# Anthropic CEO Dario Amodei outlines state of AI and impact of pace of models at Dreamforce 2026

*USA TODAY* · https://www.youtube.com/watch?v=KLH4Y-S2lMc

## Summary

At Dreamforce, Salesforce's Marc Benioff interviews Anthropic CEO Dario Amodei about the state of AI and his recent essay on pacing. Amodei explains his three-step plan for industry safety standards, shares that the pace of economic adoption has been his biggest surprise, and argues that even frozen in place, the technology is only 5-10% diffused. He closes by encouraging the audience to unlock more of AI's value, illustrated by Anthropic's own use of Claude with Salesforce data.

## Key points

- Amodei outlines a three-step response to safety incidents: examine your own record, organize industry-wide standards, and add an international component.
- He says Anthropic has committed to the first step and will open a dialogue with the rest of the industry on the others.
- His biggest surprise over the past decade is the pace of progress on the economic side, not just the technology.
- He formulated the scaling laws with co-founders, predicting models get smarter with more compute, but underestimated how fast companies and products would grow.
- Even if technology were frozen in place, Amodei estimates we use only 5 or 10% of its possible value.
- He gives the example of Anthropic's go-to-market lead using Claude with Salesforce data to analyze the 10 biggest deals, deals most likely to be lost, and how to win them.
- Benioff highlights the data layer, application and semantic layer, agentic layer, and interface layer, with the first interface on co-work.
- Benioff notes Anthropic has become the number one AI in the world, three times larger than the nearest provider.

## Notes

### Safety and Pacing Essay
- Amodei uses a car company analogy: when a competitor has a safety incident, the responsible response is not to attack them but to examine your own record, improve practices, recommit to transparency, and invest more in safety.
- His three-step plan: improve your own standards, organize the rest of the industry to set shared standards, and add an international component.
- Anthropic has committed to the first step and will have a dialogue with the industry about the others.
- He frames this as leading by example: everyone can always be better.

### Biggest Surprise of the Last Decade
- The surprise is the pace of progress on the economic side, not just the technology.
- He formulated the scaling laws with some of his co-founders: AI models get smarter the more compute you throw into them.
- The prediction of capability was there, but the speed at which companies grew and products like MCP, co-work, and Claude skills became central to everything was unexpected.
- Adoption is fast because the economic value is so great.

### Recommendation to the Audience
- Even if the technology were frozen in place, we are making use of maybe only 5 or 10% of its possible value.
- Example: Anthropic's go-to-market lead uses Salesforce with Claude daily, asking about the 10 biggest deals this week, the 10 deals most likely to be lost, the themes in those deals, and how to win them.
- The fraction of people who have seen this is small compared to those who could gain utility from it, so much diffusion is left to do.

### Salesforce and Anthropic Partnership
- Benioff praises the data layer, application and semantic layer, agentic layer, and the new interface layer, with the first interface on co-work.
- Benioff notes Anthropic is the number one AI in the world, three times larger than the nearest provider, and that Salesforce alumni help run Anthropic.

## Transcript

[00:00] Here we have the CEO of Anthropic right here, Daario. Hello, Ario. How are you? [cheering] >> Hey, >> great to have you. Fantastic. [applause] >> Great job, Dario. You're doing an amazing job leading the AI industry, guiding it all right from our Slack headquarters building. We watch you every day. >> We're renting it from you. [laughter] >> It's all good karma, isn't it? And uh we're watching you doing an incredible job. You become the number one AI in the world. Three times larger than the nearest provider. It's been amazing to be partnered with you. It's all been very karmic and kismmet and good. And one of the things I have to ask you before we get into the interview, I want to talk to you about AI force. I want to

[00:41] talk to you about Claude Force. I want to talk to you about how you're running your own business. But wow, you wrote, you know, you I mean everything from your first essay that you wrote, you know, about loving grace to to the essay this weekend on pacing has been incredible, but give us your state of the AI industry and pacing and where your mind is right now. >> Yeah, I I think it's maybe easiest to explain it by an analogy. So, let's say you're running a car company and another car company, not yours, they have some kind of safety incident, right? Something goes wrong with the brakes. uh you know something with the manufacturing now you believe you have the best safety record in the industry

[01:20] what do you do how do you approach it obviously it's very tempting to attack your competitor and say these guys are unsafe we're safe but I think the more responsible way to respond to it is to say hey first of all let's look let's look at our own record right we may not have had this big high-profile incident but I'm sure we're not perfect let's look at everything and we can always be better let's make our practices better um Let's recommmit transparency. Let's invest more in safety. Then let's organize the rest of the industry and say, "What can we do to set standards for everyone? What can we do to make everyone's standards better?" And then finally, there should be an international component to it. And and

[01:59] that's essentially the the three-step plan that I laid out that I laid out in the essay. The the first of which we committed to and the the others of which we're going to have a dialogue with the rest of the industry about. And I think I think that's the way to lead the industry forward to set an example to say that everyone can always be better. >> Let me ask you a question Dario. You know it's been amazing what you have been able to accomplish like I said but there are a lot of surprises also that you have gone through. So, you know, just tell us what has been your biggest surprise over the last decade and you know, you've obviously been heads down in AI now for 10 years really in earnest. What's your biggest shock about

[02:42] where we are right now? You know, it's it's been the pace of the progress not just of the technology but of the economic side of it. So we, you know, I formulated the scaling laws along with some of my co-founders which tell you that AI models get smarter and smarter the more compute you throw into them. And so we had a prediction of just how capable these models would be. But I think what we didn't appreciate is that it would lead to these companies growing so fast to all these incredible, you know, products that we're working together on and you've been talking about today and and how quickly they would become central to everything that's going on in the world. Right. I just watched the talks before that

[03:22] where, you know, you guys talked about MCP and co-work and claude skills and these are all new things that people are just beginning to use, but they're being adopted so quickly because the economic value is so is so great. So, I' I've just been surprised that such a pace is possible in the business world. >> Well, I'll tell you at Salesforce, we are huge co-work fans and also when I was in Europe, I met with so many customers who are co-work fans. It's incredible what you've done done with the product. I noticing over your shoulder here, you've got a few Salesforce alumni behind you who are helping you to run your company, which >> you train them well, Mark. You train them well. [laughter] >> Trying to do my best to help the whole

[04:05] industry. You know, [laughter] >> I measure myself by how many COs and CRO are in other companies that all came from Salesforce. >> You you do a fantastic job of that. No, truly truly you do. >> Thank you. But I what I want to say is this. I I want to say, you know, yes, we're have to keep our eye on the pace of AI. You know, it's been amazing what's happened over 10 years. No one could have written the story. Now, these interfaces are truly amazing. The demo we just saw is remarkable. I think every customer here wants to have that same interface on all of their systems. Everybody wants to unlock the trap value, you know. What what is your kind of recommendation for everyone in this room and the virtual

[04:52] room? Obviously, you can see we have 10,000 people all over this room, but 10 million joining us online. You know, what actions do you want us to take? >> Yeah. So, you know, I would I would just say when I look at the technology versus how much it has diffused, even if we freeze the technology in place, which we're not doing, facing the frontier is not not doing that at all. the technology will keep getting better. But just suppose we freezed it in place. Even if that were the case, we're making use of maybe only 5 or 10% of what the possible value of the technology is. So I'll I I'll give you just an example that relates to what we're doing together and to the uh CCO you just uh you just pointed at which is that Paul

[05:38] uh you know every day now he he wakes up and if he wants to look at you know our our head go to market prospects he still uses Salesforce but he uses Salesforce with Claude. He talks to Claude using Salesforce's data about you know what are the what are the 10 biggest deals we're doing this week? What are the 10 deals we're most likely to lose? what are the themes in those deals? How do we win those deals? You know, just just making use of all the information fluently with Claude and and it's it's really incredible and and the fraction of people who have seen this and seen that this is possible is still small compared to those who could gain utility from it. And so there is still so much diffusion left to do. Well, we are

[06:19] really excited to wake everybody up at the show to show show people what the possibilities are for the future, how we're going to show them that data layer, which as you know, I think is really critical, the having that grounding of the data. We've talked about that so many times, the application and semantic layer, giving it the AI even more intelligence capability, the agentic layer that you've really have helped pioneer, and now this incredible interface layer, and we're so excited that the first one is on co-work. Congratulations. Well done. >> We're going to do incredible things. You're doing a great job. Thanks for everything you're doing for the whole industry. We're so grateful to you.

[06:54] >> Thank you for having me. >> Really thrilled. Thank you so much. All right. And >> thank you for coming to Dreamforce. Please thank the anthropic team. [applause]