Ethical AI with The Wall Street Journal & David Schmaier at Davos 2024 | Salesforce

I'm Josh Stinchcomb, chief Revenue Officer at the Wall Street Journal, and really excited to kick things off with David Schmeier, the President and Chief Product Officer of Salesforce.

Thanks for being here, Josh. Thanks for having us. And it's great to see everybody here, bright and early, caffeine included.

Yeah, exactly. We could spend our whole time Talking about your 30 years in the industry and the impact you've had, but we'll take that as a given and just jump right into the topic of AI, because there's a lot to unpack.

Let's do that.

Yeah. And I'll start simple. Or maybe this is not simple, but how would you describe. How would you describe AI and what it's able to do today?

Absolutely. So AI has been around since the dawn of computing in the 50s, and it really took off when we went from expert Systems to predictive AI. And if you've ever bought something on Amazon.com and it suggests what you should buy next, that's predictive AI. And Salesforce was the first company to really go long and to include AI in our software. So we branded our AI. We call our brand Einstein, which is our trusted AI brand. And we now do 1.4 trillion predictions every month. And so it's. Every one of our customers all around the world use AI, and so that's predictive AI. But the whole world changed when generative AI came out. And generative AI is all spawned from the 2017 paper by Google, Google Brain, and they came out with this new architecture that. That essentially allows you to predict not the future things that people might want to buy or the predictions for other kinds of discipline. But it predicts the next word or the next sentence or the next paragraph, or it can write prose, or you can even apply that same algorithm to video or to imagery or what they call multimodal AI. So we think generative AI is a big, big deal. I personally believe this is bigger than the Internet, and. And the world will never be the same.

Yeah. Well, I was going to ask you. Three decades in Silicon Valley, you've seen extraordinary hits. You've seen unfulfilled hype. This is real. This is big.

This is definitely big. This is a big deal. It took 75 years for 100 million people to use the telephone. It took two years for 100 million people to use the Apple App store. It took two months for 100 million people to use ChatGPT. This is the big thing.

So, as a company, how are you working with your customers who might be at very different stages when it comes to implementing AI. How are you kind of managing the delta out there in the world?

Sure, we work very closely with our customers. We want to be the world's most trusted AI provider because trust, we think has never been more important. And it's not as simple as just plugging ChatGPT in and everything is fine and that's all you need to do. It takes you a little bit more sophisticated than that. So we work with big brands like Spotify uses our AI and their sales processes, they've improved, improved their sales productivity 40%. Or Yeti, a consumer brand, they use our AI and E commerce or leading companies to. My favorite is we think that with generative AI the world's going to go completely digital. And two of my favorite examples are Amazon and Uber. They've completely transformed their industries and, and both of them run their entire companies on Salesforce and use our AI products. So we think it's a big deal. We think there's going to be an Amazon and Uber in every single industry. And that's what we want to do, is in a trusted way, provide trusted AI so that you don't need your own data science team. You can just turn it on in Salesforce and it works. It's secure, it's trusted and it's scalable. And that's our strategy.

You mentioned data and I imagine there's a correlation between a strong data culture and successful implementation of AI. What's your advice for companies who are trying to establish a strong data culture in their business?

Sure. The data powers the AI and the AI we believe will power the customer experience of the next generation. So our fastest growing product ever built at Salesforce is our data cloud. And what it does is it takes, you know, the average company has over a thousand systems, so that means your customer names are in maybe 50 of those systems or maybe 100 or maybe 173 of those and they're all spelled differently and they use different email addresses. So this is all about the C&CRM or what I call the Holy Grail of CRM. You actually need to know who your customers are. And that's what this customer data cloud does. It takes data from those thousands of systems and gives you a single golden record, or what we call the single source of truth for every single customer. Now once you know the customer is, then you can upsell, you can cross sell, you can provide one to one service, you can do magical things. But it all starts with who's the customer.

So you've got companies using Salesforce with sensitive information in there. They're running their businesses, their customer relationship status, et cetera. You've mentioned trustworthiness a number of times and safety. So what are you doing to ensure that AI is operating in a way that is safe and trustworthy and will earn long term trust from your customers and consumers on the other end?

Sure. So the generative AI, what they call large language models, which are the foundational models that drive generative AI, they're still a work in progress. So they hallucinate, they have bias, so they're not 100% accurate as of today. And so what we built is what we call Einstein trust layer, so that we look at the questions people ask, but then before they see the output, we look at that and we screen out biased answers or harmful answers or other things that we think that customers don't want to send to their customers. And so this is part of our commitment. We have five core values at Salesforce and trust is number one. And we had to learn how to do that. SaaS is really the whole cloud started where we give you a slice of our public infrastructure and we make sure that 100%, not 99%, but 100% of your data is secure. So that's where we really developed this trust principle. Then we applied that to predictive AI, and now we provide that for generative AI. So we think it's very important that when you turn on generative AI, you, you know what it does, you know what you don't want it to do, and that you can 100% trust the output that it's going to give to your customers.

All right, I'm going to end by putting you on the spot here. Who's going to win? The US President? No. What's your most confident prediction for AI rollout over the next 12 months? What are we going to see happen?

Sure. I think Gartner has something called the hype cycle, where there's a huge boom in what's going on of generative AI. And then after that they caught the trough of disillusionment where everyone gets real and says, what can this really do? And then once people get through that, then there's a steady growth. This happened with the Internet. I don't know if people can remember when you had to do dial up and everybody said the Internet's going to change the world and it didn't right away. But these big technologies do change. So I believe in 12 months, every single piece of software will add generative AI. So everything that you use will have generative AI. And I will also predict that it's not generative AI that will take people's jobs. It's people using generative AI that will take the jobs from people that aren't using it. So it simply makes people more productive, more knowledgeable, that allows people to make faster decisions. So we think this is a big deal. We think it's going to change the world, and we do think it's going to create new jobs, too. That's the other thing that a lot of people don't talk about. I'll give an example. There's something called prompt engineering where the prompt is the question you ask children. ChatGPT that wasn't even a field one year ago.

That's going to be a job that.

If you have children and they're in college and they're trying to figure out what to do, tell them to be prompt engineers.

Yeah. All right. You heard it here. Well, David, thank you so much. I think you have a little time to hang around. If people have questions, feel free to grab him. And thanks for kicking us off, Josh.

Thank you. And thanks to the Wall Street Journal. Thank. You. Sa.