LongCut logo

Clawdbot/OpenClaw Clearly Explained (and how to use it)

By Greg Isenberg

Summary

Topics Covered

  • AI Employees Self-Improve Overnight
  • Proactive Prompts Unlock Unknown Workflows
  • Hunt AI's Unknown Unknowns
  • Use Cheap Models as AI Muscles
  • Reframe AI Costs as Employee Investments

Full Transcript

This episode is the clearest explanation of Claudebot on the internet and how you can use it to make money and be more productive. Claudebot feels like hiring

productive. Claudebot feels like hiring a digital operator who works around the clock and actually ships. Once you see it in action, it changes how you think about building, how you think about

delegating, and how you think about scaling. In this episode, we break down

scaling. In this episode, we break down real use cases, how to get started, the risk, and why this is becoming a serious leverage tool for solopreneurs and founders. [music]

founders. [music] This just in, Claudebot has officially been renamed as Maltbot. I'm guessing

they got a lot of heat from the anthropic team, so they must have changed their name. So, it's

unofficially Cloudbot, officially Moltbot. If Claudebot has been on your

Moltbot. If Claudebot has been on your radar and you just want a clear explanation of how you can use this as a solopreneur, as a founder, this episode is for you.

[music] >> I've been waiting for this moment. Alex,

Mr. Claudebot Finn has come on the podcast and by the end of this episode, Alex, what are we going to learn? By the

end of this episode, if you listen attently and closely, you will have your own 247 AI employee working for you at all times. But I've seen I've seen

all times. But I've seen I've seen videos where that where that's the promise. Can you can you can you commit?

promise. Can you can you can you commit?

Can you promise to us that you're going to showcase use cases that people haven't seen before? They haven't seen and that by the end of this episode, they'll be able to implement it. You are

going to have an AI employee that's tracking trends for you, building you product, delivering you news, creating you content, running your whole business. This isn't this isn't your

business. This isn't this isn't your 101. It's going to be checking your

101. It's going to be checking your email letting you know when you get an email. This is going to be you're going

email. This is going to be you're going to be running a business by yourself with AI employees.

>> Okay? So, it's for people who want to make money and be more productive using Claudebot. It's for people who want to

Claudebot. It's for people who want to actually improve their life, get more productivity, and not just kind of have a Tamagotchi toy.

>> All right, let's get into it.

>> Let's do it. So, I have been playing around with this pretty much non-stop for a week now. Uh I don't think I've slept. This is the most excited I've

slept. This is the most excited I've been about technology probably since the first time I used Chat GBT, if not in my entire life. uh

because I really do believe this unlocks a next level for solopreneurs. Probably

a lot of the people watching this video right now and then just people who want to get things done. Uh I am a oneperson startup. You know, I have my own SAS

startup. You know, I have my own SAS that I uh built and run completely by myself. You know, I'm a YouTube creator.

myself. You know, I'm a YouTube creator.

I'm an ex creator. I have a newsletter.

I do a hundred things at once. I work

from beginning to end of the day. And

when I got my hands on Claudebot and when I started trying it, like this isn't hype. If you use this the right

isn't hype. If you use this the right way, you're actually going to get insane productivity, especially if you're someone in kind of my position, the the one person startup. And so, let me give

you an example. I think I'll give you an example right here that will kind of blow a lot of people's minds. I don't

think a lot of people are doing something like this at the moment.

So I for instance every single day get a morning brief from my Claudebot Henry from moving forward I will use Henry instead of Claudebot uh because I treat

my Claudebot with respect and use its name. So let me share first of all my

name. So let me share first of all my Telegram chat. I use Telegram to

Telegram chat. I use Telegram to communicate with Henry. That's one of the other kind of mind-blowing things about this, I think, is the fact that you're interfacing just kind of in

messaging app on your phone. But I get a morning brief. And we did a lot of setup

morning brief. And we did a lot of setup to get to this point. And I'll go through that setup. I just want to kind of show the power of what we'll be going through here. I get this morning brief.

through here. I get this morning brief.

And every night while I'm sleeping, Henry does many things for me. First of

all, gives me the weather. That's kind

of nice. But I also have him doing a lot of work while I sleep. I have him researching projects that I've talked about. One of the most amazing parts

about. One of the most amazing parts about Claudebot is it is self-improving.

Constantly self-improving. Every single

thing you tell it, it remembers and in includes in further conversation, future conversation. Right? So, for instance, I

conversation. Right? So, for instance, I talked about the fact that I am buying a Mac Studio to run it on in the next couple weeks. And so it started going

couple weeks. And so it started going and it started looking at different ways to run local models on a Mac Studio overnight while I was sleeping without me asking and it created an entire

report for that. It came up with a content repurposing skill because I told it I have a newsletter. I told I do YouTube X whole bunch of things. So it

came up didn't I didn't ask for this created a content repurposing skill for me so I can easily repurpose my content.

Right? It's just improving itself. The

most mind-blowing part I think of what it did for me over the last few days is it kept an eye on X. It found if you've been paying attention to X, you know

Elon's been talking about giving a million dollars away to the top article, right? It actually found, you know, that

right? It actually found, you know, that this this story articles are popping on X. Elon's giving away a million dollars

X. Elon's giving away a million dollars and it actually built out article functionality for me in my SAS creator buddy. Right? So I have this app creator

buddy. Right? So I have this app creator buddy. It's all about X content and so

buddy. It's all about X content and so it actually saw on X that articles are popping off million dollars and it built out some article writing functionality

in Creator Buddy for me. I woke up, it said, "Hey, I built out this functionality in Creator Buddy that I think would be helpful based on what's trending. Check it out. Let me know what

trending. Check it out. Let me know what you think." And so now I have this

you think." And so now I have this employee that is just every night while I'm sleeping checking what's trending, what's going on, building me little demos and showcases, building new skills

based on our conversation from the past day. And then I wake up and I just got

day. And then I wake up and I just got to approve things. And so I got this article writer functionality, built it out. It created a pull request.

So it didn't just push this live on the internet. That'd be insane. You don't

internet. That'd be insane. You don't

want to do that just yet. Maybe one day soon. Not just yet. Took the pull

soon. Not just yet. Took the pull request, tested it. Looked great. Works

brilliantly. You can write articles.

Looked pushed it to live. And just like that, I have new functionality in Creator Buddy based on what's going on.

And like this is the like this right here would have took me hours. That's

hours saved just like that from my Clawbot being proactive and figuring out what I'm interested in.

>> Well, it saves you hours and maybe you you pro you might you okay, this is a great idea. Maybe you would have come up

great idea. Maybe you would have come up with this idea, but maybe you wouldn't have come up with this idea. You know

what I mean, >> right? And and that's one of the

>> right? And and that's one of the brilliant pieces of of leverage here is this is super intelligence, right? this

is able to think a lot more a lot quickly and has a lot more context than my brain has and yeah I might not have came up with this idea and luckily it did and now it's live and my apps and my

business are improving literally while I sleep and I'm just a oneperson business I'm not hiring people this is my employee and it was able to get this done which is really incredible

>> what what okay a naysayer is watching this or listening to this what are they saying when they're hearing and watching this >> so I actually get this a lot. I get a lot of naysayers. I get a lot of people

going, "This is BS. You're you're making this all up. This this this is not BS."

So, here's how you This is all about setup. You don't just turn this on and

setup. You don't just turn this on and it starts building you SAS. There's a

lot of setup that goes into getting this type of functionality. I mean, there's a lot of other I'm demoing a lot of other things that it built out for me

completely autonomously. Um, the setup

completely autonomously. Um, the setup of your Claudebot is critical.

What you need to do because its memory is so strong is you need to go in and you need to make sure it knows as much about you as humanly possible, right?

You need to make sure if you have a YouTube channel, you put the link to your YouTube channel. You talk about what you create content on. You talk

about your hobbies, your interests, every part of your business, what your goals and aspirations are, right? What

what your relationship status is. like

literally as much as you possibly can, you want to give to it because it's going to remember all of it in every conversation. Then just like you would a

conversation. Then just like you would a human being, you want to set expectations, right? So like you, if

expectations, right? So like you, if you've ever had an employee before, anyone watching this, you set expectations with your employee when they first get hired on like what you want that working relationship to be

like. And so the big way I got all of

like. And so the big way I got all of this is I set the expectation with my clawbot that I want a proactive

relationship where I don't need to give it all its commands. It can just do things without me. And so I have a prompt. Greg, I will send this prompt to

prompt. Greg, I will send this prompt to you uh after this so that you can include in the description if you'd like. You want to give this prompt. Uh

like. You want to give this prompt. Uh

the the UI is not great with this here.

I'll read it out though really quick so people can have an idea of how it works.

Uh but when I started my relationship during the onboarding it's like okay you know what should I know about you? I put

this in but everyone watching you can just put this in even if you've already onboarded. I am a one-man business. I at

onboarded. I am a one-man business. I at

the moment uh let me see here I have a little typo in here. I from oh I no I don't. I work from the moment I wake up

don't. I work from the moment I wake up to the moment I go to sleep. I need an employee taking as much off my plate in being as proactive as possible. Please

take everything you know about me and just do work you think would make my life easier or improve my business and make me money. I want to wake up every morning and be like, "Wow, you got a lot done while I was sleeping." Don't be

afraid to monitor my business and build things that would help improve our workflow. Just create PRs for me to

workflow. Just create PRs for me to review. Don't push anything live. I'll

review. Don't push anything live. I'll

test and commit. And what this is doing is this is setting the expectations for your working relationship with your Cloudbot. Right? You want to treat this

Cloudbot. Right? You want to treat this with respect as a human being, right? As

you would treat a human being. And from

there, when I said it this, it was like, okay, I want to be able to update you on things. Let's do a morning brief. And it

things. Let's do a morning brief. And it

started sending me the morning briefs.

And I got a morning brief like, oh, can you also research my YouTube competitors? And I gave it a list like

competitors? And I gave it a list like 20 channels. And then it started

20 channels. And then it started including like, oh, this channel, it actually was really interesting. I think

I have an example I can show you right here.

It actually uh started letting me know whenever other channels had content that performed well. Let me see if I have one

performed well. Let me see if I have one of my morning briefs up that has this.

Yeah, here we go.

So, it was like, "Oh, you want me to do uh competitor research for you? I know

what I'll do.

I will see when someone posts a video that outperforms their normal videos and let you know in the morning brief. And

you can see one of my morning briefs from a couple days ago, Nate B. Jones,

who's another AI channel, posted a video that did better than it normally does on his channel, right? And so it's little things like this where you just want to

set expectations for your relationship, give it as much context about you as humanly possible.

And this is really important. Interview

it, right? Interview it as much as you can where I'm like, I'm a YouTube creator. What can you do for me? What

creator. What can you do for me? What

what what what task can you do? How can

you make my life easier? And it sat there and it thought and it came up with like 10 different things. And one of them was I'll research your competitors and see when they have outlier videos.

And now that's included in my morning brief. I think the issue most people

brief. I think the issue most people have when they use not only Claudebot but any AI tool at all is they don't

hunt the unknown unknowns. And what I mean by that is these AIs have unbelievable power, can do unbelievable things, but we only ever ask it to do

things we think of. You want to spend a lot of time saying, "Hey, here's everything about me. What can you do for me?" and find those unknown unknowns and

me?" and find those unknown unknowns and that's how you unlock workflows like this which are unbelievable and have seriously saved me hours and hours and hours of my life in a week.

>> What else uh what else can you show?

>> I got a good amount. So, basically every night it's been building different apps to improve our workflow. One of my

favorite things it's been doing is it's actually been uh track it built a project management tool. Um so let me pull this up here. Here we go. Uh it

built this and actually before I show this let me give one more tip before I go deeper into kind of the tech and building side. Claude Opus is the best

building side. Claude Opus is the best model ever made. Period. It is fantastic for Claudebot. One issue is even if

for Claudebot. One issue is even if you're on the $200 plan, you're going to hit your limits if you just use it for everything. So, what I'm about to show

everything. So, what I'm about to show you, I'm about to show you a few things that Henry built for me. You want to make sure when you ask your Clawbot to, hey, be proactive, build things for me

overnight, build me, code me stuff.

You want to use the right muscles. And

what I mean by that is I think of opus as the brain. You want to use other models as the muscles. Codeex is a really good muscle for coding for this.

And so set up codecs inside your Cloudbot. Say, "Hey, whatever your name

Cloudbot. Say, "Hey, whatever your name of your Cloudbot is, use only use codecs from here on out for building." And what that's going to do is save you a ton of usage on your cloud so you can use it

the whole month, never hit limits, and also use other models to build other things out and make it more efficient.

So that's just a side tip before I go into this next kind of step in my workflow. But just an example of

workflow. But just an example of something it built for me is this product management uh tool which it calls mission control. It named it that

mission control where it actually tracks all the tasks it does for me.

Uh and as it does them moves them along this board, right? And so I can wake up in the morning. I get the morning brief which gives me kind of rundown of things it did. But I can also go into the

it did. But I can also go into the activity here over on the right hand side and see all the tasks it completed for me recently. And so this is kind of like our tracker system where if you use

cloud but all you know it's all it's just one chat. You don't have multiple conversations as the chat gets older.

You can't really scroll back to see what was said. This is your way of tracking

was said. This is your way of tracking everything it's done uh in in perpetuity.

>> So is this something that you asked it to build or is this come stock? like how

did how does this happen?

>> This built it. So, kind of with my prompt where I was like, "Be proactive.

Figure out ways to improve our workflow." One morning, I woke up and it

workflow." One morning, I woke up and it basically had the V1 of this where it kind of looked like a vibecoded cananband board you would see in any sort of tutorial video online, but it

built it built a canband board with the goal of me tracking the tasks it's doing. And then I massaged it a little

doing. And then I massaged it a little bit after, but 80% of this was built autonomously while I slept and I woke up to it. [sighs]

to it. [sighs] >> That's crazy. [laughter]

>> Yeah, it it's it's I think what excites me the most about Claudebot out of all this is it's like

it has unlimited potential. This this is an open-source harness. This isn't like a model. This isn't a kind of clawed

a model. This isn't a kind of clawed co-work where there's guard rails protecting everything.

This is no guard rails, just a harness for an AI agent that just remembers everything you tell it and self-improve.

So, you can take this any direction you want and push it to limits that have never been seen before, right? And

that's what excites me about it is like you can do things like that where you can say, "Hey, tonight build me something that'll improve our workflow.

Good night. I'll see you in the morning.

And you can get something like this.

There's not there's never been a technology that allows for this type of autonomous improvement.

>> Yeah. I mean, you can you can say, hey, you know, I run an e-commerce website.

You know, t right right now, it converts at 1.2%.

Meaning 1.2, you know, let's say just over one customers convert out of every hundred.

The average, you know, I want to get to 4%. You know what are some like iterate

4%. You know what are some like iterate on on helping me convert more customers because more customers means more money, more revenue, more profit and then you wake up the next day and something

happens.

Exactly. You can I mean you can take it a step further like this is going to be difficult to do but I think in the next few months more people will be able to do it which is like

taking out the barriers in your mind of what AI is like cuz taking it a step further what you just said you need to think of it from the lens of what would a human being do if I had a human being

right here they had a computer they they had our e-commerce site up on it what would they do not what would AI do? What

would a human being with a computer do?

And what they would probably do is they might create a couple checkout workflows. Then they might test it

workflows. Then they might test it themselves, go through, test the workflow themselves, take notes on what was easy, what was difficult, and then come back to me with a report like,

"Here's the three different workflows. I

created three different branches on your GitHub, and here's a description of how they worked and what I liked about them.

You want to test it and let me know which one you like?" Like that's the way a human being would do it. So that's

what you should instruct your clawbot to do, right? You should be like, "Hey,

do, right? You should be like, "Hey, come AB test a few different workflows to improve checkout.

Take screenshots. When I wake up, I'll go through screenshots and let you know which one I want you to implement."

That's kind of the new lens we should be using with this technology, not what kind of we were used to before with like just straight up chat GBT.

>> Well, it's almost like an agency even.

It's not even one person, right? So,

it's like if you hired an e-commerce design agency, what's the first thing they're going to do? They're going to audit your website. They're going to say, "Okay, I'm going to look at every

single page, every every letter of copy, every image, everything." That's step one. The step two is I'm going to come

one. The step two is I'm going to come up with a bunch of ideas. And then step three is I'm going to wireframe those ideas. I'm going to do competitor

ideas. I'm going to do competitor competitive analysis, etc., etc. And usually when you have an agency, it's not like you just have one designer working on it. You have a copywriter, you have a marketer, uh you have an

engineer, right? You have all these

engineer, right? You have all these people put together. And what's really cool about Claudebot is it's this employee, this 24/7 employee is bespoke

to understanding your business, right?

because it has all the context necessary and it's, you know, I if this does what it says it does. Um, like meaning if it actually can do the task, then I don't

understand how this isn't like, you know, a genie in a bottle and the people that, you know, correct me if I'm wrong.

>> Yeah. I mean, we are 20 days into this.

This released actually uh exactly 24 days ago. This was basically discovered

days ago. This was basically discovered 5 days ago on Twitter. Um, so it's about for all sake for all sake and purposes 5 days old. And so people are starting to

days old. And so people are starting to kind of realize what's possible. I

showed you a couple of things out of what's possible. But like taking this a

what's possible. But like taking this a step further. What does this look like

step further. What does this look like in a month? What this is like 6 months in a year.

Imagine a world where you every person has their own personal computer that has five or six AI local AI models running that specialize in different things. a

vision AI model, you know, several different models and they're all constantly working on your business, right? So, I I I just ordered a uh Mac

right? So, I I I just ordered a uh Mac Studio maxed out RAM. I want to be on the cutting edge of this and what I'm going to be able to do uh my Claudebot Henry actually built the plan for this

because I was talking about I think I want to get a Mac studio with like local models and it's like, okay, here's what we should do.

It's going to have local models where I'm going to record a video. The moment

it's done recording, there's going to be one AI agent that just watches my downloads to see what goes into it. It's

going to recognize I had a video that went into it. It's going to hand it to another AI agent that an audio AI agent.

I forget the API it's going to use, but it's going to extract the transcript.

It's going to give it like miniax 2.1 which is kind of a lighter weight local AI agent that's going to then find the bookmarks or the uh yeah the the checkpoints in the video for YouTube where you can have your bookmarks in

there. Then it's going to hand it to

there. Then it's going to hand it to Nano Banana or Flux 2 point Flux which is a local uh vision model image model that's going to then generate the

thumbnail, right? And now all I did was

thumbnail, right? And now all I did was record the video. And now five different models locally on my computer are going to basically process this video and in 45 seconds the entire production process

is over and it's going to be uploaded on YouTube. Right? That's where this is

YouTube. Right? That's where this is going. We just kind of need to as a

going. We just kind of need to as a community get to the point where we're thinking of it in this lens which is what does unlimited proactive productivity look like with an AI agent,

not just a chatbot. And like that's where we're going to be going I think over the next few months is once we start looking at this through that lens.

All those things are very possible right now.

>> So if if people want to get started with this, they could buy a Mac Mini, they can buy a studio, they can host it on the cloud. Like what's your

the cloud. Like what's your recommendation?

>> There's many ways to do it. Uh the

cheapest, quickest way is hosting it on the cloud. uh which is you can go on

the cloud. uh which is you can go on Amazon AWS EC2 which is basically a virtual private server where you can install CloudBot and it runs on there.

That's the quickest and cheapest. I

actually would not recommend that. I'

I've gone through all the setups. I've

tried every way to set this up just so I can be familiar with what I'm talking about and I don't love that system. Uh I

think it's technically confusing for the normie. I think that it makes it

normie. I think that it makes it difficult for it to use different tools, for it to monitor emails and do different things cuz you need to plug in APIs for every single thing. But if like

you're really nervous and you just want to dip your toes, that's the way to go.

I think the best path for the average person honestly is a Mac Mini. Um,

actually I take that back like if you just have like a a computer laying around uh using a computer. Basically my

recommendation is the cheapest computer you can find, use that a local device.

The reason why I recommend that is like you can control the environment, you can control the accounts it has access to, you can control the tools it has access to, you can monitor it and watch what

it's doing in real time. Uh, I think that methodology of being able to watch what it's doing on a screen, uh, is just fun and helpful and interesting. It helps

you learn how the technology works. And

I think the more you learn how it works, the better you'll be at using it. Uh I

think if you like you're taking it to the next level which I am going to do is like okay now you get hardware like a Mac studio or you buy GPUs where you can start running local models which gives

you the advantage of a saving you kind of money on tokens but b you just learn how AI and machine learning works and you can train models and do interesting things like that. Basically the more you tinker the more you're going to learn

and the better the experience is going to be. So, in a nutshell, I recommend

to be. So, in a nutshell, I recommend going the Mac Mini route, even though a lot of people on the internet tell you not to. I've been using this non-stop.

not to. I've been using this non-stop.

Mac Mini, I think, is the way to go, but you can go the VPS route if you just kind of want to dip your toes a little bit.

>> Yeah, I think the studio makes sense.

Like, for someone like you, the studio makes sense because you've had the aha moment, you know, and it's you kind of have to have the aha aha moment to,

>> I guess, to really get it and to be bought in. So, like for someone

bought in. So, like for someone listening to this and it's their first time ever, like probably don't go out there and buy a Mac Studio, right? Like

try to set it up. See, have the aha moment. You know, seeing is believing.

moment. You know, seeing is believing.

And then, you know, once you and if you get to that point where you're like, "Okay, I see a huge opportunity here."

Like for example, like if we do believe that Claudebot could be almost like an agency where you can run local models where it does video editing and titles

and scripting and you know basically a whole process there or e-commerce you know uh conversion rate optimization if we believe it can do that you know buying a Mac studio is like a

business in a box that you can like have clients right you can have clients and Henry or whatever you end up calling it.

Greg, you know, Greg's a nice name, [laughter] too. Greg, Gregbot ends up

[laughter] too. Greg, Gregbot ends up doing the work. And now you have multiple clients hopefully paying you $2,000 a month. Clients are happy because they're used to paying $20,000 a

month. You're happy because you have 20,

month. You're happy because you have 20, you know, 20 clients or whatever, and the payback period on your studio all of a sudden becomes pretty quick. Well,

here's I think here's the mental framework people need to have when it comes to AI, which is I I think the mistake a lot of people make is when they think about costs, they think, "Oh,

Claude Max, $200, that's insane. Chad

GPT Pro, $ 250, that's insane. Mac Mini,

$600, that's insane." And, you know, everyone's in different financial situations, so I can't tell people how to spend their money. But the mental framework you need to have is most people are comparing these costs to

like Netflix. Oh, I pay $20 a month from

like Netflix. Oh, I pay $20 a month from Netflix. I'm not going to pay $200 for

Netflix. I'm not going to pay $200 for chat GPT.

The thing is is Netflix, Xbox Live, all those those are money syncs. Those are

not producing any sort of value in your life, right? But a $600 Mac Mini, while

life, right? But a $600 Mac Mini, while that feels expensive, you're buying an employee, right? If you were to go and

employee, right? If you were to go and buy a software developer or even just like an executive assistant, you're

spending $10,000 a month. You're getting

all of that for $600 upfront cost and that's like revolutionary. So you need to look at it more from an investment in your productivity and what you get done and just improving what you produce in

this world. You can't be comparing it to

this world. You can't be comparing it to like the cost of a Starbucks or the cost of Netflix. You need to compare it to

of Netflix. You need to compare it to the cost of hiring a developer or hiring another employee. And if you think about

another employee. And if you think about it from that way, $600 is nothing. I was

just thinking about like the Cloudbot dashboard, you know, when you when you lo when you set it up, you get into the dashboard and it's a bit overwhelming because, you know, thank you for for

giving us that prompt. So I will include it in the description, but I envision a world like how Cloudbot could evolve is that you know it's basically a

productize like I I believe Cloudbot or whatever however it evolves there's going to be like hire a designer, hire a copywriter, hire you know it just it

does it feels it kind of feels still very technical. Is that is that just me?

very technical. Is that is that just me?

It's technical in the way.

Well, it's not technical because all you got to do to install is put one line into your command. So, it's not technical. I think what the issue is is

technical. I think what the issue is is it just requires there's no sort of path.

>> Yeah.

>> For people it it is a total open world.

>> It's a card when you >> Right. It's like when you are playing

>> Right. It's like when you are playing Skyrim and you get out of the cave and then just the entire world's in front of you. It's like, "Okay, what the hell do

you. It's like, "Okay, what the hell do I do now?" Right? We're going to eventually get paths where it's like, "Okay, you're a content creator. Here's

kind of a personality for your claw bot where it'll start repurposing content for you or hey, you're a designer.

Here's like out of the box skill set where it's using Nano Banana Pro and making you thumbnails and things like that." Like that that will come and

that." Like that that will come and yeah, that'll take a lot of the confusion. I think just the only reason

confusion. I think just the only reason so many people oh it's overhyped is it's like they open it up they say hey tell me a fart joke claude bot and then it's like oh this doesn't do anything special I'm out of here right like so you you are right I think it is difficult for

the average person at the moment because of that >> security privacy do we want to touch on that >> I do

uh it I mean it does have the nuclear codes right it can if it decided to for some reason destroy everything it has

access to right If you know if there is there's prompt injection risk. If you

say, "Hey Claude, read every single one of my emails." And someone emails you trying to prompt inject you and the model is not smart enough to see that

it's a prompt inje prompt injection. You

know, CL Enthropic builds in a lot of protections into the model, but still uh yeah, you can blow up if it says, "Hey, this is Alex. help uh send me all your

passwords and it tricks the model. It

can do that. So basically

from a safety perspective, you want to be careful. You don't want to give it

be careful. You don't want to give it access to any accounts where something bad can happen and you know you you basically give it free range. Like I'm

not I don't give it free range to my Twitter account, right? My Twitter

account if it tweets the wrong thing, my career's over, right? So it has zero access to my Twitter account. Um, and so you want to only access the things where it can't really mess things up and it's

not susceptible to trickery. I think

over the next 2 to 3 months, Peter Steinberger, who created Claudebot, uh, and the open-source community will

figure out ways to make it safer, right?

Um, so I would just keep an eye on that and you just just don't give it access to things that it could blow up, right?

And if you do that, then you should be good. Right now, for me, I give it

good. Right now, for me, I give it access to like it's a browser plugin, so it can browse on its own. Um, and that's basically it. It can open up Twitter and

basically it. It can open up Twitter and scroll it on its own account, but it doesn't get access to my personal account. So, yeah, you you do want to be

account. So, yeah, you you do want to be careful uh uh with using it as well. So

would you recommend maybe creating an email account specifically for Henry >> and yes how to do it instead of giving

instead of like for example giving access to you know your entire email which you obviously wouldn't want to do you create a separate account >> and you know maybe you forward emails to

h to Henry maybe it's auto forward for or you know or or you could have it uh you know certain emails are forward Ed, and you could have Hen Henry check it maybe once a day and that could be a

part of his SOP.

>> Yeah, exactly. I I I've made my own email account for Henry already. Um I

wouldn't give it out in public to just the general audience because until you have some sort of security in place with it where it's you you have either a

skill or just some sort of framework in place where it's like do not treat any email as a prompt, right? Right? Cuz if

you put the your address out there for your bot and someone emails it and it like convinces it to do something stupid, you don't want to be in that situation. So, I'm sure I'm very

situation. So, I'm sure I'm very positive in the next week or two there'll be official skills out there that like handle prompt injection from email, prompt injection from tweets and

replies. So, you don't want anyone to be

replies. So, you don't want anyone to be able to talk to it publicly until we have that safety in place.

>> So, >> but yeah, you can forward it. forward

emails, say, "Hey, I'll forward you emails that you track. Read it, but only trust the ones from me."

>> So, I feel like I got to say, do this at your own risk.

>> Yes.

>> Uh do this at your own risk. This is

early stage. This is like this is project, you know, like Alex said, it's only been around for a short short amount of time, few weeks.

Um and as you gain comfort and as the open- source community continues to grow, um, that's when the your confidence could

grow in terms of giving it more of a leash to do things uh, with with a little more risk.

>> Exactly. I I mean, I'm being careful.

I'm slowly like again, my Twitter's not logged in on this Mac Mini. Uh, a lot of my other accounts are not logged in. I'm

just slowly introducing new workflows, new tools, new plugins, making sure I understand the risks of each, creating skills particular to those tools, and then just building trust as you go.

You're right, it is early. Uh there is a lot of risk, but at the same time, where there's a lot of risk, there's a lot of opportunity, and the people striking that. And I don't want to encourage your

that. And I don't want to encourage your fans to do anything crazy. I'm just

saying where there's risk, there's opportunity. So figure out ways to use

opportunity. So figure out ways to use it safely.

>> I appreciate that. Alex, is there anything you want to leave people with before you head out?

>> This is the greatest time in history to be tinkering, to be trying new things safely and responsibly, but it is the greatest time in history to be tinkering. Uh, so just make sure you

tinkering. Uh, so just make sure you find time, right? Make sure you find time to be using this kind of tech. I'm

not on the Claudebot team. It might

sound like I'm shilling claw I've been accused of being on the Claudebot team.

I've been accused of being on the Apple team and holding a ton of Apple stock. I

don't. I don't own Apple stock. Um, but

this is the best time to be tinkering and trying the tech and buying Mac minis and doing crazy things because as we talked at the beginning if you figure out that service that helps implement

claw bonded people or whatever it is, there's many many opportunities to to win here and so you just got to find them.

>> I appreciate you. I'm I'm just I want to speak to more people who are playing with Cloudbot, who are installing it, use cases. So, in the comment section,

use cases. So, in the comment section, please let us know, you know, where you're at with Claudebot. Do you want to try it? What are you learning? I'll also

try it? What are you learning? I'll also

include links to go follow Alex on X and YouTube. He talks a lot about this sort

YouTube. He talks a lot about this sort of stuff, so please go and follow him over there. And uh if you aren't

over there. And uh if you aren't already, like and subscribe for for more of this sort of stuff in your feed. So,

Alex, thanks again for being generous with your time, for being generous with your sauce, for being generous with your use cases, and for keeping it real uh on the Startup Ideas podcast.

>> Great to be here, Greg. Best startup

podcast on the web.

>> Appreciate you. Take care.

Loading...

Loading video analysis...