30:18Greg Isenberg · Tutorial
How to Start an AI Agent Business Today
A 30-minute screenshare where boring arbitrage ideas become cash-flowing businesses with a few prompts and a Slack webhook.
May 11thThe argument in one line.
Frontier agents have crossed the threshold of true autonomous execution, and the entire gap between winners and everyone else comes down to whether you commit to 30 to 90 days of daily deliberate practice rather than one-shotting a task and giving up.
Read if. Skip if.
READ IF YOU ARE…
- You are building a solopreneur business or early-stage startup and want a framework for which agent bets to place now.
- You are evaluating agent platforms (OpenClaw, Manus, Perplexity Computer, HyperAgent, Codex) and need a principled comparison.
- You already use AI tools but still one-shot tasks and give up when the first output is mediocre.
- You want a concrete example of going from a startup idea to a V1 app and marketing site in a single agent workflow.
- You are trying to understand how to scale from one agent to a fleet with quality control built in.
SKIP IF…
- You want a technical deep-dive on agent architectures, tool-calling, or infrastructure — this is a business and product conversation.
- You are skeptical that the capability shift is real — the episode argues from the premise, not for it.
The full version, fast.
Frontier agents crossed a capability threshold roughly five months ago, shifting from AI-augmented humans to fully autonomous multi-step execution. The guest argues the real addressable market is the entire white-collar labor pool across the Western hemisphere, not Sequoia's trillion-dollar estimate. HyperAgent, built by the Airtable team, positions as the Mac to OpenClaw's Linux: cloud-native, UX-first, with skills (reusable task playbooks), rubrics (LLM-as-judge eval loops), and a command center for fleet management. The sharpest takeaway is on persistence: 99 percent of people quit after one attempt, and that gap — not capability — is the real arbitrage for whoever commits to daily practice.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Who's talking.
01:52guestHowie Liu
00:00hostGreg Isenberg
Where the time goes.
00:00 – 02:22
01 · Intro / cold open
Greg frames Howie's background (Airtable, half-billion in revenue) and announces a $1M HyperAgent credit giveaway for the first 1,000 listeners.
02:22 – 04:41
02 · Sequoia's AI agent deployment chart
Side-by-side chart review: software engineering at ~50%, most categories under 10%. Howie argues even 50% understates the frontier shift.
04:41 – 08:13
03 · Copilot vs Autopilot and the $1T+ opportunity
Sequoia's trillion-dollar framing; Howie argues the real TAM is all white-collar GDP, and the capability unlock happened with Opus 4.5.
08:13 – 11:12
04 · Agent economics vs human labor costs
Unit economics slide: AI agent at $0.25-$0.50/interaction vs human at $3-$6; 88% cost reduction. Howie tells the board memo story to reframe token cost as opportunity cost.
11:12 – 14:48
05 · Fastest enterprise adoption curve in history
Gartner chart: <5% in 2025 to 40% by 2026 EOY. SaaS took 6 years to do this. Howie on OpenAI + Anthropic reaching $80B+ revenue from zero.
14:48 – 18:03
06 · The agent command center and fleet of 20 agents
Visual mockup of agent command center running Customer Intel, Content Production, Competitive Research, Lead Enrichment, Inbox Triage, M&A Diligence simultaneously.
18:03 – 19:43
07 · What is HyperAgent?
Product positioning: the Mac to OpenClaw's Linux. Cloud-native, secure by default, UX-first agent platform.
19:43 – 22:38
08 · Live demo: hyperlocal real estate market report
Agent researches market, validates on Reddit, does competitive analysis, builds V1 app (Blockpulse) with map, dashboard, and marketing site from a single startup idea.
22:38 – 23:21
09 · HyperAgent as the founder, not just the developer
App building is a commoditized feature. HyperAgent positions as the full founder: research + build + business context.
23:21 – 24:15
10 · Street View, Zillow redesigns, visual tool power
Agent uses Google Maps Street View to find billboard locations; uses images as seed for AI image and video generation.
24:15 – 25:48
11 · Command center view across agent fleet
Fleet dashboard: agent names, run counts, costs, quality scores, last run times, deploy to Slack in one click.
25:48 – 26:30
12 · Skills as the key primitive
Skills are reusable task playbooks that give generally-intelligent models domain expertise. The Einstein analogy: give him a real estate manual and he figures it out.
26:30 – 32:31
13 · Building the Greg Isenberg contrarian AI skill live
Howie builds a real-time skill that researches Greg's voice, distills a style guide, and produces contrarian AI tweet drafts for X.
32:31 – 34:48
14 · HyperAgent vs Perplexity Computer, Manus, OpenClaw, Codex
Positioning matrix: Codex is coding-only; OpenClaw is raw and technical; Manus and Perplexity Computer are closest comps but less fleet-scale.
34:48 – 36:15
15 · Reviewing the writing skill
Howie reads back the agent-generated Greg Isenberg voice profile: "smart friend at dinner saying the quiet part out loud."
36:15 – 37:15
16 · Reviewing contrarian tweet drafts live
Tweets reviewed live: too formal. Demonstrates the feedback loop for real-time skill refinement.
37:15 – 38:31
17 · The arbitrage of persistence
The tennis analogy: messy middle, daily practice, compounding. 99% of people quit after one try. The knives-vs-internet parable.
38:31 – 42:10
18 · Confidence milestones and daily practice
Greg on the psychological checkpoints ($1, $10K/month). Howie on committing 30/60/90 days at 30 minutes daily.
42:10 – 46:55
19 · Rubrics and the eval loop
Rubric as eval framework pinned to an agent; LLM-as-judge scores outputs; trend lines on quality over time; automatic model downgrade suggestions.
46:55 – 53:03
20 · Connectors, OAuth, and custom API skills
Out-of-box connectors (Slack, Gmail, Notion, Granola); agent self-teaches new APIs (Twilio demo: voice and SMS service).
53:03 – 1:01:54
21 · How to get started with HyperAgent
New onboarding: connect to existing accounts, agent learns your context, suggests relevant use cases. VC deal flow example.
1:01:54 – 1:03:51
22 · Credit giveaway and closing
$1M in HyperAgent credits; first 1,000 builders get $1,000 each.
1:03:51 – 1:05:11
23 · Final thoughts
Howie's vision: $100B companies with fewer than 5 employees built on agent fleets.
Lines worth screenshotting.
- Frontier agents crossed a true capability threshold roughly five months ago — this is a modality shift from augmentation to autonomy, not incremental improvement.
- Software engineering at 50 percent agent adoption is actually an overestimate of the frontier mode, because most teams are still running three-year-old AI practices.
- The real agent TAM is not a trillion dollars — it is the entire white-collar GDP of the Western hemisphere, many tens of trillions.
- Stop anchoring AI costs against software subscriptions; anchor them against human time — a $150 board memo beats days of executive hours.
- Context window limits mean agents will always be partitioned into roles, making the humanoid job-role form factor of today's agents structural, not a temporary phase.
- Skills are the most important primitive in frontier agents: they turn generally-intelligent models into domain experts via playbooks without fine-tuning.
- Rubrics — LLM-as-judge eval loops — are what separate a single-agent experiment from a scalable business operation.
- The arbitrage is pure persistence: 99 percent of people one-shot a task, get a mediocre result, and quit before the compounding begins.
- One good tweet per day beats ten mediocre ones; content is a hits-driven business and human quality judgment still wins.
- Agents are converging on human job roles for the same reason robots converge on humanoid form — existing infrastructure is built for human ergonomics.
- PLG and top-down enterprise are both valid agent business models; the difference is speed versus check size, not winner versus loser.
- The gap between people who get the agent shift and people who do not is almost entirely explained by whether they have spent a full weekend hands-on with frontier agents.
- A daily 30-minute calendar block for 90 days is what produces a top-one-percent agent operator, not innate talent.
- Small shops will outmaneuver large incumbents on AI adoption for the same reason startups always do: agility and no bureaucracy.
- Every company will run a fleet of agents; the only question is who builds the quality-management systems to run that fleet reliably.
Five things that actually move the needle with frontier agents.
WHAT TO LEARN
The capability is no longer the bottleneck — your operating habits are, and the gap between top-one-percent operators and everyone else is almost entirely explained by persistence and daily practice.
02Sequoia's AI agent deployment chart
- Frontier teams now run dozens of Claude Code instances in parallel, fully autonomously — the 50% software engineering adoption number understates this shift because most companies are still on three-year-old practices.
- Every white-collar category in the Sequoia chart should reach 100% agent penetration with today's frontier models; the lag is adoption speed, not capability.
03Copilot vs Autopilot and the $1T+ opportunity
- The $1T Sequoia number is a floor; the real TAM is all white-collar labor in the Western hemisphere, many tens of trillions.
- The capability unlock happened with Opus-class models that can take a multi-day human engineering task and return a clean PR autonomously.
04Agent economics vs human labor costs
- Anchor AI costs on human time, not SaaS subscriptions — $150 in tokens for a board memo beats hours of executive labor and produces better output.
- The unit economics already show 3.5 to 8x ROI on $1 spent in leading organizations, with year-three ROI projections of 1245%.
05Fastest enterprise adoption curve in history
- Enterprise AI agent adoption is moving faster than mobile or SaaS ever did — from under 5% in 2025 to a projected 40% by 2026 year-end.
- The revenue of the AI category going from zero to $80B+ in a few years has no historical parallel in software.
07What is HyperAgent?
- The Mac versus Linux framing is useful for any agent platform choice: some people want to configure every detail from the command line, others want something that just works securely by default.
- Cloud-native agents remove the Mac mini dependency — the computer the agent runs on is not your problem.
08Live demo: hyperlocal real estate market report
- A single well-scoped prompt can drive an agent to complete market research, Reddit validation, competitive analysis, a V1 app build, a marketing site, and ad creative in one workflow.
- Medium-sized markets ($1-5B TAM) are often more attractive than massive ones because large incumbents ignore them while you can build a lucrative niche business.
12Skills as the key primitive
- Skills work because frontier models are generally intelligent enough to apply any domain playbook — the leverage is in the playbook, not in the model.
- A skill built interactively through feedback improves over time; one-shot prompts do not.
14HyperAgent vs Perplexity Computer, Manus, OpenClaw, Codex
- Choose your agent platform based on whether you need a single powerful agent (most tools do this well) or a managed fleet with quality observability (far fewer tools do this).
- The real differentiator is fleet-scale management: deploy to Slack, track quality trends, auto-improve skills — not just one great agent run.
17The arbitrage of persistence
- The SEO-versus-door-to-door knives parable: two months of zero revenue while learning the new paradigm beats years of slow decline in the old one.
- 99% of people quit after one mediocre agent output — that quit rate is the arbitrage for anyone willing to push through the messy middle.
18Confidence milestones and daily practice
- The first internet dollar is a psychological unlock, not just revenue — it proves the model works and rewires what you believe is possible.
- 30 minutes in your calendar every day for 90 days is sufficient to reach top-one-percent agent operator status; consistency beats intensity.
19Rubrics and the eval loop
- Pinning a rubric to an agent is management 101 applied to AI: you cannot review every output personally at scale, so you need automated quality checks.
- LLM-as-judge scoring can also suggest model downgrades — if Sonnet scores only slightly below Opus, you just got a 5x cost reduction for free.
20Connectors, OAuth, and custom API skills
- If there is no pre-built connector to an API, the agent can read the documentation, build a skill, store it, and use it from that point forward.
- Always-on agents in Slack that listen to team conversations and chime in when relevant represent a qualitatively different kind of coworker relationship than a chatbot you query.
21How to get started with HyperAgent
- Connect the agent to your existing accounts first so it learns your actual context before you ask it to do anything — this eliminates the blank-slate problem.
- Start with a specific real problem from your own work rather than a toy task; the agent needs real stakes to produce useful outputs.
Terms worth knowing.
- Skills (HyperAgent)
- Reusable, composable task playbooks that tell a frontier agent how to perform a specific domain-expert role. They are built interactively, self-improve through feedback loops, and can be pinned to agents or shared across a team.
- Rubrics
- Eval frameworks pinned to an agent that define what good output looks like. A separate LLM acts as judge, scoring every output against the rubric dimensions so you can track quality trends without reviewing every run manually.
- Command Center
- A fleet dashboard in HyperAgent that shows all agents running simultaneously with their run counts, costs, quality scores, last run times, and health status at a glance.
- Live Mode
- An always-on agent behavior (heartbeat polling) that continuously monitors inputs and pushes outputs to the user via Slack, email, or Telegram without requiring manual triggers.
- Frontier agents
- Agent systems running the most capable frontier models (Opus-class) with full tool use, multi-step reasoning, and browser/code execution — as opposed to chatbots or simple AI autocomplete.
- Copilot territory
- Sequoia's framing for AI use cases where humans remain in the loop and AI augments specific tasks, such as autocomplete or draft generation.
- Autopilot territory
- Sequoia's framing for AI use cases where agents complete full workflows end-to-end with minimal human intervention — the stage Howie argues we have now entered.
Things they pointed at.
02:22linkSequoia AI agent deployment chart and $1T software map
19:43productBlockpulse
32:31productPerplexity Computer
32:31productManus
32:31productCodex
32:31productOpenClaw (Claude Code)
55:15productTwilio
53:03toolGranola
03:57linkAndre Karpathy blog post on AI coding inversion (Oct/Nov)
Lines you could clip.
06:51
“The TAM for that is not even a trillion. It's probably the whole GDP of all white-collar labor, which is obviously many tens of trillions.”
10:28
“I got feedback that that was the best memo from some of our best investors that I had ever written. And I'm like, yeah, because an agent did it. And by the way, I got to do it in ten times less time.”
37:40
“Don't give up after the first shot. The agents are powerful enough to do almost anything you want. The issue is not whether it's capable. It's whether you are able to invest the time and coaching and curation to get it there.”
41:45
“When you make your first Internet dollar, no matter what it is, it rewires your brain.”
1:04:55
“I want to see your listener base generate a hundred billion dollars of legit companies with less than five employees.”
Where the conversation goes.
00:00 – 11:12denseMacro case for agents: under-penetration and TAM
08:13 – 14:48denseUnit economics and reframing token cost
18:03 – 34:48denseHyperAgent product walkthrough and live demo
25:48 – 46:55denseSkills, rubrics, fleet management
37:15 – 53:03steadyPersistence, confidence milestones, getting started
46:55 – 1:01:54steadyConnectors, API skills, onboarding
1:01:54 – 1:05:11sparseCredit giveaway and closing vision
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
00:00Howie Lu is an absolute legend. I mean, this guy started Airtable. Half 1,000,000,000 in revenue, a billion dollars in the bank growing quarter after quarter.
00:09So he's one of those people that when I wanna know where is the world going, I call Howie. This episode is structured into two parts.
00:17First, where is the opportunity when it comes to AI agents? I think that there's a trillion dollars up for grabs in AI agents.
00:26Does he think there's more? Does he think there's less? Spoiler alert, he thinks there's way more and we get into it.
00:31The second part of the episode is where he reveals hyperagent.com. Now Hyper Agent is an AI agent builder that allows you to build digital employees, allows you to build apps on different ideas, and I don't know why more people aren't talking about it.
00:47So I had him just give us the tips and tricks for how to use HyperAgent so that you can outperform 99.9% of people.
00:56I got good news. Howie is gonna give you a thousand dollars of hyper agent credits, no strings attached. You just log in to the account, there's gonna be a thousand bucks right there to go and build the business of your dreams.
01:08The catch is, first a thousand people do it, get the thousand dollars. He's committing a million dollars. How crazy is that?
01:15Just writing a million dollar check of tokens to you, to the Startup Ideas podcast community, to play with HyperAgent, to automate some stuff, to do some research, to build their business. So thanks, Howie.
01:25You know, all I ask is you'd like and comment on this video. Show some love for Howie for doing such a cool thing. We need more entrepreneurs, more builders and it's I'm stoked to see him support you all.
01:36Thank you to Airtable for sponsoring this episode. You guys are legends. Enjoy the episode and have a creative day.
01:52Feeling really lucky right now because we've got Howie, he's the co founder and CEO of Airtable, and today we're gonna talk about agents. He's gonna do a little show and tell of his new product that I've been using for the last few weeks.
02:06But first, Howie, I have been I haven't been sleeping very much, to be honest.
02:12It's the agent psychosis.
02:14Yeah. Exactly. And I've I I just need your reaction to to just some things I've been thinking about.
02:20Yeah. So this chart over here is by Sequoia.
02:24In what domains are AI AI agents deployed? You can see software engineering's at almost 50%, back office at 9% Yeah. Marketing and copywriting, 4%, sales and CRM, 44.3% and down.
02:36When you see this, like, what's your reaction?
02:39I mean, I think two things. One is I think it absolutely reflects the under penetration of AI in industries that clearly could already be disrupted or benefit with even today's AI capabilities.
02:56If you took like frontier agents today and deployed them into every one of these categories, you should get to 100%. And then two, I think even the higher numbers like software engineering is actually kind of an overestimate.
03:08Meaning, as I think frontier developers and companies applying frontier agentic development practices are finding, the new model of software development is not even just every engineer using AI autocomplete, like TAB autocomplete, which like we all figured out like three years ago, right, with even GitHub Copilot.
03:25But it's now like, you don't even need the IDE. Right? Like, the the way I develop on HyperAgent is I have like 30 different Cloud Code instances running in parallel, and each one is coupled up to like a browser, fully autonomous.
03:37It can go and get other agents to comment on any PRs it creates. And so this modality shift of no AI to kind of what I would call gen one AI, which is basically AI augmentation for still very human driven development workflows.
03:55Andre Karpathy talked about in October, November is when he completely inverted from mostly still human written code with AI augmentation to completely the opposite.
04:06Right? And that's what we've seen, like, the frontier companies leap into. Like, I think even the 50% is an underestimate because the number of companies and even people who have switched into that new frontier mode is actually definitely less than 50% of software engineering today.
04:21So I think what we're actually seeing is the frontier is advancing so quickly. And many companies and many industries and many functions are barely catching up to, like, the three year ago state of the art, let alone, like, disrupting themselves and their industry with the new state of the art.
04:42Right. Well, I mean, another way to think about it is, like, there's copilot territory. These these charts are from Sequoia.
04:48Right? There's copilot territory. There's autopilot territory.
04:51Like, how do you see you know, you look at this. Right? Uh, this, you know, this is what Sequoia says.
04:57There's a there's a trillion dollars up for grabs within agents. Yeah. But they're very different.
05:02What's your reaction to this? I mean, look. I I think
05:06to me, it's like these agents really reached a breakthrough really, call it like four or five months ago.
05:13Right? And I think developers felt this with Opus.
05:17Opus 4.5 just kind of set a new high watermark of like, woah, this thing for the first time, like, really feels like a true software engineer that's able to work, like, on a task that would have taken a real human engineer, like, maybe many hours, if not days. It can go do it completely autonomously, and it ships me a perfect clean PR that I can just review like a reviewer would.
05:40Right? And I think that that experience is going to be unlocked and already is unlockable across every single other domain.
05:49Right? Because kind of just reached this point where, like, the models are more than smart enough. Right?
05:53Like, you talk to these models even in, a more synchronous, like, chat interaction, not like an autonomous agent interaction. And you like, you can ask it the most advanced things.
06:02Give it, like, really complicated subject matter content, right, like management consulting. You give it like, you know, kind of some some really hard meeting problems in the context thereof, and it gives you really smart answers that truly are like expert level.
06:15And so it's clear that the model intelligence is there. The models are smart enough also to kind of coherently execute across multiple terms with lots of tools and context. And so I think it's more of just a matter of how and how quickly we can deploy agents into every role in industry before we can, like, truly just almost do anything that humans could do in each of these functions with agents.
06:40And I mean, the TAM for that is, like, not even a trillion. It's, like, probably, like, the whole GDP of, like, all white collar labor, which is, like, obviously many tens of trillions. Right?
06:48Like, in in even, the Western Hemisphere alone.
06:51Right. Which is sort of like, I don't understand how you're not how people aren't motivated to create startups right now in that sense. Like, the the person listening to this is like, yes.
07:02Yes, Howie. You know? But it just feels like, you know, I can't think of a better time to be creating a startup than now.
07:08Totally. When this right? I think like I mean, yeah.
07:11I think the weird thing is like,
07:13it's almost like using is believing. Right? Like, it's really hard to fully grok the power here if you haven't actually gone and hands on spent, like, at least a full weekend playing with agents.
07:24Right? Like and that means more than just a superficial. Like, you did, like, some naive, like, one shot thing.
07:29Like, hey. Like, you know, who's gonna win the next presidential election? Like, kind of question that you could have asked a chatbot.
07:34Like, I think people are not actually coming in and when they're doing light experimentation, they're not actually putting in an ambitious enough prompt or task in front of the frontier agents.
07:45And they're still kind of using it like they use gen one chatbots. And until you actually experience the full power and autonomy of these frontier agents, I think it's hard to fully extrapolate what types of companies can be built now that were possible for structurally?
08:04How could you build, like, a multibillion revenue business with one human and, like, hundreds of agents? Right? Like, you have to use it to to get it.
08:13Also, you know, this is another chart I can't stop thinking about, which is the unit unit economics just absolutely crushed. When you look at a human a Yeah.
08:21Human person versus an AI agent and what it costs, like, you can create some serious gross margin businesses on top of this.
08:29A 100%.
08:30And this is the funny one because, I've seen kind of a lot of people complain about the cost per token of the frontier models, right?
08:39So like Opus 4.6, now seven, clearly the most expensive model, right? And then like GPV 5.4, very good, still kind of expensive. Even open source, it's cheaper, but it's not free, right?
08:53And I think people are Some people are struggling, I've seen, to adapt to this mental model of In the old days of software, like a lot of stuff was free.
09:02Like, you could get like I mean, even ChatchBT has a free version, right, that you could use however much you want. You get a cheap dumb model. But like, you're not expending that many tokens because it's not actually doing like autonomous multi turn work and expending like a billion tokens like every few days.
09:18Right? Like, it's much more token cheap or token token lean. And I think that, like, we have to get over this hump of like, you know, anchoring our price expectations for AI on like traditional subscription software where it's like, oh my god.
09:33I have to pay like $20 for Netflix per month now instead of whatever it was, $12.99 to 4. And instead think of this as like, yeah, to your point, how much would it have cost a human to do the thing? If I wanted to go and create an entire marketing campaign, um, or actually in my CEO role, like, it's funny.
09:53Like, one of our recent, uh, board memos that I wrote, uh, and sent out to our entire board and and kind of major investor list, like, a lot of it was researched and crafted by HyperAgent, right? Obviously, with, like, my kind of instincts and context and whatever imbued into the agent. Um, and of course, I I oversee it at the end.
10:12But like, I got feedback that that was the best memo from some of our best investors that I had ever written. And I'm like, yeah, because an agent did it. And by the way, I got to do it in 10 times less time.
10:23And so even if it cost me, let's call it $150 of tokens to generate that output, think about the opportunity that cost my time.
10:31And so I think that is a real reframe moment that's needed is let's think of this as like, what is the human equivalent time cost versus, wow, a $150, that sounds really expensive versus like a $10 per month sub.
10:46A 100%. Yeah. I think the way I always think about it is like, I anchor it around value.
10:51Right? What's the value I'm getting out of that? I mean, the truth is with your, you know, your board deck or whatever, like, it probably was the best of you know, it probably was the best because you had you had so much research support.
11:03Yeah.
11:04Totally.
11:06Two more quick graphs, and then I wanna get into HyperAgent. Percent of enterprise apps with embedded AI agents Mhmm.
11:17You know, this is the fastest adoption curve in enterprise history. Right? So, like, when you see this, you know, how do you react?
11:26I am not surprised. Mhmm. And I think even this reflects the pace at which, like, incumbents can even, like, integrate AI into their products.
11:36Right? I And think even that is, like, stimmied by, like, just incumbency and, like, you know, kinda how how seriously did enterprises, you know, enterprise apps or enterprise app makers or internal app teams, like, take this.
11:50I think the real show of how profound this growth curve is is like if you take the aggregate revenue created from from zero of all the leading AI companies, right, or companies like doing AI things, like take OpenAI and Anthropic alone.
12:04Right? Let's just say they have a combined revenue probably of like 80,000,000 plus, right?
12:09Or 80,000,000,000, sorry, plus right now. Up from like basically zero a few years ago. Like, in in the history of software, has there ever been an industry where any company, let alone or even in aggregate, across all the companies, you've got a category that went from 0 to 80,000,000,000 plus, right?
12:32And that's not even including all of the other AI providers, inference providers, and tooling, etcetera, out there. The revenue of, I think, the AI category is an even sharper curve.
12:46And I think that really reflects, like, just how profound this lightning in a bottle is. Totally. And just from an opportunity perspective, it's like selling
12:54to these enterprises and helping them figure it out and and just helping them transform is just a huge Totally.
13:04I think it's probably the one of the bigger cash grabs in business history is there's kind of two angles, I think, to create a very valuable business right now with AI as a wedge.
13:18One is PLG. And obviously, we see a lot of these PLG products that kind of put OpenClaw itself in this category because even though it's not actually a monetized business, it is getting this massive amount of adoption.
13:29And just the raw token consumption through OpenCLaw is, I'm sure, in the many hundreds of millions, if not billions already.
13:38And likewise, other products in the PLG genre. So that's one way. Just like let people use the AI thing that actually works.
13:45You're gonna get profound growth. But the other is like to come in top down, Palantir style. This is why OpenAI and Anthropic and, like, you know, the the big guys are also doing it.
13:53There's new companies as well going after this opportunity, which is go pitch to every enterprise board and CEO. Like, we will fix your AI problem. Pay us a massive check.
14:04Like, give us a $100,000,000 plus check, and we will purportedly solve your problems for you. Like, that is a existential risk mitigation that every large company incumbent should be willing to pay.
14:17Because frankly, the CEO's choice is like, either I pay it and I risk wasting a $100,000,000 and maybe getting fired over it. Or like, I don't do anything with AI, and I'm definitely getting fired over it.
14:28So on a game theory level, it's like everybody's gonna pay it. Now whether that actually results in long term substantial structural kind of transformation to the business that probably could be run now with, five people maybe instead of, like, 50,000, right, in some cases?
14:46That's a bigger question.
14:48Yeah. And and this this is you know, it sort of speaks to my my last point too, which is like if you can help a company, you know, run a fleet of 20 agents doing customer intel, content production, competitive research, lead enrichment, like all these different things like this is the future of work like in one image.
15:06Right? An agent command center. Right?
15:08Yeah. So when you see this, your reaction.
15:11I mean, look. That literally is a view in HyperAgent.
15:16I look I feel like I'm looking at HyperAgent, and I think this is the future. Right?
15:19Like, we are building towards a world where it may not be that every company is literally one person, right?
15:27And we have a lot of one person companies. But I do think every company will have a fleet of agents. And what's interesting to me is actually that agents are converging on these purposeful they almost map to job roles that humans were playing.
15:46And maybe it's a little bit like, why are robots, like hardware robots, converging on a humanoid form factor? And part of it is like, well, a lot of the infrastructure of everything we have in our homes, in construction sites, in factories are built for human ergonomics.
16:04So for the robot to effectively insert themselves seamlessly with the current infrastructure, they have to have human scale kind of capabilities.
16:16Right? And so I think there's a kind of very similar phenomenon happening with agents, which is it's not like I guess, like five years ago when people talked about superintelligence, I always imagined there's going to be just this single omnipotent AI that just figures everything out and looks at everything all at once.
16:35Like everything everywhere all at once. Right? And I think now I'm more and more of the belief that there are gonna be fundamental and always kind of present limitations on like context windows, for instance.
16:49Right? I just don't think we're ever gonna get to a point to where like a an AI model can like have infinite context window. And I think there's a physics to that.
16:58You can just literally only have so much attention and so much context at once. And I think what that means is that for the same reason why we partition humans into different roles and work structures so that not everyone in the company has to know everything and work on everything all at once, I think the same is true for agents.
17:19And so hence, you get this overview of agents that actually maps like, to kind of intuitive human played roles really well.
17:28And that's the really kind of interesting emergent phenomenon for me. You know, I just recently, like, spent some time playing around with paper clip, which is kind of fun because it literally creates the org chart metaphor.
17:40But I think this is really exciting, right, where it's in a way, it's both familiar because we're not just completely upending everything we knew about job functions and roles in the old world to the AI world.
17:55And yet, like, there is a rethink and reapplication of, like, okay. How do I play that content production role with an agent?
18:02Right. Well, I think we should get into HyperAgent. Let's do it.
18:06Now is the time. Right? So, for, you know, for the listener, like, what is HyperAgent?
18:11Why are you building it? And Yeah. This is a show and tell podcast, so, you know, by the end of this part of you know, by the end of this episode, like, you know, can you commit to, you know, giving all the sauce around how to use HyperAgent to to sort of build a business?
18:26Sure.
18:27Yeah. Let's let's go for it. So this is HyperAgent.
18:29I'm currently in a thread. I'll zoom out in a second and kind of show you what, like, the entry point looks like. But think of HyperAgent as if all of these other agent products out there, OpenCloud, etcetera, are kind of more like Linux.
18:43HyperAgent is our take on the Mac version of it. We want it to just work to be secure.
18:49It's cloud native. You don't have to run a Mac mini. And perhaps most importantly, HyperAgent is applying a lot of the same design philosophy and obsession with great UX that we applied to the no code app category ten years ago, but now to agents.
19:08Right? Meaning apps are kind of complicated.
19:11Right? If you're a developer, even at that time, could build a Rails app. You had a data layer, a logic layer, a view layer.
19:17But it was kinda technical. Right? And we're very technical.
19:20And the whole idea of Airtable was to distill that into a really intuitive experience. In fact, we were very inspired by, like, the Macintosh, the GUI, like, taking terminal based command line computing and making it into something that people could just grok immediately.
19:34And so HyperAgent is really intended to be a very intuitive and visual way of using agents.
19:42So this is actually a task thread that I ran a little bit earlier. And this is actually one of your startup ideas, Greg, that we had a hyper agent work on. And basically, the pitch was hyper local market reports for real estate agents generated from public data.
20:01And and so basically, this agent went around and did research on the landscape of the market.
20:09It ran a bunch of analysis. It's got full coding capability. It's got a full sandbox environment.
20:14So it is running a full computer. It's just one of the cloud, not like kind of your own computer. And you can connect it to all your accounts if you want.
20:22Like, it can access your Slack and granola and email. It can send stuff if you want it to on your behalf or just pre draft emails. You know, it's got, uh, already preconfigured ability to do things like pull from Twitter, use, uh, advanced tools like generate imagery or use Google Maps, etcetera.
20:39But basically, what happened was it went around. It did all of this.
20:43It researched the opportunity. Right?
20:46And then created this research brief. And let me just show you what this one looks like. This is kind of the business case for for the idea you pitched.
20:56Right? I kinda love it because, like, I actually think these what I would call medium sized markets, it's not like a $100,000,000,000 market, which is gonna be super competitive and there's gonna be massive incumbents going after it.
21:09But I really love this idea of the kind of like, maybe it's not micro. It's more like mini or medium market, like, couple billion TAM large, which is to say you can build a very lucrative business even capturing, like, a double digit percent chunk of this. Like, you can make a few 100,000,000 per year.
21:25And yet, like, it's small enough to where really big guys are not coming after it. Right? So, um, you know, this this, uh, this agent created kind of a business case for it.
21:34It found some really cool, um, like, user validation of the problem. So it's, you know, looked up Reddit, like, you know, and found, like, some real real estate people who are actually saying, like, I need this product.
21:46Right? So it's kinda validating the market need. Here's actually the current problem.
21:51Uh, I didn't even know about this, but, like, apparently, I guess there is some, like, legal thing that kind of changed kind of the dynamic of the market.
22:01People don't want more software, like another tool with an interface, and did, like, some competitive analysis. Here's who who who else is out there.
22:10And then kind of just put together the case for this. But then better yet, you don't just have to stop there. You can go and actually tell it to go and just build a v one of the product.
22:20So in this case, because HyperAgent has full coding capability, it just went ahead and, like, created a v one of this product, right, which I think this will actually work.
22:29Like, where do you farm? Like, here's my report style.
22:35It also looks really clean. What's that? Yeah.
22:38I mean and look. Honestly, a lot of this is just like if you have a good frontier agent running a frontier model, I. E.
22:45Like, Opus 4.7 or 5.4, like, it just does a lot of this really well out the box.
22:51So any frontier agent powered by frontier model should be able to create an app of this quality. What's unique about HyperAgent is that it can do that perfectly well, but then kind of do that in the in the workflow of like, it's not just an app builder.
23:05App building is just a feature now. It's a commoditized feature. And what it can actually do is go and research the end to end of like, here's actually the business context of what I'm trying to do and then build the app informed by it.
23:16Right? So it's more like HyperAgent is the founder in this case. It's not just the developer.
23:21It's the founder. Um, one of the cool thing I like about, uh, HyperAgent is like, it just comes out of the box with, like, really powerful tools. So it has, like, you know, Google Maps as a tool.
23:31And it can actually go and, like, let's say I think I already did this, but, um, like, I wanted it to go and actually find, like, real Street imagery of billboard locations.
23:42So it knows how to use Street View to, like, find actual points of interest. And then to take that image and use that as a reference seed image, uh, for, like, a, uh, AI image generation or video generation.
23:55Right? So, like, I mean, another cool thing you can do with HyperAgent, uh, is you could tell it, like, take this house and, like, I want you to redesign the house using interior photos from Zillow or, like, the exterior shots. And it will do that, like, really, really well.
24:09Right? So that's HyperAgent in a nutshell.
24:12I can walk through some of the other stuff here. Once you actually build like a lot of agents, then you get like this this ability to start looking at like, well, what if I wanted to see not just my one agent, sorry, but but an overview of all of my agents.
24:29Right? So this is not like a very built out account. This would be like your first week of HyperAgent use.
24:35But like literally that command center view that we talked about, like, you know, we want you to be able to create many different agents that each play a role. Here's the content marketer. Here's here's the market researcher.
24:44Here's, like, the, like, customer email responder. And, like, just manage and oversee an entire fleet of agents, constantly improve them because we actually have this ability to go and, like, curate memory and skill improvements from every run that you do.
25:01And then finally, to be able to deploy them into a team setting as well. So if you wanted to take any of these agents and actually give it the ability to talk in Slack.
25:10Right? So I can actually say, like, let me put this into Slack. Let me have it always on, always listening, in fact, and, you know, just sit there in my channels, listening to everything I'm talking about, my team's talking about.
25:21And when I have something relevant to add to automatically chime in and then people can interact with me truly like I'm a virtual coworker, right? And I think that's kind of part of the open clock experience I've seen some of the power users achieve.
25:35That's really quite magical. Like your Slack coworkers are now agents in addition to humans, and they're really smart, and they have their own, like, expertise and context. Like, you get that with a single click out of any agent that you build in HyperAgent.
25:48So you you mentioned skills. You know? How how does skills work on HyperAgent, and how how should people think about it?
25:55Yeah. So skills
25:57are, I think, like, the most important concept or primitive in the frontier agents world. Meaning, the models are generally intelligent enough.
26:06It's like, find like Albert Einstein, who's like obviously super smart in a general sense. And he may not know like how to solve problems in real estate. But if you gave him like just the right, like kind of briefing on like, here's a playbook.
26:19Here's a manual to learn everything you need to do to know to do this job in real estate. Like, he's gonna go and figure it out pretty well. Right?
26:25And so what's really powerful about skills is like they're a really, really composable concept. Like, you can interactively create skills. So let's say I'm actually gonna create, like, a new thread here to keep this super clean, but, like, help me create a skill that posts Greg Isenberg like AI content.
26:49Okay. And so what's really powerful about this is like No. Don't create this.
26:52Don't create No. No. Okay.
26:54But but worse enough that, you know, we don't take Greg's business. Exactly.
26:59But what's really cool about this is like, it's not gonna just like go and like say, okay, I'm just gonna have a prompt that pretends to be Greg Eisenberg. It could actually go and research how you actually do content.
27:14So it's coming up with a plan. The plan is like, I'm gonna first go and research your style, figure out, like, what platform I care about, like, look at some of your actual posts, and then distill all that into a skill that I can then pin to an agent or, like, just use on demand at any point.
27:30Right? So let's say, um, just for fun, uh, like, what, um, what platforms, uh, do you wanna post to?
27:37Let's just say x for now. Uh, we're gonna have the skill only generate drafts, it's not gonna auto post for you. Is there any kind of content you want your agent, Eisenberg, to
27:48to be focused on? Yeah. Let's do contrarian AI takes.
27:51Okay. Cool.
27:53And then any topics beyond that, like?
27:56Solopreneur, bootstrap, life And
28:00then how do you wanna use this agent if if you end up using this agent? Like, you know, do you wanna, like, start with an idea?
28:08Do you want it to just, like, come up with ideas for you?
28:11I don't wanna do it anymore.
28:13Right? Like, we'll go full autonomous. Right?
28:15Like, someday, we're gonna have to see if like real Greg is actually just sitting at the pool all day. Right. It's just created the the Greg avatar version of you and is doing everything on its own.
28:27But okay. So now it's like gonna go and like do some research about you and figure out like how to distill distill the perfect skill for for Greg, like, into this skill.
28:38How should people think about, HyperAgent versus Perplexity Computer versus Manus versus OpenCLaw itself?
28:48Yeah. So Codecs.
28:50Like Yeah. Yeah. How do you how do you see it?
28:52So I think
28:54against Codex, you know, it's quite simple. Like, HyperAgent is a more general purpose agent platform. Right?
28:59I think against open OpenClaw, like, this is much more turnkey, ready to go, safe and secure by default, cloud native, like, you know and and I think just much more focus on, like, great UX.
29:11Right? OpenCLaw. Like, we actually have to go to configuration or, like, if you're trying to edit memories or do any kind of curation or, like, kind of configuration, it's quite raw, right?
29:21It's like a very kind of raw product. Kind of feels like it's more for very technical people who've become expert at it. I think Perplexity Manus or Perplexity Computer and Manus are the closest comps for HyperAgent.
29:35The key difference is like, one, HyperAgent has more powerful tools out of the box. And also, it has more focus on UX out of the box.
29:46Right? Like, I spent some time playing with both of those products. I think they're great products.
29:50And you know, at their time and, you know, or at least when Manus first came out, it's truly groundbreaking. Right? Like, was the first kind of real, like, holy crap, like, yellow agent.
30:00Like, look at everything it did. Kind of like before even OpenClaw. Right?
30:04Long before OpenClaw. And so I think they were really kind of pioneers in this space. With HyperAgent, like, we've just taken a very UX focused approach.
30:12So for people who like seeing visually and be able to interact with the outputs and see more visually what the agent is doing and have a more visual way of defining skills, deploying skills, creating agents, etcetera.
30:28HyperAgent is just much more of the Macintosh experience versus the Linux. I think secondarily, we've also kind of done a lot more to make HyperAgent immediately ready to run not just like one, like, agent.
30:41Like, I think the nominal experience for Manus and Perplexity Computer is still like, you use those products and you kinda have this, like, agent that's pretty awesome. And, you know, you use it directly. Right?
30:52You can do that with HyperAgent. That's exactly what we're doing here. But it's also designed from day one with much more of, like, the scalability and deployability story in mind.
31:00So meaning, like, once I have an agent that kinda works for me, I can now deploy it one click into my Slack channel. And now everyone in my company can benefit from this agent just always on, like, kinda chiming into conversations. You know, they can ask it questions.
31:13They will respond. You have the command center, that fleet view, where it's not just one agent. You can oversee your entire fleet of multiple agents.
31:20And we even have things like, you know, the ability to oversee and curate, like, the learnings that that keep making each agent better. So, like, they kind of have this automatic self improvement loop where over time, they're accumulating not just new memories, but also suggesting to you, hey, maybe you should add this additional skill or update or tweak this skill.
31:39Or even maybe you should go and actually try changing my agent system prompt or give me access to different tools so I can do this type of job better. And best yet, like, we actually have this concept of what we call rubrics, which is exactly what it sounds like.
31:54It's like a eval rubric. And what you can do with rubrics that's really powerful is actually, like, define what does good look like for a certain type of task. Right?
32:02So I could create one here that's like, is a rubric for great Greg Eisenberg content? And what it basically does is I can then have a full eval loop where every time my agent runs, like once the Greg Isenberg skill is ready, I could say, like, I'm creating the virtual Greg agent, and I'm gonna pin a rubric to that agent that then says, every time Greg creates a piece of content, I wanna score that content along the dimensions that you care about using a separate L11's judge, um, that fires off.
32:31And then I can literally oversee, like, how well is my agent doing over time. Right?
32:36And if I wanna double click in and inspect any one task run to see, like, how did it get scored, I can do so. So we basically, you know, have this complete full loop of it's not just like you get a day one agent or thread experience that works really well out of the box. And it's not just like you can curate agents and deploy them and, like, improve them over time, but it's that you have this complete, um, observability layer and kind of this this orchestration story where you can actually just, like, look at all of your agents running all the time and see how they're doing.
33:08Um, and so if I pinned the the eval rubric to any one of these agents, I would see like the trend line of how it's scoring. I could then automatically like suggest, hey, maybe I can reduce the model quality.
33:21So I drop from Opus to Sonnet, get a five times reduction in cost, and the score didn't go much down. Right? So just once people actually start running agents at scale, these kind of secondary capabilities become really critical because it's not just about can I get one agent to do one thing, but how do I like oversee and run an entire business with many different agents and ensure consistent quality?
33:46Which is a big deal because, you know, for example, if you're using Manus, who is the judge around the output? The judge is you, a human being.
33:54Right? It's not Opus 4.6. It's like Exactly.
33:56Winning. So if you're trying to actually create what we were talking about before, which is like an agent first business, you know, managing a ton of agents, you're realistically,
34:06you're not gonna have the bandwidth to be looking at every single output Totally. At all stages. Right?
34:12Yeah. It's kind of like management one zero one, right, but like applied to agents now where it's like, as you scale up, if you're the CEO of a business, like, you just literally don't have time to go and, like, look at every single thing that every single person in the company has done. And so you need to create better automated checks and balances to oversee what the agents are doing, right, and inspect quality of work.
34:35This would be like if you actually had a giant army of human content creators, you would want some way of, in a scalable way, like, to detect, like, if they're posting good or bad content or not.
34:47Right? And then know, like, okay, we gotta tweak, like, the guidelines for each of these people. Okay.
34:52So now we have the Greg Isenberg contrarian draft skill. And I'm gonna go ahead and save this skill. And I'm gonna try seeing like, okay, let's do a dry run.
35:03It's gonna scan today's AI and news and trends and then create some contrarian drafts.
35:08Right? And the whole idea here is like, look, like, it's probably gonna do an okay job on like the the first effort here.
35:14Like, it did some research about you. It kind of like, you know, has a lot of like context about how you work.
35:20Right? And if I wanted to see more about this skill, I could actually open it up. Here's what it should be used for.
35:26Here's the actual kind of skill contents. Um, Greg's voice is a smart friend at dinner saying the quiet part out loud. Uh, not a corporate communicator.
35:34Would agree with that. Um, you know, you've been inside all these companies, uh, blah blah blah.
35:39Like, doesn't mean be a jerk. I I I I think it's very astute.
35:44Like, you're loud, but, like, not annoying or, like, you know, kind of rude. And then, actually, I'm curious if you agree with some of these stylistic things.
35:55Right? Like, gotta hook in the first seven words. You know, you don't want, like, long blocks of text, which I'm guilty of.
36:01So I I I should take some of this Greg Greg skill and apply it to myself. You love ordered lists.
36:08Never end with what do you think, which is super generic. So let's just say, like, this is a pretty good v one. Like, maybe it's, like, 50% of the way there.
36:18But the idea is that, like, these skills should be evergreen. Right? It's not like you do one and done.
36:24The whole point is every time I use this skill, either automatically using kind of the LLM generating learnings and suggestions to improve itself or because I am looking at the content and saying, oh, that's not quite right.
36:38Here's why you got that wrong. You can interactively tweak and improve the skills and performance of the agent over time.
36:45So I think this is the challenge that a lot of people face is they one shot something. It's not quite as profound as what they hoped for, and they kind of give up.
36:54Right? And I think like my kind of strong guiding and urgency to folks, and I think this is very aligned to how you've thought about it, is don't give up after the first shot.
37:05Right? Like, because it's very, very clear that the agents are powerful enough to do almost anything you want it to do. And the issue is not whether it's capable of and whether you should, like, give up on it.
37:16It's whether you are able to invest the kind of time and coaching and curation to get it there. And I think that it is well worth it, right?
37:25If you get it there, it's obviously going to be so much leverage for you that what's the value of like having an always on now employee that just like does the things that you care about, like behind the scenes at all times and like, you know, runs for trivial costs relative to like the the cost of hiring a new employee?
37:46Well, it's like real life too, which is like, you know, when I first started playing tennis, I was bad at playing tennis. And when I, you know, would go to play tennis, I I almost didn't wanna go because I was like, I'm bad at this.
38:00But you sort of you you go through the messy middle and you get better and better and over time, then you end up, wow, this is a lot of fun. So I think that once you get to the point where it's a lot of fun and and it does feel like the outputs are really good, The truth is 99% of people don't are not putting in the work to get to get the great outputs.
38:20Right? So, you know, the this is the arbitrage. It's for people to actually, you know, actually invest in spending time to optimize and get it to a place where it's high quality.
38:32Absolutely.
38:33Yeah. It's funny. One of the benchmark partners sent out this this memo about, like, you know, it was basically a wake up call to all of the the portfolio companies to, like, get with the program and really radically rethink how you operate your business immediately with AI.
38:48And the assumption is you're probably you think you're doing some or some things for AI. You have an AI kind of center of excellence.
38:58You have this AI feature, but it's not enough. Right?
39:01And the kind of parable that they ended with was like, imagine there's two friends back in, call it like, 2003.
39:14And they're both going door to door selling kind of knives. Right?
39:18Or or some other kind of in person kind of offline product. And one of them decides, every night and weekend, I'm gonna spend 30 minutes trying this new Google AdWords thing and trying to get some extra leads for my business, a supplementary.
39:35And one month, they grow a little bit of revenue from the SEO or the SEM thing. Next month, they get a little bit more.
39:42And the other person is like, this thing is awesome. SEM is awesome, and it's early, but I need to figure it out. And so they stopped going door to door and selling knives at all.
39:52And they just spend, like, the next few months, like, just focused on, like, how do I get this entirely Internet business to work, right, in the early days of of it? And two months, they have zero revenue.
40:05They're living off their savings. But they slowly start to get this thing to start get humming. And they get really versed in the best of SEO and SEM techniques and how do I create an e commerce kind of website that allows people to transact directly there versus just giving them a number to call me?
40:23And the end of the story is like, okay, project forward five years. Where do you think each of those people is?
40:30And the obvious answer is the second person has probably built one of the early multibillion dollar e commerce businesses and just carved off the next Amazon.
40:41And the other person is probably still selling door to door, which is getting harder and harder. And that market's shrinking.
40:47And so I think it is one of those things where it's like you kind of have to hit a reset moment. And what feels like maybe experimentation and not actually bringing home the bacon actually is the most profound thing you can do to create real business leverage in the not even two year time frame, but maybe even the six month time frame.
41:07And I'm curious in your experience or when you see solopreneurs doing this, where do you see or like how often like, what is the average breakeven point? Literally, either in terms of like, you get to the point where you can like self sustain a full time kind of like business, right?
41:24Like, and that becomes your paycheck or just even where it like even feels like it's starting to to pan out.
41:31I think that there's like multiple milestones that people hit where they you know, it's a game of confidence, you know. Yeah. When you make your first Internet dollar, no matter what it is, it rewires your brain.
41:43Yeah. So if you can take an idea and make $1 a stranger, just $1 Yeah.
41:50It's gonna rewire your brain. Then I think once you get to like 10 k a month, just something about that number, You know, for the most part, once you hit that, you're probably quitting your job, you're probably going all in, you're probably like, okay, there's something here and there's a path to something bigger.
42:06Totally. I think that with respect to like agent products and products like this, you know, the mistake I think a lot of people make is they try it too sporadically.
42:19So what I encourage people to do is to actually try the product, you know, every single day for a certain amount of time. So commit to thirty days, sixty days, ninety days, some amount of time so that every single day it's like in your calendar.
42:36Like, literally I have in my calendar like thirty minutes here, thirty minutes there. Right? Yeah.
42:41And that's what gets you to be a top 1% agent builder. Right? Yeah.
42:46Because it you make it a part of your workflow, and then you end up seeing, like, you know, outsized returns because it compounds.
42:54Makes sense. I mean, it's kinda like, you know, I'm not a writer, but like, I've heard from writer friends, like, the most important thing is not to like wait for like the one weekend where you're gonna have like the spurt of brilliance and write the whole screenplay or the whole book all in one get go.
43:07But it's like, you have to force yourself to write like some pages every single day. Like, no stopping.
43:13Like, of them are gonna be crappy pages, but like the forced habit, like, just gets you better and better and better. And then it becomes like natural. And so I could see that being very applicable and and kind of like analogous here for the world of like getting agent savvy.
43:27So do we have some tweets?
43:29So okay. Let's look at this. Let's see.
43:33The consensus narratives are oh, this is not loading for some reason. The consensus narratives are getting louder. Every Medium post reads like the last one.
43:42Okay. So here's here's one.
43:45The ten k month AI solopreneur boom is mostly content from fiction. They cite 82% of US businesses have zero employees. What do you think about this one?
43:55I mean, what I like about it is, you know, when I do tweets, because I'm a human being Yeah. Largely, there's no data. It's just like, I have a hot take.
44:04Yeah. So what's cool about this is there's research. The truth is,
44:09though, you know, people people obviously want data associated with their tweet. Yeah. Maybe with with a team of hyper agents doing all the research for you and, like, coming up with content ideas, now now you have time.
44:21Oh, this is kinda cool. Is this true that MedVee is actually not a legitimate business?
44:30I actually I hadn't I followed, like, the first arc of of that story, which is, oh my god. This thing is, like, so massive.
44:36But I mean, it's a it's a little let down for, like, the the billion dollar start up story. But, like, you know, maybe there's a take on it that says, like, no.
44:45But, like, it's still possible for real. This guy just kinda, like, gave us all a bad reputation. Your AI agents didn't replace your VA.
44:53Blah blah blah. It's kinda interesting. I mean,
44:57these are all what I would call, like, kernels
45:01Mhmm. For really good for great tweets. Yeah.
45:04Like PYR key. And and the cool thing is, like, I could give it feedback. So, like, you know, as an example, like, let's let's say, like, I want to give you feedback on your skill.
45:15What's, like, one thing that you wanna, like, give it some feedback on?
45:20I would say, you know, the the tweets that tend to do well are sound sound very friend to friend.
45:30And is there like, do these all just feel like a little too, like like, they're not, like, colloquial? They feel Exactly. Yeah.
45:36These feel a little too formal or, like, stiff or something.
45:44Don't know. Exactly.
45:46Yeah. And that's something, like, I would notice that. Right?
45:49Yeah. And so what we can do
45:53like, we would put this in the eval. Right? Yeah.
45:56You could do both. So one is, like, you could immediately go and turn this, like or update the skill based on this feedback. You could also have it immediately just, like, turn around, like, a new draft of these tweets, right, to sound more colloquial.
46:09And then finally, to your point, I could go and create a rubric that actually says, like, okay. Like, here's the five dimensions I care about And then auto evaluate every future output.
46:19Right? Um, so you kinda have a number of different options, like, depending on how far you wanna go right now. Like, if you just wanna get your job done right now, you don't wanna bother with Rubric.
46:28You don't have to. Right? But eventually, like, you get to the point where you wanna set up a scalable system for this to just constantly work and get better and better.
46:36And that's the point at which you would do a Rubrik, which is not that hard actually. Like, if you know, you can either go in through the UI and build one, or you can actually in this chat, like, say, help me build a rubric to score great Greg style content, which I'll queue up for after it updates the skill.
46:52Um, and and then it will go and help me create that rubric, save it, pin it to this agent or to this skill, and then automatically run every future time I create content.
47:05And is it possible to, for example, get an email every single day at 8AM with, you know, some ideas?
47:14Like You yeah. You absolutely can. So the way to do that would be in fact, you could just tell it in the thread, like, can you turn this into a recurring daily email at 8AM?
47:26And so then what it's gonna do is say, I wanna now save this thread into an agent. And the agent is gonna be given a run schedule of every day, 8AM, go and do this thing. We're actually about to ship something that we're calling a live mode, which is kind of inspired by the open claw, like, kinda heartbeat behavior where you could already have configured an agent to do this just by saying, like, I wanted to pull every thirty minutes.
47:48But we're making it much more of a first class thing within HyperAgent where you can literally just click a button, turn any agent or any thread alive. And then the feeling is gonna be that like, wow. This thing is just like constantly on and looking at all of the, like, new tweets out there, coming up with new ideas, and then pushing them to me either via Telegram or over email or in Slack whenever it comes up with new stuff.
48:11So like the UX or the mental model is meant to be like, wow. This just becomes like a always on, like twenty four seven agent that that pushes ideas to me, or even, like, can go and, like, preemptively draft and post content.
48:27Like, if you wanted it to go full YOLO, you could actually have it just go and, like, tweak the content itself. Right?
48:33Good old full YOLO mode. Yeah. Yeah.
48:37I don't recommend full YOLO mode just because, I mean, there's no need for for for something like this.
48:47Right? Like, in order for X specifically, in order to win, if you can get one good tweet out every single day, that's all it is. Yeah.
48:55No one, you know, and that just means that you could and you can batch these.
49:02You can schedule it up, but just look at it and make sure that that it's it's high quality, meets your bar. Yeah.
49:10I think I think it's it's definitely worth it for for this specifically. Yeah. That's fair.
49:15I mean, I think that it know, content is a very hits driven business. And so, like, fewer high quality hits is is what matters.
49:23But, you know, there are there are tons of use cases where, like, maybe for my own emails. Right? Like, they're a subset of emails that, like, are low stakes that I just, you know, want HyperAgent to just automatically not only draft a reply, but, like, if it feels confident, it's, like, not a sensitive, you know, kinda situation.
49:42Like, then just go ahead and respond to it. It could be simple inbound inquiries from internal folks saying, hey, when you have time to meet, you can just preemptively go ahead and suggest a time, or even prebook it on my calendar.
49:58Or customer emails that are, like, innocuous or, like, asking for, like or trying to give input on a feature. You could just compile all that feedback for me as a report, but then respond, like, with a smart personalized acknowledgment to the user or even ask for, like, clarification.
50:15And I think
50:16you you all have, like, a ton of connectors built into HyperAgent. Right? Yeah.
50:21So what's really cool actually is that not only do we have a ton of connectors that just work out of the box, you click a button OAuth in in the thread. Right? So maybe starting a new one, I could say, like, what's a tool that you wanna use with with HyperEngine?
50:35Could be like Winola. Or Notion maybe? Okay.
50:38Yeah. Connect can I connect to Notion and pull in all my notes?
50:45And so it will just in the thread, like, say, hey. Here here's an OAuth link. Like, connect to your Notion.
50:52But arguably, of the most powerful parts is, like, even for things that we don't have a connector to, like, let's say there's some, like, very obscure API that you're trying to work with. Right? Um, you could basically have HyperAgent go and learn that API.
51:08So I actually, I'll say, like, actually, never mind on this. Can you instead help me build an API integration too?
51:17What's some, like, fairly new tool that you know of that has an API?
51:22I'm assuming well, do you have Linear built in here?
51:25We do have a connection to Linear, but actually, maybe maybe Twilio could be a good example. Right? Like, where Okay.
51:30I don't if you can OAuth into Twilio, so it has to be an API skill. And we may have a prebuilt connector, but I'm I'm gonna have it, like, build a custom skill regardless.
51:39So can you still help me build a custom skill to integrate with Twilio via API? Right?
51:46And so now what it's gonna hap what what's gonna happen is, like, it can go and, like, research the Twilio API docs, create a skill for itself to use the API, and then actually ask me to enter my credentials in a safe way and then be able to like use the Twilio API fully.
52:03Right?
52:06So I think the powerful thing now is a Frontier agent should be able to literally do anything.
52:14But it's just a matter of you have to give it access to the right context. And you have to tell it like, hey.
52:21Like, yeah. You should build a skill for this. So then it can do it every single future time effortlessly.
52:27Okay. Let's say, like, what we wanna do, SMS, voice for now, maybe phone numbers. We'll do an API key off and any specific workflows.
52:38Think like maybe actually, I wanna build a voice and SMS service that can call restaurants for reservations
52:52or something. Right? If you're listening to this and you're not fired up about building a business right now, like, the fact that you can do this is crazy.
53:02Yeah. If someone has heard about you know, this is the first time they're hearing about HyperAgent.
53:08Yeah. They wanna they wanna get started and they you know, what's a plan for them to like, what should they do?
53:14How do they get started? How do they get the most out of HyperAgent?
53:18I I think, like, the most often, like, the hardest thing to get over is not, like, how to use the product.
53:25Like, I think, you know, our users have said, like, wow. This product is, like, super intuitive. Like, I can usually just, like, ask the agent to figure something out, and a dope goes and does it.
53:33So it's not like I have to learn, like, a ton of new, like, configuration or UI or anything. I think the hardest part is actually, like, picking, like, the right problem or, like, the right business opportunity you want to try to attack with HyperAgent, which, like, HyperAgent actually can help you brainstorm that. In fact, we just shipped a new better onboarding flow where instead of just, like, landing you into a generic, you know, kind of, like, empty canvas where you have to, like, just pick, like, a new thread.
54:01And, you know, we have some, like, templates and so on. Like, now when you first land in, it's going to suggest like, hey. Do you wanna, like, connect me to all of your contacts?
54:08So, like, connect me to your, uh, Gmail, and to your Slack, and to, like, your Notion and Granolah. And what I'll offer to do is actually go and research you in your context.
54:20So I'll read through a bunch of your past week's emails and Slacks and look at your past Granolah meetings. And of course, all that context is private to you.
54:29But now HyperAgent is gonna be able to suggest to you, hey, based on everything I've learned about you, here's some use cases that might be relevant to you. So it seems like you're a VC. Maybe you're doing a lot of deal flow.
54:41I could create an agent to just go and automatically, like, you know, kinda summarize and do research on every investment pitch that you get. Right? So, like, you can turn me on all the time.
54:52Like, I'll just run-in the background and then, like, ping you every single time you get a inbound pitch. You can even have it learn the behavior to thread a private reply to any email that you get inbound from a founder.
55:06Right? So you get an inbound pitch. HyperAgent, on behalf of you, sends you and only you a just threaded reply within that email chain saying, hey, I researched this company.
55:15I also summarized all the materials. Here's what you you should know about them. Right?
55:18But the whole idea is that HyperAgent itself can help you identify use cases. Or you could come in just with a really broad prompt, like, kinda interested in building building a solopreneur business.
55:30I don't know. I'm kind of interested in, like, real estate. I wanna pick one of Greg's, kind of ideas that are open source.
55:36Help me plan this out. Right? And it will do a very good job of going and running with you on that.
55:40So I think the main thing is don't get stuck in the blank slate starting point problem. Just come in and figure out some place to start.
55:48Maybe it's your personal contacts. Maybe it's you come in with an idea.
55:52But once you start getting into it, it just sucks you in even more because you realize all of what you can do.
56:01And it's just so powerful. You won't help but to get better and better at it.
56:06Last question before we head out. I was just talking to someone on another pod on actually this podcast talking about Yeah.
56:16Her message agent. And one of the things we're talking about is when you're picking one of these platforms, be it OpenClaw, HyperAgent, Codex, whatever, you're sort of like investing in an ecosystem.
56:30My question for you, Howie, is why should someone, you know, invest in the HyperAgent ecosystem?
56:38Like, where do you see HyperAgent going over the next few years?
56:42Yeah. So we have a lot of experience building great PLG products.
56:48I mean, obviously, Airtable itself is a PLG product that also scaled up into real serious kind of like businesses. Right? Like, there are companies that still run their major operations, whether it's like really, really large, like, you know, kind of Walmart scale companies, like, you know, the opening eyes of the world.
57:07But also, like, you know, we have, like, like, really innovative, fast moving SMBs, some of the, like, fastest growing companies like Recor, run a lot of, like, stuff on Airtable. And, you know, I think, like, the the the experience that we have of building a product that's both extremely low floor and intuitive, but then also has a very high ceiling and scales up even as you need to scale up the number of agents you have, how you deploy them, um, how you oversee them.
57:33Like, that's our commitment is that we are going to be the best at giving you both a low floor and a high ceiling, especially as you want to actually run a serious business or operations with HyperAgent. Right?
57:45So I think that's that's gonna be kind of unique where I see the landscape fragmenting into, like, there's gonna be really easy, fun kind of prototyping tools and products that are kind of, like, easy to get started with.
57:57But then ultimately, don't scale with you as you wanna become, like, a real serious enterprise built around these agents.
58:04And then conversely, there's gonna be more, like, heavy kind of agent builder products, right, with like configuration and like controls and all that stuff that are gonna be better from like a control plane standpoint from be able to like oversee a fleet of agent standpoint, but make the initial experience and the graduation path a lot more clunky, right, or just a really sharp wall to overcome.
58:28So I think our commitment is this product is going to be the best combination of low floor and high ceiling. And we're always gonna have this obsession with great UX.
58:38Like, that's our DNA. That's like what I obsess over. And the only kind of company that I want to build is one that wins in a product category where the value of the software or the technology is very, very high, but the accessibility is really kind of the key differentiator that we win on, right?
58:56So agents are going to be powerful. We're not going to be the only powerful agent product out there. Like, I think Frontier agents are all going to get better and smarter and faster and and so on.
59:07But what we can do is use really great product design just like Apple did with computing to make the powerful experience also really accessible. Yeah.
59:16It really is the most
59:18HyperAgent's the most visual agent builder I've ever seen. It reminds me of a a desk. Like when, you know, I'm looking at my desk.
59:27It's a wood desk right now and I've got I'm like, I have a paper over here and some scribbles over here and my iPad over there. To me, that's what HyperAgent kind of feels and looks like. It feels like a desk that I'm like visualizing it.
59:41So I think for people who like, you know, connect like that, and I'm I'm certainly one of those people, I think a lot of people are just gonna be like, sign me up. Totally.
59:51Yeah. I mean, look, like, know, for people who don't like UI and wanna just, like,
59:56use their computer through the terminal, like, all day every day, Like Well, some people Yeah.
1:00:01Howie, some people are like, they're they're, you know, they're what they love doing is like obsessing over tuning every single detail and stuff like that.
1:00:12And they're those people, the you know,
1:00:15that an OpenCLaw might be for them. Right? If you if you are more Like yeah.
1:00:20That but I believe that, like, you don't have to sacrifice the tunability, right, or the, like, the power. And so, you know, one of our strong design philosophies here is that, like, HyperAgent still does give you a lot of control.
1:00:33Like, you can go and tweak, you know, kind of like agent configuration if you want to. If you wanna, like, choose the exact model and system prompt and tools and, like, give it a lot of refinement, you can. And, like, you can go quite far in terms of curating memories.
1:00:46Uh, we actually just shipped yesterday a kind of like a defrag tool for your memory so that as you accumulate more and more memories across all these different agents, you have this, like, really elegant way of defragging them.
1:00:58Right? Where we can auto suggest here related memories clustered by both keyword as well as embeddings similarity so that we're actually understanding the content of the memories You can consolidate them.
1:01:10But they're like you know, we wanna really serve both people who are like power users who want control over how the agent is set up so they can get maximum bleeding edge performance. But then also, you know, like, shouldn't have to do all that to get value out of the product. So it really is about the range.
1:01:27I think it's more just that if you are truly happy just doing it all yourself through a very low level command line interface kind of experience and you're okay not having the control plane, like the deployability, the ability to oversee many agents and deploy them at scale and manage across a team, then, you know, may may those people, like, aren't gonna appreciate HyperAgent as much.
1:01:53Totally. Well, I'm stoked to see how it evolves. Thanks for doing a little show and tell.
1:01:58Got me fired up. Oh, thank you. How we I'll include links where to follow you, but also where to sign up to HyperAgent in the description, in the show notes.
1:02:10Yeah.
1:02:11And we're gonna do a really generous credits giveaway for your listeners. I mean, of the benefits of launching HyperAgent within Airtable, which is a half billion revenue business, we're gonna generate 100,000,000 of free cash flow this year.
1:02:27Like we have over $1,000,000,000 on our balance sheet. That's not to just be pretentious about it.
1:02:32But is that we've built a good and growing and profitable business with Airtable that allows us to be even more generous and liberal with, we just wanna get people to really adopt HyperAgent, get value out of it.
1:02:49And we want it to come to standard, right? Like, we want it to become like the iPhone. And so we're willing to be very, very generous.
1:02:56Like, we're not trying to make money and nickel and dime people on pricing. In fact, like, we're giving away multipliers to your audience and early adopters for both just straight up cash that gets applied towards real model costs, including like Opus, which now, as a lot of the OpenCLaw community has gotten kind of sad about, like you can't get subsidized credit for use in OpenCLaw.
1:03:23But you can use Opus. You can get the Frontier models. And you can get it much more cheaply because we're willing to subsidize it through HyperAgent.
1:03:32Well, we this is a group of people who are listening to this who appreciate that because this is a group of people who actually you know, they listen and then they actually go and build stuff. So Yeah.
1:03:43Thanks for the love, Howie. And you have to you have to I love this
1:03:49the the solopreneur and, like, you know, small early stage, like, startup and small business owner audience.
1:03:56I think it is where more AI innovation is gonna happen far faster than, frankly, within many large kind of incumbent companies, right? You just have the agility.
1:04:07And the only thing keeping you from going and deploying agents everywhere is like just your willingness and like putting in a little bit of time. Right? But we're already seeing in our early adoption base with HyperAgent, like, some of these like small shops have become super sophisticated really, really fast and are running their operations in a kind of game changing way that, frankly, a 50,000 person company would not be able to do for a much, much, much longer time.
1:04:34Right? And just has all kinds of kind of reasons why they wouldn't be able to go and pivot on a dime. So I think this is a really, really awesome audience.
1:04:43And I kind of live to see entrepreneurs do awesome stuff.
1:04:48Right? So super exciting to to be plugged into to the community.
1:04:52And, like, I wanna see, you know, your listener base generate, like, you know, a $100,000,000,000, you know, kind of legit companies with, like, less than five employees.
1:05:04Your lips to God's ears, baby. Thanks a lot, Howie. I'll see you next time.
1:05:09Awesome. See you.
The Hook
The bait, then the rug-pull.
A trillion-dollar framing is still too small. Howie Liu, the co-founder who built Airtable into a half-billion-dollar revenue business, sits down to argue that the real addressable market for AI agents is the entire white-collar labor supply of the Western hemisphere — and that the capability unlock already happened, quietly, about five months ago.
How they asked for the click.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
OTHER LINKSAlso linked in the description.
- If you want more workflows and tactics to build a business with AI, check out this free workshop ↗
- If you want more workflows and tactics to build a business with AI, check out this free workshop ↗
- latecheckout.agency ↗
- The Vibe Marketer - Resources for people into vibe marketing/marketing with AI ↗
- X/Twitter ↗
- Instagram ↗
- LinkedIn ↗
- X/Twitter ↗
- Hyperagent ↗
- Airtable ↗
