AI Agent

“Actually I, I'd say 50% of our end users are now agents and the rest 50% are humans.”

“This is the context that any agent, actually human or AI agent needs to be able to solve this problem for the customer. Knowledge, expertise, and norms. So if you had to build this in a business, you need to be able to bring together first the map of the busin…”

“Investing in the Orchestration layer or the model layer, like that's the easier problem to solve. Building an agent is easy. How do you actually make it accurate? How do you actually manage your context? That's going to be the hard problem.”

“And so we've made some simple tools where with a single Command. You just prompt, you paste it into your AI agent. It can now have its own financial account and it's using a self custodial wallet and crypto rails.”

“It's going to be possible to create a separate agentic account with segregated funds on Coinbase in the near future. And you can give your agent access to just that to do trading and payments on your behalf. That's kind of linked to your identity as human.”

“And in the future, I think you're going to want to have these agents have their own spend accounts, kind of like having a personal assistant or an employee who has their own spend account. It's going to need probably like a stablecoin-backed credit card for ba…”

“This is just how I, by the way, side note, this is how I'm thinking before I hire anyone. I'm like, should I just do, get an agent to the, To do this. Sometimes yes, sometimes no, but that's been the default of how I've been thinking as I'm building my company…”

“I do still believe that, like, AI agents are not that good today actually at doing something that is very relationship driven. But they're getting much better at it, right? So, like, the problem is when you look at LLMs, they look at the distribution site, and…”

“And the reason for that is we think as these models get smarter the same way that your teammates, you know, you wouldn't ask for second by second updates of exactly all the buttons they're pressing and websites they're going to. I think it's too much to ask yo…”

“And so in order to actually deliver on that promise, it's actually not a lot of feature product roadmap style stuff. It's like hill climbing five really important problems in the back end that many users don't directly experience, but you totally feel when, yo…”

“I think we're going to see more of that type of stuff where the agent or the bot should actually be more proactive to you than you reaching out to it. And I think that will be the next shift in AI.”

“It's moving way faster than any other technology adoption that I've ever seen.”

“Really an agent is the ultimate settings screen for any software and code is the ultimate setting screen for any software.”

“And compared to a few years ago, You do not have the need to go and hire people that you do not trust, that you do not trust to have high agency or the right skills, because you can use an agent to do those things.”

“I think you'll trust your agent to decide what data to share with other people without you intervening in five years.”

“I think it is the human's responsibility to first allocate the resources. That means to say how many tokens are we willing to spend to try to get a goal to set the goal right as well. So you set the goal and the budget and then to monitor, right? Like the over…”

“The loop will help you climb to the local maxima, but then it plateaus, and you need some sort of out-of-distribution thinking, you need human intuition, you need somebody to actually help you land at the base of the next hill.”

“So for the next 10 years, we're going to have security issue after security issue. But there's going to be a moment where agent written exquisite code exceeds human written error filled code.”

“Number two is I think the concept of agents where you don't really need A clean API per se. You can make do with unstructured data and take inputs and shoot out outputs that are processable, human readable, but also machine readable. And that to me is going to…”

“The learning and here's the meta learning guardrails aren't enough. You have to have a lock and key. Guard, it doesn't matter whether you build 80 gates, a hundred gates, 200 gates, like, they're not enough, and they get, and then, you know, what's even worse,…”

“There's no, you can't take Jason's agent out to a steak dinner to get him to buy your product. It just has to be better.”

“I think at the same time, you've got sort of agents that are operating on their own. That's probably the smallest of the categories right now. I think that that will grow and become a pretty large category, but right now that's sort of Agent to agent kind of c…”

“If every one of those agents ran in a traditional container, like, you know, which is not even a VM, not even its own server, but just a container, right? You would need more than 40 times the current CPU production in an, in a calendar year in order to suppor…”

“We think that somewhere between one and 10% of those will day one be monetizable through agents with a with a micropayments transaction, which means that if you're going to build something like this, you've got to build a payments infrastructure that can day o…”

“And then by the way, we get to worry about the next problem, which is that AI is this massive divider between, you know, the rich and the poor. And if we're not in an ad based business model in anymore, which I think is likely if you're spending thousands of d…”

“If you're running your model at over 4000 tokens per second. Now the, you can do, you know, more agentic loops. You can do more reasoning. Ultimately you get significantly more capable, more intelligent agents.”

“Atlassian just launched. They finally were able to really monetize their AI agent and it's crushed their stocks up.”

“So we're seeing a lot of pull towards agentic solution, but also caution because they are not willing to kind of cede control completely to agents for their core infrastructure and core security systems.”

“People are going to make mistakes more than agents if we build our agents with the right guardrails around it. And so our belief is that actually we'll make fewer mistakes in an agentic world We'll tell people what the right solution is. We'll keep, take more …”

“I think the complexity is going to go up in order of magnitude when these agents are out.”

“Agents are moving from answering questions. They're operating tools and phones is one of the biggest tool services in the world.”

“And you train these agents the same way you train a new hire. You start with small tasks, you watch at work, you correct the mistakes, You add the correction to memory, because now we have memory, and then you expand the scope as you increase your comfort leve…”

“If you look out three to five years, those agents are going to be fully autonomous.”

“And then that job will be made irrelevant five years from then, because AI agents will govern themselves.”

“I do think that increasingly the future of work will look more like steering than rowing in the sense that there will be agents that you'll, you'll be able to work with that do a lot of the rowing and your role increasingly becomes steering the ship in the rig…”

“People are increasingly thinking about agents that are persistent, that feel like teammates, that feel like coworkers. Where you can work with them the way I might work with someone on my team, which is they do a whole bunch of work. I provide input and then t…”

“It turns out that if you have multiple agents, then the agents develop more context and expertise. So they get more specific”

“I think in the past, the way that most SaaS Products thought they spent a lot of time on their user interface, and now they really got to think about the agent interface, not just user interface.”

“I mean, look, I think that SaaS products can offer their own agents, and they can offer their own AI and copilots and all that, but I think the reality is, Is that most of the action with agents is going to happen outside those products and simply want to inte…”

“In CX and support if I'm talking to an agent on someone's website as a potential customer and that agent keeps me forever, I might not need, there may be no need for the CRM or CX system or SFA that undermined it. Used to underlie that interaction because the …”

“The thing about headless is it also makes it much easier to bifurcate our data and put some into outside of the system of record and some in because the agent, the headless version of your CRM or SFA or system of record, well, it has no problem up to a point p…”

“You're going to lose in the vendor war if you're not the one that not just is recommended in AO, but where the agent says, I will just do it with this API.”

“The email agent thing, I was on with an entrepreneur, actually two in the last week, and he's like, well, I have, like, the, I call the agent, and I text the agent, and I was like, yeah, but email's by far the best, and he's like, yeah, it is. Email's sick. Th…”

“There are over a billion knowledge workers in the world, right? And with that, it's a long ways to go for that demand. And by the way, the work that they do, all this workflow and automation, So on, and then you get to the back office, which is all the agents.…”

“Next year, there's going to be backlash. I can see it in my best portfolio companies where we can't go back anymore. We can't live. We can't go back in time. And I need my 10 sub agents running 24 hours a day to do my job or I quit.”

“Agents, agents, they kind of just guess what to do, right? And if they're smart enough, they guess right most of the time, but sometimes the agents guess wrong. And the second thing is agents are using tokens, right? We probably heard token maxing, right? And …”

“Well, so you could, but the thing is, you're only gonna be able to keep up with, what, 20 agents? Cause it's gonna keep assigning you tasks and more things to do and more things to do. And so there still is like a, there still is like a limit.”

“Yes. I feel pretty confident that that is the direction we're heading to.”

“And increasingly, I think we will have agents do some unit of work, take as many shots on goal as they can, and however many tokens they use to get there is going to be kind of a dependent variable depending on the task. So you think about, like, agents that s…”

“We're already at parity in terms of agent traffic to human traffic.”

“I think it was Cloudflare this morning said that in five years from now, we'll have 1000 times the amount of agent traffic to human traffic.”

“Why I like rock bot and how I use it is I only add an agent if we really like needed, I just, I'm so anti idea creep and agent creep because that's how you burn tokens. That's how you'd spin your wheels. That's how you don't make process progress.”

“This technology around simulation, creating accurate representation of people ought to precede the more complex agents that would automate the world that we live in.”

“Well, I think we are starting to see the added dimensions of the alignment challenge that open up once you have systems that are sophisticated enough because the space of possible strategies that you can pursue is a function of your cognitive capacity. Like yo…”

“And stable coins are today relatively unergonomic. So if you're a human, you know, you have to go jump through a bunch of hoops. But if you just give an agent a budget, they can easily, you know, take dollars, move into stables, or, you know, just use a shared…”

“So the way that we're going to need to reinvent what that kind of interactive interruptible work looks like in the agent layer, which is separate from the work going on, I think in the, like in the model layer. And in this like interactive model space, but I t…”

“So the best companies in PLG focus right now will also be best positioned, in my opinion, for the agentic layer of essentially decisions being made without that human in the loop when tools are being selected for certain tasks.”

“I took all the crits, all the feedback that we've given over the last couple of years, right? These like Zoom meetings or raw notes, we just dumped it in a database, right? Gave it to an agent and said, look, from here on out, before we even get to those revie…”

“The reality is agents are going to come into our lives, like for every business, much faster than I think we're going to be able to build up a population of people that know how to build them.”

“Now everyone is 2.2. We've been on the journey for a bit now. Right. So everybody is now observe 2.2. We took the older product, right. That we built completely replatformized it. It's now all agents and whatever new builds all agents anyway, sellers are all a…”