All right. Welcome everybody. Welcome. Okay. I'm Matt Berman. And of course we have here Marc Benioff, co-founder and CEO of Salesforce. Absolute legend in tech.
Lately one of the loudest voices talking about agents, how they're going to transform work, how they're going to transform communication. And I'm so excited to talk to you today, Marc. Thank you for joining.
Well, I'm so happy to be here. Thank you. And we're so happy to have everybody here for this incredible event. We just had a great program downstairs here at the St. Regis and it's been an awesome day.
Yeah, a lot of features launched. Slackbot. We're going to talk a ton about that. A lot of happy customers and partners.
That's right. I am especially excited to talk to you because I think about a lot of these questions all the time and now I get to sit across from you and actually ask the man himself about how he thinks about the future. So let's start out with the AI interface.
You Salesforce acquired Slack a few years ago and it becomes...
Five years ago and it is increasingly obvious that agents should be everywhere that work happens, not trapped in a single interface. And Slack seems like the perfect place to do that. And over time, how do you think about the tension between Slack being the primary interface and then also Slackbot needing to live everywhere?
All credit really goes to Peter Schwarz on our team who's our chief futurist at Salesforce. A lot of people know Peter who wrote or co-wrote or participated in the development of movies like Minority Report, War Games, Deep Impact, and other things. Universally regarded in our industry as being an incredible futurist. Well, he came to us now almost 10 years ago and said the number one thing that we should focus on is acquiring Slack. At the time, no one in the company really agreed with him and no one in the company really understood why that would be. But his number one focus was that the world would go to AI and the world would go to agents and it was his words exactly at that time. What he foresaw correctly was that we were going to get breakthroughs in the models that we were seeing being pioneered in Silicon Valley that we would need an interface to those models, something that is conversational, something that is open, something that has a broad ecosystem. I think what no one could have seen though is not only that Slack has persisted as a great user interface to AI, but also this ecosystem is so rich. The ecosystem of Slack is beyond any of our expectations. It was never kind of foreshadowed that this would be one of the great parts of Slack and it has become that way. And also Slack has become a great interface not only to the AI community, not only to all these companies including the AI companies, but also to Salesforce itself that all of our applications are becoming Slack first.
I mean, let's continue on that thread. So your co-founder Parker hinted at this or kind of spoke about it during the keynote today. Do you see a world in which the traditional Salesforce interface almost fades away a bit and people are really just interacting with their agents, their Slackbot, and then kind of sitting on top of your data layer. Is that the future you see?
There's no question. We've said this to our whole company now several times that we really over time really want to kind of position Lightning in where it gets amplified, where it's strongest. But at the same time, Lightning has a role, but Slack has a role. And Slack as a conversational interface is just a great place for Salesforce users to live. Not just Salesforce users, even Tableau users, other and of course the users of all these ecosystem products as well. In fact, I just had a great demonstration of Writer AI, which is this amazing new AI tool. A lot of folks are using it in the marketing community. It was 100% in Slack. So, that idea that products are coming along to be 100% Slack first. We are completely behind that.
And then are you comfortable with people taking their Slackbot off of Slack wherever else they may be working? Do you see a future in which that's going to happen?
Huge part of it. I think that Slackbot should be kind of a highly composable object that can be dropped into every capability and we're committed to working deeply with Teams from Microsoft or every other collaboration tool like Workspace from Google. But in addition to that, the idea that it can get dropped into every Salesforce app as well. So when you're in SalesCloud, which is our number one namesake product or our service cloud, the Slackbot is with you as well.
Yeah. Okay. I mean that makes a lot of sense. I want to talk about how this transition...
Good. Will you communicate that to my product people as well?
Nervous laughter from the audience.
So I want to actually continue on that thought because all of these changes really change how work happens. They change how companies are structured. How do you think team structures change from here? Do you think we're going to see fewer or sorry more smaller teams? Do you think we're going to have bigger, more ambitious companies? You see like a SEAL team being the best application of the human team inside of a company. What do you think is the right way to think about this?
All of the above plus a huge explosion of agents that are getting coordinated and commanded either through the AI or by humans. As we've been saying for the last two Dreamforces and I think we got this right. Humans and agents are working together. This idea that we have a role and the agents have a role. When you look at the power inside of a company for the agentic layer to become an agentic enterprise is what we say. That idea that all of a sudden now I have my agents who can help me service my customers, sell to my customers, qualify my customers, market to my customers, have conversations with the customers because agents are built on large language models. The key word is language. It's good at speaking language. So things that are basically impacted by language like a service conversation are going to be dramatically better using these tools. Even coding, coding has now become a language. When I was coding it was not a language. We were down at the machine level and we're doing 68,000 assembly language at Apple in 1984 that was not about language that was about a whole different kind of mathematical construct. Now coding is language and so these tools are very good abstracted at that moment.
Yeah. And so I'm facing this at my company. So I want to ask you, agents are getting better. They're producing more work. They're able to almost anticipate what I need. I feel like I'm becoming the bottleneck. I still have to... I certainly am. Right.
I wanted to see if you felt that way and like what is the solution there because we're still having to prompt and we're still having to verify which is incredibly important to human in the loop but...
Over time the human also is the bottleneck and I'm already feeling that. Are you...
It's fine with the human is the bottleneck because these large language models are still wildly inaccurate at times. All of us have had the experience where all of a sudden we're using a large language model and it gets totally confused and it's taken a right turn and it doesn't know what we're talking about. So it's very critical that human beings at least at this stage of AI stay in the loop. Maybe not forever but right now I think it is extremely important and we can see that because the accuracy levels are just not there. So having that ability to really kind of go as far as you can go. A good example is if you go to help.salesforce.com right now you know that you're working with Agentforce to resolve your customer service issue but for about half the calls it hits a moment where that customer is like I need to speak to somebody and the agent goes you're right and boom through our omni channel supervisor goes to a human who then gets a screen in front of them and that screen it could be in Slack it could be in Lightning then is looking at all the conversations all of the information and humans are really good at that moment looking at screen going actually the problem needs to be this. We're just very good at synthesis and that human even won't be 100% accurate but the large language model is still needed a very high percentage of the time.
Is there still a set of tooling to be built to help humans verify at a higher scale? I don't like being the bottleneck and I'm wondering...
Well one thing is yes and then the second thing is new models that'll be more accurate over time as we have new AI baked for this. Remember we are in a continuum of models starting for decades we've had more and more models and then now we're in the large language model stage and we'll move to world models but we're going to move to other kinds of models that we can't even describe yet that will just over time radically increase the accuracy. And our chief scientist Silvio Savarese basically has the vision that we're moving into multi-sensory models. We haven't really seen them yet. But the reason why it's important is the model just has a lot more data to be able to make a decision. Not just language, but all these other kinds of sensory data that we have. Like we have our eyes and ears and our brain and our memories and all these things and we're able to kind of make these decisions. Eventually, the models will get to that point.
Interesting. And so let's continue on the human in the loop. For a long time to rise up in a company as an individual contributor, you needed to specialize. But it seems like we're having a renaissance of the generalist. Are you seeing that inside Salesforce? What do you think about generalist versus specialist in the age of artificial intelligence?
Well, I think this is especially true in software engineering. Obviously, we have 15,000 software engineers at Salesforce. Each one of them is an A player. And the key thing about these 15,000 engineers, they're all out there all over the world. All of them can now be hugely augmented with these coding models. It could be Anthropic Claude. It could be OpenAI Codex. It could be Cursor. It could be others. But when they start to use these models, they're now working not only with the AI but agents to help them code. And they can even become somewhat supervisory over these agents. But still those engineers are needed. The model still cannot operate autonomously. We're not at that level yet of AI. So it's really critical. So our engineering organization is probably more than 30% more productive, but I wouldn't call it 100% more productive. And that's why even in the top AI companies, if you go to their job boards, you'll see they are hiring a lot of engineers. They're hiring a lot of everybody because they have to build companies of humans. Even though these top AI companies, we can go through the list of them again, have these unbelievable models, they need a lot of humans. And that's probably that's the canary in the coal mine that we know that the models are not at that level yet.
So I know you're a big believer in humans will be much more productive. You will still need to hire a lot of humans. We're seeing that in companies like Anthropic, like OpenAI, but there is the flip side where there are other companies and there are divisions of opinions here that they're cutting. And like what do you think about the reason might be for that kind of split opinion and different approach?
Well, Salesforce right now has hit a new record number of employees which is more than 83,000. And but the balance of those employees and where they are placed is different than where it was 5 years ago because we went through an uncomfortable period ourselves over the last 5 years of rebalancing our workforce. That's just a difficult thing for any company. In some cases, these companies are cutting because their costs are just too high. In other cases, these companies are cutting because they've made financial commitments specifically to data centers that they have to pay for. And in other cases, these companies are cutting because they need to rebalance their workforce to reflect the changes in artificial intelligence. These are different reasons. So you cannot bucket all these companies together. If you do, you're making a fundamental mistake, I think, in business. And even though I've spoken about this, I think somewhat aggressively, I don't think most people still really understand what is going on. And it's too easy to kind of basically take AI and make it the scapegoat. And I think for some CEOs, it's the lazy way out. That's up to them. I think it's better just to say, here's what's really going on and trying to be specific as you can. You're going to take bullets no matter what because that's your role as CEO. And then you have to kind of go forward and put everything back together.
Yeah. I mean the CEOs that I'm talking to they're doubling down on the teams that are leveraging artificial intelligence. They are investing more in them. They're hiring more. As you if you were to give advice to somebody fresh out of college, you said that maybe you are rebalancing or had rebalanced inside Salesforce. What organizations should people skill up to?
This is a very very good question and I am very worried about this actually because I was just at several universities just last week including MIT in Boston and I sat with an incredible woman who's a junior at MIT, top computer scientist in her class and she's asking me if she should change majors and I walk through with her the realities of AI and the models just to remind her of her own technology strengths of where the technology actually is and the reality is I'm trying to recruit her, try to bring her, but all of her colleagues should come to Salesforce. We're hiring interns. We're hiring that freshman class. We want to send our recruiters out to those top 25 high academic threshold universities like MIT and bring those top computer scientists and others to Salesforce because we badly need that talent. And that idea that those people are so critical for a company like ours, that is really part and parcel for what I think the discussion is because for a lot of them, they've been told, well, there might be a difficulty getting a job for the summer or even getting a job when you graduate. But the reality is that companies like ours badly need these people. And we need to kind of put two and two together to really show that this is still going to be a critical part of our workforce and a lot of workforces going forward.
