity paradox where we've seen many times before with 0:12:29.280 --> 0:12:32.120 other technologies. You saw on our study ninety six percent 0:12:32.120 --> 0:12:34.760 of global leaders are saying generative AI should be making 0:12:34.800 --> 0:12:37.880 the workforce more productive, but the reality is seventy seven 0:12:37.920 --> 0:12:40.880 percent said is actually adding to their workload. And then 0:12:40.920 --> 0:12:43.559 I think, maybe even more alarming, forty four percent said 0:12:43.559 --> 0:12:45.600 they don't even know how to get these productivity gains, 0:12:45.760 --> 0:12:48.280 and that's resulting in one in three employees looking to 0:12:48.320 --> 0:12:50.880 exit the organization because they're feeling so overwhelmed by their 0:12:50.960 --> 0:12:51.520 job today. 0:12:51.880 --> 0:12:54.199 Wow, so what's going on here? 0:12:54.240 --> 0:12:54.320 Like? 0:12:54.400 --> 0:12:56.840 Why where is the failure happening? 0:12:57.960 --> 0:13:01.400 Yeah? Great question. So a couple of things going on. One, 0:13:01.520 --> 0:13:03.960 this is still a very new technology and we are 0:13:03.960 --> 0:13:06.720 still in net learning phase. I think the mindset shift 0:13:06.720 --> 0:13:09.000 that needs to take place is leaders are looking for 0:13:09.000 --> 0:13:11.400 this to drive efficiency and that's how they're measuring it, 0:13:11.440 --> 0:13:13.280 and that's where they're putting the pressure in their workforce, 0:13:13.679 --> 0:13:16.520 when in reality is there's so much unknowns and this 0:13:16.600 --> 0:13:18.680 is a frontier and so what really needs to be 0:13:18.679 --> 0:13:21.240 taking place is employees thinking of this through an R 0:13:21.280 --> 0:13:23.120 and D lens. What do I need to be learning? 0:13:23.160 --> 0:13:25.280 How do I drive innovation? How do I rethink what 0:13:25.320 --> 0:13:28.079 I'm doing? But instead, what they're feeling is that leaders 0:13:28.080 --> 0:13:29.960 are asking them to do more of the same but 0:13:30.120 --> 0:13:32.360 faster and cheaper, and that right now is just a 0:13:32.360 --> 0:13:34.319 big disconnect because that's not where we are from a 0:13:34.360 --> 0:13:35.360 technology perspective. 0:13:36.000 --> 0:13:39.160 I want to understand, though, exactly what you guys looked 0:13:39.200 --> 0:13:43.800 into in terms of and how what were employees doing 0:13:43.840 --> 0:13:46.040 with AI as they try to figure out how it 0:13:46.080 --> 0:13:48.600 can either help or is just adding to their workload? 0:13:48.640 --> 0:13:51.040 Like what did you exactly look into the research study 0:13:51.080 --> 0:13:53.840 or was it just a case of questioning, you know, employees, 0:13:53.840 --> 0:13:56.760 because I'm curious about who these workers are and what 0:13:57.000 --> 0:13:58.680 they how they were using AI. 0:14:00.600 --> 0:14:02.720 Yeah, so our study is based on knowledge workers and 0:14:02.760 --> 0:14:04.800 this was a global study, so we have representation in 0:14:04.840 --> 0:14:09.240 the US, Australia, Canada, in Europe. And so what we 0:14:09.280 --> 0:14:11.440 really focused on were those workers who are already using 0:14:11.440 --> 0:14:14.440 computer technologies to do their job and so they had 0:14:14.440 --> 0:14:17.000 to have some level of exposure to even understanding AI. 0:14:17.640 --> 0:14:19.560 And what we found was, but they're actually spending a 0:14:19.560 --> 0:14:21.920 lot of their time on right now is auditing the 0:14:21.960 --> 0:14:24.760 AI content that's coming out. They're then having to train 0:14:24.960 --> 0:14:26.800 the AI content because a lot of these are large 0:14:26.880 --> 0:14:29.280 language models that need to be fine tuned and trained. 0:14:29.640 --> 0:14:31.880 And the real disconnect that we're seeing is the majority 0:14:31.880 --> 0:14:35.520 of these individuals are self teaching themselves. There is no 0:14:35.560 --> 0:14:38.880 formal structure with an organization that's helping them upskill and 0:14:38.960 --> 0:14:40.680 actually learn how to do these things, which is why 0:14:40.720 --> 0:14:42.160 I think there's so much friction today. 0:14:42.240 --> 0:14:44.680 Is there any sort of parallel that we can draw 0:14:44.680 --> 0:14:48.520 in recent history where we've seen a technology that in 0:14:48.600 --> 0:14:54.360 your view, actually will be this just has been this 0:14:54.480 --> 0:14:58.400 disruptive and companies have had to upscale their workforce. 0:15:00.760 --> 0:15:02.520 Yeah, so we've seen this, I think in a few 0:15:02.520 --> 0:15:04.240 different times. And now AI has been around since the 0:15:04.320 --> 0:15:06.680 nineteen fifties and I know it's gone through periods of 0:15:06.720 --> 0:15:08.960 winters where it's been overhyped. And now we're in a 0:15:08.960 --> 0:15:11.360 period summer period where there's a lot of investment and 0:15:11.400 --> 0:15:12.920 there's a lot of hype and what this can cause. 0:15:13.360 --> 0:15:15.080 And I think this is a good thing, but the 0:15:15.120 --> 0:15:17.880 reality is all technologies before it have gone through this. 0:15:18.000 --> 0:15:20.880 A lot of people compared this to internet computing, so 0:15:20.920 --> 0:15:23.360 when businesses were mostly done on site and then everyone 0:15:23.400 --> 0:15:26.040 had to quickly become a digital some version of e 0:15:26.240 --> 0:15:29.920 commerce business in order to reach customers. I think over 0:15:29.960 --> 0:15:31.640 time we're going to see that being a pretty good 0:15:31.680 --> 0:15:34.640 comparison and fundamental shift. We're pretty soon you're going to 0:15:34.680 --> 0:15:37.320 hear the conversation around all businesses have to be AI 0:15:37.600 --> 0:15:40.520 powered to some degree. We've seen that with cloud computing. 0:15:40.880 --> 0:15:42.480 I think the problem is right now, this is just 0:15:42.520 --> 0:15:45.160 happening so fast, and we're not giving people enough time 0:15:45.240 --> 0:15:48.040 and enough runway to really make this pivot and change, 0:15:48.040 --> 0:15:49.520 and we're expecting too much too soon. 0:15:49.640 --> 0:15:50.960 Why do you think it's important that we have this 0:15:51.040 --> 0:15:52.080 conversation right now? 0:15:54.160 --> 0:15:57.120 I think right now there's so much pressure of business leaders. 0:15:57.160 --> 0:16:00.800 I mean, they're managing such great macrock. I'm a concerns 0:16:00.880 --> 0:16:04.640 whether it's rising inflation, whether it's global turbulence, and I 0:16:04.640 --> 0:16:07.120 think right now the workforce is looking for something to 0:16:07.520 --> 0:16:10.320 grasp onto to help make their work lives easier. They've 0:16:10.320 --> 0:16:11.960 been through a lot as we think about coming out 0:16:11.960 --> 0:16:14.320 of the pandemic. And what my concern is and why 0:16:14.320 --> 0:16:16.680 I think this conversation is so important is this is 0:16:16.720 --> 0:16:19.840 a human plus machine conversation. There's so much headlines about 0:16:19.840 --> 0:16:21.800 dis generative AI and how it's going to save the day, 0:16:22.160 --> 0:16:24.560 but in reality, humans are going to help save the day, 0:16:24.600 --> 0:16:26.920 and we really need a new model of leadership and 0:16:27.000 --> 0:16:29.040 talent and work mix in order to get this right. 0:16:29.080 --> 0:16:31.440 And I think we're at a critical cross point that 0:16:31.480 --> 0:16:34.280 if we have too much technology investment without enough human 0:16:34.320 --> 0:16:36.760 development investment, we may be leaving people behind. 0:16:37.000 --> 0:16:39.040 So I want to get to some takeaways here before 0:16:39.040 --> 0:16:41.680 we let you go. Takeaways for knowledge workers who are 0:16:41.680 --> 0:16:45.160 out there dealing with this brave new world, people like 0:16:45.240 --> 0:16:47.840 us who are trying to understand and incorporate AI into 0:16:47.840 --> 0:16:49.480 our lives. What should we be doing. 0:16:50.680 --> 0:16:52.720 Yeah, there's two things that I think are critical that 0:16:52.760 --> 0:16:55.480 our research found that where people were using AI and 0:16:55.520 --> 0:16:57.480 it wasn't burning them out and they felt more productive. 0:16:57.920 --> 0:17:00.240 The very first thing is seeking outside help. I mean, 0:17:00.320 --> 0:17:02.240 this is, to your point, a brave new world. And 0:17:02.280 --> 0:17:04.359 there's other people who are much further ahead of this. 0:17:04.880 --> 0:17:07.440 We've seen this in the answer community. They are much 0:17:07.440 --> 0:17:10.560 more likely to be adopting AI incorporate it into their workflows. 0:17:10.600 --> 0:17:12.680 I mean, their livelihood depends on staying on top of 0:17:12.720 --> 0:17:15.119 the skills, and so this is a great community to 0:17:15.160 --> 0:17:17.760 go after to really seek the outside help to both 0:17:17.880 --> 0:17:21.080 educate and understand how you can be incorporating AI. And 0:17:21.080 --> 0:17:22.439 then the second thing, and this is a little bit 0:17:22.480 --> 0:17:24.800 more called to like a business leader perspective, is how 0:17:24.840 --> 0:17:27.640 we actually begin to measure productivity. I worry if we're 0:17:27.800 --> 0:17:31.240 purely looking at efficiency, hours worked, output, we're missing the 0:17:31.240 --> 0:17:34.000 big conversation that we should be having instead. When we 0:17:34.040 --> 0:17:36.920 saw on our research people want to be measured against creativity, innovation, 0:17:37.000 --> 0:17:40.440 how they're adapting to change their contributions to the company strategy. 0:17:40.800 --> 0:17:43.080 And that's the conversation I want to see Jenai play in. 0:17:44.280 --> 0:17:46.840 Yeah, it's interesting, you know, and I do wonder. 0:17:47.960 --> 0:17:49.760 I thought what was interesting also that one in three 0:17:49.760 --> 0:17:52.280 full time employees saying they will likely quit their job 0:17:52.359 --> 0:17:54.080 in the next six months because of the strain that 0:17:54.080 --> 0:17:58.200 they are feeling from rising productivity demand. And I don't 0:17:58.240 --> 0:18:01.400 know what that says maybe more strongly, more significantly about 0:18:01.480 --> 0:18:04.000 kind of the job market, the economic outlook right now. 0:18:04.960 --> 0:18:06.720 If we don't get this right, then what Because I 0:18:06.760 --> 0:18:08.560 just do wonder if it's just I mean, this is 0:18:08.640 --> 0:18:11.919 the you know, typical strain. I feel like when we 0:18:11.960 --> 0:18:14.280