AI agent

“I think there'll be better architectures and better data collection ways to make AI agents more productive. And I think we will see that. Out of top research labs and top startups in the coming, you know, 1218 months.”

“I think where we have seen it actually working at some sort of scale would probably be coding, customer support, or sales. And generally the ones that have worked better are very verticalized. So they're doing a clear workflow really well. They're not a genera…”

“I think in the next few years, we're going to see more AI agents surfing the web for people than we'll have people surfing the web.”

“That's what's going to happen when intelligence gets cheaper and cheaper. We'll just be all much more efficient. And most of us have to become managers of AI agents doing work for us.”

“If your agent is like 95% accurate on each step and you multiply .95, like times 15, then like half the time that process will fail, right? If it's not like implemented in a way that's very robust to errors in each step.”

“I do think the year, I mean, 20, 25, everyone's been talking about agents. I think you'll start to see agents that actually work pretty well.”

“And I think in twenty-twenty-five, going back to your first question, we'll make some pretty material advances there, where the predictability is much better, and that will lead into, I think, twenty-twenty-six, where we'll start to see pretty significant ROI …”

“And we used to have agents overseas that would do it. Now the agents are AI.”

“There's a version of this actually happening kind of in AI agents as well, which is, like, will the agents browse the internet like humans do, which is sort of the humanoid side of things, or will we be Changing how the internet works so that agents can intera…”

“The biggest thing holding back AI agents from really browsing the web reliably is authentication, and there needs to be some innovation there, probably from a couple very well-established companies.”

“So if it hallucinates a refund policy, if it spews Nazi propaganda, if it leaks your customers' data, if it discriminates against the people you're hiring, the enterprises are going to be on the hook.”

“Once Walmart has like, 50 AI agents deployed, they're not going to want each individual provider to ensure them. They're going to want to have one big coverage policy.”

“So now we have agents all night going through testing all the different possible things that a customer might do the next morning when they wake up, and then we catch the errors before they do.”

“And we are headed into a world of agents that are going to be spewing out even more telemetry information about are they working well or are they not working well. Lots of complicated pieces interacting with each other.”

“The way we're thinking about it today is we want to make sure there's an ass in seat. Like, are you physically watching this thing?”

“So forms of shopping that are pure entertainment are never going to go entirely robotic.”

“I don't think people are ready to give their money to Open Claw and say, make me money now, please. You know, like, I think we're far away from that, and you know, there's a reason why everyone wants to have a little Mac Mini, because they want to have these t…”

“And the cool thing is now with AI, we can have AI agents essentially look at searches, et cetera. What are the most common requests? Customer service, same thing. When we get calls, all of that can be summarized and can be put into what could be a potential ro…”

“And so I think that that is just creating a massive need for more CPUs and more memory. And we're seeing that we're actually just at the beginning of that. I think agents is really what is driving that.”

“What's changing is now you're getting persistent memory with all of the next evolution of accelerators. And so I think that you're seeing memory content is increasing. The agent is not going to have amnesia anymore. The agent is going to be able to remember.”

“Your footprint, you're like a semiconductor or technology footprint as an individual is about to expand by like 10 X because you used to have a couple of devices. Maybe you're someone has an iPad too. So you have X a little more, but like three or four devices…”

“If you or me in, in a year or two, we're gonna have a claw that's running hundreds of agents at all time. I'm gonna have a similar impact at work. I'm gonna have hundreds, maybe thousands of agents running at all time.”

“Agents, agents launching agents, I think is one of the most underappreciated unlocks that has happened, because you can give an agent a project now, and you can just go away, and you can come back, and it has done If not all of it, really, really far along.”

“I think you have to, because I think a lot of times not for all software, but for some software, like you're not going to have time to make a decision for what vendors you use. [1925] Shensi Ding: We notice this with packages a lot, also in cloud code, scarily…”

“Agents are pushing a ton of code. You don't have enough humans to read all that code, and so things are slipping by, things are getting in, and one of them was a vulnerability that gets injected into an open source package that everybody relies on and uses, an…”

“We don't trust agents. We say, Hey, we can try to set rules, but there needs to be hard guardrails and blocks for things like sensitive data being sent across.”

“You see a lot of people trying to build, like, a custom harness, or, You know use something like a workflow builder to build agents that are repeatable and all of that. And I just ultimately think none of it matters because we're almost at the point now where …”

“I think you're going to start seeing things like this for token billing, where there's going to become this whole ecosystem of like, how do you optimize around this? Because you kind of have these weird misaligned incentives from the model providers, right? Be…”

“And you have to, it's like a shift in mindset where, you know, let's say that it creates a draft pull request for some new feature, and it's not quite perfect. Your instinct typically would be to go in there and actually edit the pull request to make it better…”

“We're also going to see AI agents have their own financial accounts, and we're powering that with the base, like, MCP API so they can have their own self-custodial wallet now and start to have a true agentic economy, paying for goods and services, hiring other…”

“We built out a lot of infrastructure for AI agents to actually get their own financial accounts and connect them into your own, you know, your Coinbase account.”

“In the future, you're increasingly gonna Talk to one agent. That agent is going to orchestrate hundreds of thousands of other agents. It's going to pay for goods and services with companies that it needs to engage with to get work done on your behalf.”

“Well, I mean, I have started writing some code again using AI agents. I test out the tools every few months, make sure I can ship a few pull requests and like still get in the weeds.”

“Agents start using inference in an exponential way. The demand curve goes way up for what we, what we're going to need from an inference perspective.”