WEBVTT - Marc Benioff’s Race to One Billion Agents 0:00:00.880 --> 0:00:03.360 I'm Emily Chek and this is the podcast version of 0:00:03.400 --> 0:00:07.480 the Circuit my show on Bloomberg Originals. There's one building 0:00:07.720 --> 0:00:11.160 that rises above the rest of the San Francisco skyline. 0:00:11.320 --> 0:00:14.720 It's home to a software giant whose story begins somewhere 0:00:14.960 --> 0:00:16.040 more tropical. 0:00:16.480 --> 0:00:17.160 I heard something. 0:00:17.200 --> 0:00:19.720 Did you really come up with the idea for Salesforce 0:00:19.760 --> 0:00:21.200 while you were swimming with dolphins? And? 0:00:21.239 --> 0:00:24.400 Fright? Did I did I do a great dolphin imitation? 0:00:24.480 --> 0:00:25.560 Did you know that? Can you? 0:00:27.960 --> 0:00:28.880 How do they respond? 0:00:31.440 --> 0:00:34.800 This is Mark Benioff, CEO and founder of Salesforce, the 0:00:34.840 --> 0:00:39.839 world's number one customer relationship management or CRM software provider. 0:00:40.159 --> 0:00:44.120 They sell cloud based business tools for sales, marketing, e commerce, 0:00:44.280 --> 0:00:47.360 all that good stuff. It might sound niche, but their 0:00:47.360 --> 0:00:51.360 clients include the likes of Apple, Boeing, Amazon, and more 0:00:51.400 --> 0:00:55.000 than one hundred and fifty thousand companies. Benioff has been 0:00:55.040 --> 0:00:58.960 called a master networker, a force of nature, mister disruptor, 0:00:59.200 --> 0:01:03.320 and the fresh of software. Let's just say subtlety isn't 0:01:03.360 --> 0:01:06.440 really his thing. At its core, this is a story 0:01:06.480 --> 0:01:10.280 all about software. A self taught teenage programmer turned visionary 0:01:10.360 --> 0:01:13.240 tech billionaire, and believe it or not, it all starts 0:01:13.240 --> 0:01:16.400 with a customer service job. In nineteen eighty six, you 0:01:16.440 --> 0:01:19.280 could find Benioff answering customer calls at a not so 0:01:19.440 --> 0:01:23.600 little future competitor called Oracle, under the wing of longtime 0:01:23.640 --> 0:01:27.039 CEO Larry Ellison. Benioff's snack for sales took him straight 0:01:27.080 --> 0:01:30.120 to the top. Today, Benioff is all in on the 0:01:30.200 --> 0:01:33.960 AI boom with a new product, AI Agents, a digital 0:01:34.000 --> 0:01:37.880 workforce to autonomously complete tasks. He isn't the only one 0:01:37.959 --> 0:01:41.560 betting big on agentic AI, but true to form, he's 0:01:41.600 --> 0:01:42.000 all in. 0:01:43.800 --> 0:01:44.360 Good to fore you. 0:01:45.040 --> 0:01:47.320 Thanks for having us here. We are back in the 0:01:47.360 --> 0:01:48.000 Ohanna room. 0:01:48.720 --> 0:01:50.840 Welcome back, Thank you, Ohanna. 0:01:51.000 --> 0:01:53.040 I'm so happy to have you. I know you know 0:01:53.080 --> 0:01:56.120 what jwhana means great? How do you know that it means? 0:01:56.320 --> 0:01:57.640 How do you know what ohana means? 0:01:57.680 --> 0:01:58.440 I'm from Hawaii. 0:01:58.480 --> 0:02:01.480 You're Hawaii? Are you a native Hawaiian? 0:02:01.480 --> 0:02:02.920 I'm a Kamma. 0:02:03.400 --> 0:02:04.360 Were you born in Hawaiian? 0:02:04.360 --> 0:02:06.080 I was born in Hawaii, born and raised. And you 0:02:06.120 --> 0:02:06.680 love Hawaii? 0:02:06.800 --> 0:02:07.640 I love Hawaii too. 0:02:07.720 --> 0:02:09.240 But next time, if we do this, I want my 0:02:09.280 --> 0:02:10.120 ties on the Big Island. 0:02:10.160 --> 0:02:10.680 Let's do it. 0:02:10.840 --> 0:02:11.080 Deal. 0:02:11.240 --> 0:02:11.679 Come over. 0:02:12.720 --> 0:02:16.120 I've interviewed you many times from your false cloud days, 0:02:16.360 --> 0:02:18.720 many many dream forces. But I want to go back 0:02:18.760 --> 0:02:20.959 a bit to young Mark growing up in San Francisco. 0:02:21.080 --> 0:02:23.959 Yes, Young Mark, right. 0:02:23.240 --> 0:02:26.440 So, your parents ran department stores. You grew up your 0:02:26.440 --> 0:02:27.760 first job was cleaning. 0:02:27.560 --> 0:02:29.560 Jewelry Beniof's department store. 0:02:29.320 --> 0:02:33.880 Benioff's department stores. Who was Mark Benioff before Mark Benioff? 0:02:35.040 --> 0:02:38.160 Do I even exist? Is this an existential thing? Should 0:02:38.200 --> 0:02:41.480 I bring in some spiritual. 0:02:40.639 --> 0:02:42.560 But seriously, tell me about you as a kid. 0:02:43.160 --> 0:02:46.919 I was born right there, yeah, literally about six blocks 0:02:46.960 --> 0:02:52.560 away at Mount Zion Hospital, UCSF, now right onto Visadero Street, 0:02:53.280 --> 0:02:57.639 nineteen sixty four, sixty years ago, and I grew up 0:02:57.840 --> 0:03:01.919 right there for two years. Then I moved right there 0:03:02.120 --> 0:03:05.200 for a while, and then I went to college right 0:03:05.280 --> 0:03:08.200 down there. So this is your hood and this is 0:03:08.240 --> 0:03:10.239 where I, Yeah, I feel very comfortable here. 0:03:10.800 --> 0:03:14.800 You basically invented the SaaS industry software as a service 0:03:14.960 --> 0:03:18.320 and made it possible for everything from Netflix to Spotify 0:03:18.400 --> 0:03:20.240 to businesses at Google and Microsoft. 0:03:20.400 --> 0:03:22.560 What sparked the entrepreneur in you? 0:03:23.160 --> 0:03:26.440 I don't know what sparked the entrepreneur because certainly my 0:03:26.639 --> 0:03:30.800 family was entrepreneurial in the Benioff Department stores. On the 0:03:30.800 --> 0:03:34.480 other side of my family. My grandfather was a supervisor 0:03:34.520 --> 0:03:38.480 and worked right there in City Hall, and he also 0:03:38.800 --> 0:03:41.880 was an attorney at six ninety Market Street right there, 0:03:42.440 --> 0:03:45.160 and we used to walk right back and forth there, 0:03:45.440 --> 0:03:48.320 and he was doing civic leadership, and my father was 0:03:48.400 --> 0:03:52.400 a retailer selling clothing, so it was two different worlds. 0:03:53.160 --> 0:03:56.520 And then what happened was I was just fifteen years old. 0:03:57.000 --> 0:04:00.680 I was cleaning jewelry cases not so far away from here, 0:04:00.680 --> 0:04:04.040 in a jewelry store, doing a lot of very hard work, 0:04:04.360 --> 0:04:08.400 like cleaning the floors. I remembered really well polishing the cases. 0:04:08.880 --> 0:04:11.640 And I walked across the street to a radio shack store, 0:04:11.720 --> 0:04:14.480 which doesn't really exist anymore, and I found a computer, 0:04:15.280 --> 0:04:18.520 and in that computer I became fascinated as a fifteen 0:04:18.560 --> 0:04:22.080 year old, and I taught myself how to program, and 0:04:22.160 --> 0:04:25.839 from then I bought other computers kind of based on 0:04:25.880 --> 0:04:29.800 this income from this jewelry store. And my grandmother would 0:04:29.800 --> 0:04:33.159 say to me, listen, I'll match you, and which was 0:04:33.279 --> 0:04:36.240 very exciting. Yeah, the computer was four nine oh of dollars, 0:04:36.640 --> 0:04:38.800 so I made two hundred and fifty dollars. She matched 0:04:38.800 --> 0:04:42.160 two hundred and fifty dollars I bought the computer. I 0:04:42.200 --> 0:04:44.839 sold my first piece of software when I was about fifteen, 0:04:44.960 --> 0:04:48.320 called how to Juggle, to a little software reseller in 0:04:48.400 --> 0:04:54.760 Santa Barbara, and then I leveraged that into an Apple 0:04:54.800 --> 0:04:57.560 Too computer. I didn't like that. I sold that to 0:04:57.640 --> 0:04:59.920 a friend of mine who's now our chief technology off 0:05:00.000 --> 0:05:03.400 are here, who I've worked with for forty five years, 0:05:04.000 --> 0:05:05.720 and that was that. 0:05:06.080 --> 0:05:09.039 The rest is history. You founded the company now twenty 0:05:09.040 --> 0:05:09.960 five years ago. 0:05:10.200 --> 0:05:12.919 So you've seen the rise of mobile and social and 0:05:12.960 --> 0:05:14.320 the cloud and now AI. 0:05:14.640 --> 0:05:16.440 How has the CEO job changed for you? 0:05:17.320 --> 0:05:21.599 Well, the CEO chop is really changing fast because it 0:05:21.680 --> 0:05:23.760 used to be I felt very alone at the top. 0:05:24.480 --> 0:05:27.240 But now like I just finished writing the business plan 0:05:27.320 --> 0:05:29.800 for this year, and I always do that with someone else. 0:05:30.200 --> 0:05:33.239 I take one of our executives, And for the last 0:05:33.279 --> 0:05:37.039 three years, I've also found a new partner in AI, 0:05:37.400 --> 0:05:40.640 and I bring AI in and I have an AI partner. 0:05:40.720 --> 0:05:43.120 So I have an AI partner, I have a human partner, 0:05:43.480 --> 0:05:45.200 and it's the three of us who are writing the 0:05:45.240 --> 0:05:47.279 plan together, so it's a little less lonely at the top. 0:05:47.400 --> 0:05:50.000 You said you'll have a billion agents on salesforce before 0:05:50.000 --> 0:05:50.599 the end of the year. 0:05:50.680 --> 0:05:53.280 Yes, are you on track? What are they doing? I mean, 0:05:53.320 --> 0:05:54.040 will they be doing? 0:05:54.240 --> 0:05:57.440 Remarkably on track because we now have about five thousand 0:05:57.520 --> 0:06:01.720 customers deploying this and it's amazing how many agents we 0:06:01.720 --> 0:06:05.599 see getting spawned every year. We're tracking very closely how 0:06:05.640 --> 0:06:09.359 these conversations are growing, so week by week we're looking 0:06:09.400 --> 0:06:12.400 at how are the conversations growing, and it's amazing. So 0:06:12.560 --> 0:06:15.120 I think it's going to be incredible. I mean, by far, 0:06:15.160 --> 0:06:17.799 it's the fastest growing, most exciting thing we've ever done. 0:06:18.040 --> 0:06:19.880 You've thrown around this term AI nirvana. 0:06:20.800 --> 0:06:23.640 What is that? When do we get there? And how 0:06:23.640 --> 0:06:24.640 do we know? What won't feel like? 0:06:24.720 --> 0:06:24.840 Hell? 0:06:25.960 --> 0:06:28.839 Well, I mean, nobody exactly knows what's going to happen 0:06:29.040 --> 0:06:31.719 when we get to this next level of AI. I 0:06:31.760 --> 0:06:34.560 can tell you in the current level of AI, between 0:06:34.560 --> 0:06:37.680 this combination of large language models and these new reasoning 0:06:37.720 --> 0:06:40.520 models and where we are, we're able to provide this 0:06:40.600 --> 0:06:44.400 kind of digital labor revolution. It's an opportunity for our 0:06:44.400 --> 0:06:49.520 customers to improve productivity without hiring more employees, and to 0:06:49.640 --> 0:06:54.320 let them redeploy existing employees into functions that need more attention, 0:06:54.440 --> 0:06:57.520 more growth. I'm doing it myself. It's amazing, and that's 0:06:57.560 --> 0:07:00.400 really important because it's hard to hire right now, isn't 0:07:00.400 --> 0:07:04.599 as many people available. Unemployment is remarkably low. Interesting, so 0:07:05.040 --> 0:07:07.720 you know, we have a point where we're going to 0:07:07.760 --> 0:07:11.600 be able to add in digital labor in addition to 0:07:11.640 --> 0:07:15.160 physical labor, so productivity itself will improve. 0:07:15.680 --> 0:07:18.400 You bought the name Einstein, he's the face of your AI, 0:07:18.960 --> 0:07:20.600 but how smart are these agents really? 0:07:21.280 --> 0:07:23.960 Ten years ago we had the vision that AI would 0:07:24.000 --> 0:07:28.200 surround us, and we bought that Einstein name, or really 0:07:28.240 --> 0:07:30.320 licensed it. You know, you can't really buy his name, 0:07:30.360 --> 0:07:33.600 but we licensed it for artificial intelligence because we really 0:07:33.640 --> 0:07:37.120 believed that this is where we were going to end up, 0:07:37.120 --> 0:07:39.360 in this incredible place, which is why this week we'll 0:07:39.360 --> 0:07:42.480 probably do about a trillion Einstein transactions for all of 0:07:42.520 --> 0:07:45.080 our customers around the world just this week. Yeah, and 0:07:45.120 --> 0:07:48.640 now it's evolving into agent force, this idea that it's 0:07:48.680 --> 0:07:53.080 these agents, these digital laborers, digital employees who are out 0:07:53.080 --> 0:07:56.880 there doing this work, servicing the customer, selling the customer, 0:07:57.320 --> 0:08:01.680 marketing to the customer, partnering with me to help run 0:08:01.720 --> 0:08:05.280 the company, partnering with me to do the analytics partnering 0:08:05.280 --> 0:08:08.520 with me to help do the marketing, the branding. So 0:08:08.680 --> 0:08:09.800 that's exciting. 0:08:10.000 --> 0:08:12.120 So what do you think Einstein would think? Would Einstein 0:08:12.560 --> 0:08:13.080 be proud? 0:08:13.600 --> 0:08:16.640 I think Einstein would be using AI in his work, 0:08:17.040 --> 0:08:20.040 There's no question. You know, he could have his insights, 0:08:20.080 --> 0:08:23.520 his capabilities and then probably would validate them using AI. 0:08:23.720 --> 0:08:27.160 Yesterday I was with a Nobel Prize laureate who has 0:08:27.560 --> 0:08:31.280 received this incredible Nobel Prize for his work in proteomics, 0:08:31.360 --> 0:08:34.840 and then he's envisioning now how he's designing new drugs 0:08:34.960 --> 0:08:37.720 using AI. I mean that is a leap that we 0:08:37.840 --> 0:08:40.520 didn't think any of us a lot that we would 0:08:40.520 --> 0:08:42.640 get there as quickly as we are, and now we 0:08:42.679 --> 0:08:45.559 can see that scientists have all types. So yes, Einstein 0:08:46.160 --> 0:08:49.920 or whoever is going to have the ability to go faster, 0:08:50.360 --> 0:08:53.959 be more productive. Because of this capability, you're just able 0:08:53.960 --> 0:08:54.640 to do more. 0:08:55.120 --> 0:08:56.280 How secure is all this? 0:08:56.480 --> 0:08:59.800 I mean there's malicious code, there's hallucinations. 0:09:00.120 --> 0:09:01.560 What are the risks for businesses? 0:09:01.800 --> 0:09:04.080 Well, those are two different things. I think security and 0:09:04.120 --> 0:09:08.680 hallucinations are just slightly different. So security, there's no finish line. 0:09:09.280 --> 0:09:13.040 We're in a world of cryptography and changes to cryptography 0:09:13.120 --> 0:09:16.079 and the security issues, and you can never have extreme 0:09:16.120 --> 0:09:19.839 confidence on security, you have to be completely paranoid. Hallucinations 0:09:20.160 --> 0:09:23.720 is a different story. That is because of the nature 0:09:23.760 --> 0:09:25.920 of the models that we're using right now. They're called 0:09:26.000 --> 0:09:30.000 large language models, combined with these reasoning models, these two 0:09:30.040 --> 0:09:32.240 things together and kind of the current state of the 0:09:32.360 --> 0:09:35.559 art of AI is not one hundred percent accurate. And 0:09:35.600 --> 0:09:37.640 you know that because anyone who's used any of these things, 0:09:37.679 --> 0:09:40.440 so wait a minute, that's not well, like everybody probably 0:09:40.440 --> 0:09:42.400 puts their name in and goes, wait, that's not me, 0:09:42.520 --> 0:09:44.360 or that's not my exact bio. 0:09:44.720 --> 0:09:45.760 Doesn't it get messier. 0:09:46.200 --> 0:09:48.959 It could get a lot more accurate. It could get 0:09:49.000 --> 0:09:54.319 more accurate because the models themselves are stacking their intelligence, 0:09:55.240 --> 0:09:59.360 so hopefully they're honing what the truth is, and that's 0:09:59.400 --> 0:10:01.840 the hope. Right now, it's not one hundred percent accurate. 0:10:01.880 --> 0:10:05.560 In our own work. We're doing hundreds of thousands of 0:10:05.600 --> 0:10:09.800 conversations just ourselves and our company with our customers, and 0:10:09.840 --> 0:10:12.440 we have about ninety three percent accuracy. Even for a 0:10:12.559 --> 0:10:14.640 large brand or a large company that we work with, 0:10:14.679 --> 0:10:17.400 like Disney, it's about ninety three percent. It's not one 0:10:17.480 --> 0:10:18.000 hundred percent. 0:10:18.320 --> 0:10:20.920 Ninety three percent though, that's not good enough, is it. 0:10:20.920 --> 0:10:23.720 It's pretty good. You're able to do a lot, but 0:10:23.800 --> 0:10:26.520 you have to be realistic that we're not at one hundred. 0:10:26.520 --> 0:10:28.640 You'll never see me saying that we're at one hundred percent. 0:10:28.880 --> 0:10:30.720 But I think for a lot of the other vendors 0:10:31.400 --> 0:10:34.160 maybe they're at much lower levels because they don't have 0:10:34.200 --> 0:10:37.240 as much data and metadata and they're not able to 0:10:37.320 --> 0:10:39.640 kind of put as much information into the model to 0:10:39.679 --> 0:10:41.199 get to a higher level of accuracy. 0:10:41.280 --> 0:10:43.960 You started out as a coder at Apple in nineteen 0:10:44.000 --> 0:10:46.320 eighty four. Thank you were an intern in the Mac division. 0:10:46.840 --> 0:10:48.160 Well, hey, you've done your homework. 0:10:48.200 --> 0:10:50.680 I have done my homework here in twenty twenty five, 0:10:50.679 --> 0:10:54.239 though you've said you won't hire any more coders, Well. 0:10:54.040 --> 0:10:56.800 You're right. When I was fifteen, I was writing that 0:10:56.920 --> 0:10:59.880 software on that Tierrasad. Then it became an Apple too, 0:11:00.040 --> 0:11:02.120 and become an Atarian Hinder, Then it became a Commodore 0:11:02.200 --> 0:11:04.920 sixty four, then it turned into a Macintosh, and I 0:11:04.960 --> 0:11:08.600 wrote like ten, twelve, fifteen different pieces of software that 0:11:08.640 --> 0:11:11.440 I ended up selling commercially. And I was just a kid. 0:11:11.520 --> 0:11:13.040 I didn't know what I was doing, and I was 0:11:13.080 --> 0:11:18.320 self trained. I didn't really have any formal computer science education. 0:11:19.000 --> 0:11:21.480 It was cool to work at Apple in nineteen eighty four. 0:11:21.520 --> 0:11:24.280 I was a sophomore in college, such an iconic time. 0:11:24.640 --> 0:11:27.400 Steve Jobs was walking around, but all of the original 0:11:27.440 --> 0:11:29.600 players and the Mac team were there. I mean, it 0:11:29.800 --> 0:11:33.560 was an imprint in my subconscious of what that cool 0:11:33.800 --> 0:11:37.880 company was and who that cool leader was. And I'll 0:11:37.920 --> 0:11:41.560 never forget that. I am so deeply grateful that I 0:11:41.679 --> 0:11:44.520 had that experience. I'm deeply grateful to Steve Jobs. He 0:11:44.600 --> 0:11:47.040 had a huge impact on my life. He had a 0:11:47.080 --> 0:11:49.640 huge impact on Salesforce. When I started Salesforce, he would 0:11:49.640 --> 0:11:52.680 call me all the time and give me advice. And 0:11:52.760 --> 0:11:54.680 I actually wish I had taken more of his advice. 0:11:54.720 --> 0:11:56.040 I mean a lot of the things he was saying 0:11:56.080 --> 0:11:57.800 I did not understand, but he was right. 0:11:58.040 --> 0:12:01.520 So you said you won't hire anyway MOK coders at Salesforce, 0:12:01.880 --> 0:12:05.040 and you've said today's CEOs will be the last to 0:12:05.120 --> 0:12:07.360 manage all human workforces. 0:12:07.640 --> 0:12:09.640 What does this mean for businesses? 0:12:09.960 --> 0:12:11.560 Well, I just had a meeting with my head of 0:12:11.600 --> 0:12:15.400 engineering and we're looking at productivity levels thirty to fifty 0:12:15.440 --> 0:12:20.920 percent this year, and key functions like engineering, coding, support. 0:12:20.640 --> 0:12:23.440 You're saying service AI is doing thirty to fifty percent 0:12:23.480 --> 0:12:24.000 of the work. 0:12:24.400 --> 0:12:26.640 AI is doing thirty to fifty percent of the work 0:12:26.640 --> 0:12:30.760 at Salesforce now, and I think that that will continue. 0:12:30.800 --> 0:12:32.200 I think that all of us have to get our 0:12:32.240 --> 0:12:35.760 head around this idea that AI can do things that 0:12:35.880 --> 0:12:38.200 before we were doing and we can move on to 0:12:38.240 --> 0:12:39.319 do higher value work. 0:12:39.440 --> 0:12:42.640 So Salesforce is seventy five thousand employees now, is it 0:12:42.679 --> 0:12:44.120 half that in the future. 0:12:44.840 --> 0:12:47.480 I'm not willing to make a projection exactly like that. 0:12:47.679 --> 0:12:51.120 I do think probably will rebalance. There's no question that 0:12:51.480 --> 0:12:54.559 we have this opportunity to take advantage of the technology 0:12:54.600 --> 0:12:56.440 to get to a new place, and I think every 0:12:56.440 --> 0:12:58.520 company is going to be able to do that. Look 0:12:58.520 --> 0:13:01.640 at our own support. Just four months in we have 0:13:01.679 --> 0:13:04.199 help dot Salesforce dot com, which is our agent force 0:13:04.240 --> 0:13:07.520 doing our support. We have nine thousand human support engineers. 0:13:07.559 --> 0:13:10.480 They're still involved, they're still doing their work, but they're 0:13:10.480 --> 0:13:12.679 just starting to do different kinds of work. Because we've 0:13:12.679 --> 0:13:15.880 already done more than five hundred thousand conversations soon a 0:13:15.920 --> 0:13:19.800 million through agent force. That's just for us, and then 0:13:19.840 --> 0:13:23.160 when you extrapolate that with all the agents, you know, 0:13:23.200 --> 0:13:27.360 the billion agent dream. That's a big idea that all 0:13:27.400 --> 0:13:31.200 these companies can do all these incredible things. Like I 0:13:31.240 --> 0:13:36.439 was saying this idea of Disney or Lenar, the nation's 0:13:36.520 --> 0:13:38.720 largest home builder, one of them, They're able to do 0:13:38.760 --> 0:13:42.960 all their sales and service and support with agents. Or 0:13:43.200 --> 0:13:46.680 you take another exciting customer, there's like so many of 0:13:46.679 --> 0:13:48.840 them open table Like I'm going to make am going 0:13:48.880 --> 0:13:51.120 to go to an open table and make a reservation 0:13:51.200 --> 0:13:53.520 for Saturday night. I do all the time. And this 0:13:53.679 --> 0:13:56.560 week we've been working with their employees now for six 0:13:56.600 --> 0:13:59.920 months with agent Force now it runs on the concer 0:14:00.360 --> 0:14:04.160 side as well. So that is just very exciting that 0:14:04.200 --> 0:14:07.440 we're able to impact our customers in a positive way, 0:14:07.520 --> 0:14:10.960 making them more productive and helping to change that balance 0:14:10.960 --> 0:14:11.240 of work. 0:14:11.360 --> 0:14:14.720 So Salesforce is marketing it's AI tools on their ability 0:14:14.760 --> 0:14:17.760 to replace human labor. Do you have any ethical qualms 0:14:17.800 --> 0:14:18.200 about that? 0:14:18.679 --> 0:14:22.520 Well, it's a digital labor revolution. That is, we're probably 0:14:22.560 --> 0:14:25.760 looking at three to twelve trillion dollars of digital labor 0:14:25.840 --> 0:14:28.440 getting deployed, and that digital labor is going to be 0:14:28.440 --> 0:14:31.840 everything from AI agents to robots. We've seen all the 0:14:31.880 --> 0:14:33.760 kind of robots that are coming We've seen the movies 0:14:33.800 --> 0:14:36.600 for a long time. Right now we're seeing it deployed. 0:14:37.120 --> 0:14:41.040 And I think it's really just technologies marching forward. It's 0:14:41.040 --> 0:14:44.280 getting lower costs, it's getting easier to use. And I 0:14:44.320 --> 0:14:47.280 do think to your point, CEOs have to make sure 0:14:47.320 --> 0:14:49.520 their values are in the right place and that values 0:14:49.560 --> 0:14:52.119 bring value. But we're becoming more automated. 0:14:52.640 --> 0:14:54.240 Could an agent replace you one day? 0:14:54.640 --> 0:14:57.480 I hope so you could already say to my point, 0:14:57.520 --> 0:14:59.920 which is that some of my job is already be 0:15:00.160 --> 0:15:02.760 done by AI. I mean, I'll be honest with you. 0:15:02.960 --> 0:15:06.560 I'll write my plan, my business plan, and then I'll 0:15:06.600 --> 0:15:08.440 say to the agent. Now, I mean, I've been doing 0:15:08.440 --> 0:15:11.280 this now for almost three years. Hey, great it What 0:15:11.280 --> 0:15:13.480 do you think of my plan against my competitor. I 0:15:13.480 --> 0:15:16.560 won't tell you which competitor, but against this competitor, which 0:15:16.840 --> 0:15:17.960 you know, how do you think I'm doing it? And 0:15:17.960 --> 0:15:20.960 I'll say, oh, Mark, I think you know it's about 0:15:20.960 --> 0:15:23.240 a B plus And I'll say, well, what does it 0:15:23.280 --> 0:15:25.200 need to do to get to an A? And I 0:15:25.200 --> 0:15:27.200 think that idea of knowing how to talk to the 0:15:27.240 --> 0:15:30.240 AI and communicate with the AI. And I've been training 0:15:30.280 --> 0:15:32.040 a lot of my friends on how to do that. 0:15:32.080 --> 0:15:34.360 I just had a friend of mine who left the FDA, 0:15:34.520 --> 0:15:37.280 you know, took one of these Trump leaves, and he's 0:15:37.320 --> 0:15:39.360 thinking about what does he do next? How does he 0:15:39.400 --> 0:15:42.200 think about it? And he hadn't really worked with an AI, 0:15:42.960 --> 0:15:45.040 and so I'm working with the I and I'm saying 0:15:45.040 --> 0:15:48.520 to him, Hey, ask the AI these questions, you know, hey, 0:15:48.520 --> 0:15:51.440 what about this? And he's using very technical terms that 0:15:51.520 --> 0:15:53.960 I don't even know what The technical terms are based 0:15:54.000 --> 0:15:58.520 on certain biological substances that he understands and chemical substances. 0:15:58.720 --> 0:16:01.520 But the AI understood a lot of what he's talking about. 0:16:02.120 --> 0:16:04.680 And then I watched him interact with the AI for 0:16:04.720 --> 0:16:07.080 about fifteen minutes, and together they came up with a 0:16:07.080 --> 0:16:10.480 company idea named it brand It built a slide deck. 0:16:10.960 --> 0:16:14.200 And he walked away kind of shaking his head, going, yeah, 0:16:14.240 --> 0:16:16.080 that is a new potential future for me. 0:16:16.280 --> 0:16:18.920 So why do you hope an agent replaces you one day? 0:16:19.880 --> 0:16:23.480 Well, I think that this idea that AI can do 0:16:23.520 --> 0:16:25.720 a really good job and a lot of these functions. 0:16:25.760 --> 0:16:27.920 So I mean, of course, I'm partially kidding you know that, 0:16:28.360 --> 0:16:31.000 But I'm like, I do think, Look, how long will 0:16:31.000 --> 0:16:34.720 it be where there's an AI also interviewing me with you. 0:16:34.720 --> 0:16:37.080 You know that is we could be talking. It knows 0:16:37.120 --> 0:16:39.880 my whole background and knows all of my interviews, and 0:16:39.920 --> 0:16:41.680 then you know, I want. 0:16:41.560 --> 0:16:43.880 The real thing that agent I don't want AI. 0:16:43.720 --> 0:16:48.080 Mark, Agent Emily. Okay, and Emily both are interviewing me, 0:16:48.800 --> 0:16:51.400 and you could turn to Agent Emily and say, Agent Emily, 0:16:51.760 --> 0:16:54.200 what question did I not ask Mark? Like you've probably 0:16:54.240 --> 0:16:56.480 asked me a thousand questions, and then say, hey, what 0:16:56.640 --> 0:16:58.480 question should I ask him? Now? Based on all of 0:16:58.480 --> 0:17:01.040 the history that I've had with Mark, it might have 0:17:01.080 --> 0:17:03.920 a really unique answer. I was at a conference yesterday 0:17:04.200 --> 0:17:06.920 and there was a political leader there and we were 0:17:06.960 --> 0:17:10.520 having this dialogue with him, and at the end of 0:17:10.560 --> 0:17:13.320 the conference, they said, hey, we've had AI listening to 0:17:13.359 --> 0:17:16.360 this conference and just listen to all of your questions, 0:17:16.359 --> 0:17:18.159 and we are going to have now AI ask you 0:17:18.240 --> 0:17:21.440 the last question. It was able to sum all of 0:17:21.480 --> 0:17:25.080 the questions that we had already asked, plus that person's 0:17:25.119 --> 0:17:28.160 whole individual history, and then ask that final question. 0:17:28.560 --> 0:17:29.840 Sam Altman's you open. 0:17:29.880 --> 0:17:32.480 AI recently told me it's going to get really weird 0:17:32.480 --> 0:17:34.560 when we're walking down the street with a pack of 0:17:34.720 --> 0:17:38.440 humanoid robots how do you see robotics and AI agents 0:17:38.520 --> 0:17:39.240 coming together? 0:17:40.560 --> 0:17:43.520 First of all, I'm just trying to like help this 0:17:44.240 --> 0:17:47.719 come along right now in the present moment. And I 0:17:47.840 --> 0:17:50.120 think that it's not a huge leap to think that 0:17:50.119 --> 0:17:51.920 that's going to be a weird idea. I mean, we've 0:17:51.960 --> 0:17:54.880 seen the movies, so we can kind of envision some 0:17:54.920 --> 0:17:56.680 of this is going to happen, and we can see 0:17:56.720 --> 0:18:00.159 where the robots are and where the software is. But 0:18:00.200 --> 0:18:02.520 no robot is walking home with you today. But you 0:18:02.600 --> 0:18:05.199 might get in a weaimo, and maybe a weaymo is 0:18:05.240 --> 0:18:07.600 not that far away from a robot taking new home. 0:18:07.720 --> 0:18:08.280 Is that weird? 0:18:09.560 --> 0:18:12.119 Maybe maybe not? I could get used to it. 0:18:12.160 --> 0:18:12.440 I yeess. 0:18:12.480 --> 0:18:14.400 I mean it's the car that's just driving itself around 0:18:14.440 --> 0:18:16.359 San Francisco. I think a lot of people haven't driven 0:18:16.400 --> 0:18:18.680 in an autonomous vehicle. And when they come here, they're 0:18:18.680 --> 0:18:20.480 hanging out with me. They're staying with me in my house. 0:18:20.520 --> 0:18:22.359 I'm like, I'm going to call a weimo for you, 0:18:22.400 --> 0:18:23.560 and a lot of people are like, I don't know, 0:18:23.800 --> 0:18:27.000 can I get in there? And I put them in 0:18:27.000 --> 0:18:29.560 and whisk them away and said okay, bye bye. It's 0:18:29.600 --> 0:18:31.679 that experience. It's a robot on wheels. 0:18:32.080 --> 0:18:33.240 So I don't know if you know this, but our 0:18:33.320 --> 0:18:36.959 last interview at Dreamforce, someone spoofed it, and Salesforce has 0:18:37.040 --> 0:18:39.480 kind of become a meme. There's these jokes out there. 0:18:39.520 --> 0:18:41.639 People who use it say it can be confusing. They 0:18:41.720 --> 0:18:43.399 joke that some people don't even know what it is. 0:18:43.960 --> 0:18:45.959 How do you feel about your company being both a 0:18:46.000 --> 0:18:47.879 lifeline and a punchline. 0:18:48.000 --> 0:18:51.080 I love it, And also like even Elon Musk texted 0:18:51.119 --> 0:18:55.960 me and said, hey, I don't understand what does Salesforce do? 0:18:56.040 --> 0:18:59.000 And I'm like, what is this the meme? And he's like, yeah, 0:18:59.040 --> 0:19:01.119 you also have slash, Yeah we also have Slack, you 0:19:01.119 --> 0:19:03.840 have Tableau, Yeah we have Tableau. You also have MuleSoft. 0:19:04.040 --> 0:19:06.520 Yeah we also have all these other products. It's like, 0:19:06.600 --> 0:19:09.080 why do you call this company Salesforce? I'm like, what 0:19:09.119 --> 0:19:13.040 should we call it? Agent Force? So yeah, that's all right. 0:19:13.080 --> 0:19:16.320 You've been throwing some shade at Microsoft lately. You've said 0:19:16.400 --> 0:19:20.240 Copilot is the new Clippy? What's the beef? What did 0:19:20.240 --> 0:19:21.199