Hermes Agent is blowing me away...
542 segments
Okay, I have to be honest with you.
Hermes Agent is now better than Open
Claw. Open Claw has fumbled the bag over
the last couple months. Every update
they put out has broken the system. I
have spent more time fixing Open Claw
than using Open Claw. Now, this is fine
for me as a technical person. I don't
care. I can fix things pretty quickly.
But for most people out there, having to
fix your Open Claw every time you update
it has become a major pain in the wazoo.
If you were to learn just one AI agent
at the moment, it's got to be Hermes
Agent. In this video, I'm going to show
you how to set up your own 24/7
Hermes AI employee. An AI employee that
will literally do work for you at all
times of day and night, saving you tons
of time and creating tons of value. I'll
go over all the differences with Open
Claw, show you how to set it up, and
then show you three use cases that I
promise will change your life. By the
end of this video, you're going to walk
away with your own personal AI employee.
Now, let's lock in and get into it. All
right, so here's the story on Hermes
Agent, why it's taken the world by
storm, and why I'm now recommending it
for the time being over Open Claw.
Things change quick, but number one, and
this is the most important thing, it's
just plain reliable. Open Claw with
every single update, basically every
single update, breaks. I update Open
Claw, it breaks. I have to spend 20
minutes fixing it. It's a pain in the
butt. They release tons of updates. To
their credit, they are constantly
shipping, they're constantly updating,
but I don't know what's going on over
there. Are they not doing testing? Are
they not testing out their own tool?
Because the thing's constantly breaking.
Hermes Agent, while they do less updates
and ship a little bit less, they're more
reliable, and I would much, much, much
rather have less updates, but they don't
break and I don't spend tons of time
fixing the gosh darn agent. The updates
are better. It seems with every Open
Claw update, it's just shipping the
entire kitchen sink, right? It's a
hundred different things. They don't
really tie together. It's just a whole
bunch of things thrown together
hodgepodge style. With Hermes agent,
every update has a theme. Every update,
all the features tie together. As you
can see here, every release they do, it
has a theme, it has a name, the tenacity
release. And all the features in that
update will tie together, right? So,
this one had a Kanban board, goals, all
those things tie together. It seems like
every update is very well curated. It
doesn't seem like it's just throwing
things together. Everything kind of ties
together, which is really nice and also
just shout out the aesthetic of the
whole thing is very nice as well. So, it
it it it it feels much more well
curated, which leads to much more up
time and reliability, which I really
like. It's self-improving. This is a big
one. They've put huge emphasis on
self-improvement. How does the
self-improvement work? Check out this
chart here. Shout out Mr. Anon. I found
your newsletter. I saw this chart. It's
beautiful. You send a message to Hermes
agent. It does some sort of execution.
So, whatever you tell it to do, it
triggers a skill. So, everything you do
comes from your skills. And then, if
that skill doesn't exist or you ask it
to do something new that it's never done
before, it improves or creates that
skill. So, everything you're doing makes
the Hermes agent better, improves the
Hermes agent. So, the more you use it,
the better it gets. They put tons of
emphasis on this, which is really
amazing and you see it in real time. You
can see it here. I had to take a
transcript of one of my YouTube videos,
and you can see it goes, gets the skill,
then it takes the skill, improves it,
writes at top of it, adds whatever it
did, and it's just always improving.
Everything it does is a little bit
faster and a little bit better. It's
really, really cool. And I also just
like that they show you a lot of
transparency behind all the tool calls
it does and everything it's doing. So,
you can really track your AI agent
employee. And lastly, there's an
emphasis on experimentation. I really
like this. These people at new research,
it seems like they truly are like
hackers. I really love it. They give you
tons of tools around using local models
and things like that. It's training
models, Laura's, tons of tools around
machine learning and experimentation.
So, if you're a little bit of a tinkerer
like I am, it's really, really cool. If
you're buying tons of hardware and GPUs
like I've been telling you to do for the
last couple months, there's going to be
a lot of cool things for you to play
around with here. They are very pro
local model and tinkering. And one last
thing I'll add before we get into
installation and all the really cool use
cases I promise will blow your mind.
Swarms, they really built this to be out
of the box made from using multiple
agents, spinning up new agents quickly,
and like sending swarms of agents to do
things. All you have to say is build a
new Hermes profile and boom, it spins up
a new agent that you can now talk to and
use however you want. I do tons of
multi-agent stuff, and I'll show you all
that in a second as well. So, it's
really built for swarms, which I really
love as well. All this being said, this
is why Hermes has kind of taken the lead
is just this focus on reliability and
just working has really been amazing.
Now, let's talk about installation. I
won't spend too much time here. If
you're already set up and installed,
feel free to skip between the chapters
down below, really easy to do. But, much
like every other AI agent out there like
OpenClaw, it's just copying a command
and then putting it into your terminal.
So, I do a new terminal here. I press
paste, you hit enter, and it installs.
It gives you some options on how to
install. So, it's going to give you
options around what messaging service to
use. Just as I recommend with every
other AI agent including OpenClaw,
Telegram is the absolute best. Telegram
is king for AI agents. They're
constantly adding new AI functionality
to Telegram. The formatting is really
nice. You can see every tool call they
give you here. They have built-in
approval, so they ask your approval and
you can just hit accept in the buttons.
Everything is built around Telegram
really, really nicely. As for models,
this is the big question I get all the
time. This is where the big debate is.
After many, many, many tests, I am still
saying Opus is the best model for AI
agents. It is still the best model for
open claw and Hermes agent. It just does
things best. It has the best taste. It
builds the best UI. It gets the work
done the most reliably and efficiently.
If you have money, if you have a little
bit money to dedicate to this, I say use
Opus. And listen, you might say, "Oh
man, I'm going to spend like 3 to $400 a
month on Opus API. That's a lot of
money." Do some reflection here. Using
Hermes agent is going to improve your
life. It will add more efficiently, it
will add more efficiency, it'll help you
get more done, it'll make you better at
your job, your career, everything you're
doing. This is an investment in
yourself. If you're spending money right
now on Netflix, if you're spending money
right now on Xbox Game Pass, that's not
an investment in yourself. That's an
investment in you getting distracted,
right? So, before you go, "Oh, $300 a
month, that's a lot to spend on Opus
API."
This isn't the same thing as buying
Netflix. You can't compare those two
services, right? This is an investment
in yourself, which I think has a decent
chance of having a return on investment.
So, just keep that in mind. ChatGPT,
I'll be honest, ChatGPT 5.4 was
completely useless for AI agents. It got
nothing done. It was just a pain in the
wazoo. But, the good news here is
ChatGPT 5.5 is actually quite good. Is
it Opus level? No, it's not. It's not
Opus level. The user interface it
creates stink. A lot of decisions it
makes is kind of questionable. But, as
opposed to ChatGPT 5.4, it's actually
usable. It gets the job done. So, if
you're tight on budget, you can use your
chat GPT subscription. You could be on
the $20 chat GPT plan and still get
pretty good usage. So, I'd say uh you
know, if you're really tight on budget,
chat GPT gets the work done and you'll
still be able to get really good results
out of your Hermes agent. I'm not taking
you step-by-step in the terminal setting
up. It's really easy. Just copy and
paste the line in, hit enter, and it'll
take you through step-by-step. Really,
the only two decisions you have to make
are models and how to talk to it. But,
Telegram, Opus, those are ways to go,
but you can get away with chat GPT here.
You also have the option for models uh
to do local models. It's built for local
models, too. I have a video coming out
in the next few days that shows you how
to set up local models and power your
Hermes agent with it. They get the job
done. They're really good and they're
completely free. This is my Hermes agent
powered by Qwen 3.6, which is running on
my DGX Spark, and it is great. It does
really good coding. It can answer my
questions quickly and I'm not spending a
single dollar on it uh other than the
power going into the computer. So, this
is another option as well if you have
really good hardware. My recommendation
is if you're going to go local, set up
two agents. So, have two agents, one
local, one cloud, so you can give the
right task to each agent. If it's a
super complex task you need done
quickly, probably want to go cloud. If
it is a simpler task, then you can go
local and save tons of money. So, let's
get into the first things you should be
doing after you set up your Hermes
agent. I'm going to show you how to make
it personalized and make it your own.
Once we do that, we'll get into the
actual execution and the use case.
Again, feel free to jump around down
below if you need to. Also, if you
learned anything at all so far, make
sure to leave a like down below. Tell me
in the comments what you want me to make
a video on next. Hermes use cases, vibe
coding stuff with Claude code. Super
curious, let me know down below. Also,
subscribe and turn on notifications cuz
all I do is make amazing videos about
AI. I have a second channel I just
started, Alex Finn Lab. Subscribe to
that down below. I'm going to do more
off-the-cuff, just talking about what's
on my minds when it comes to AI. And
then the last plug, I swear to God, Vibe
Coding Academy. If you want to join my
community, I do live bootcamps every
Friday. I do Hermes bootcamps in there.
Link for that is down below. Join, join,
join. It'll be the best decision you
ever made in your life. So, here's the
first things you need to do when you set
up your Hermes agent. These are the
first things you need to do, no matter
what AI agent you're setting up. Super
critical, sets the stage for what you do
next, makes every task your agent does
way more impactful. And that is two
things. Two things we're about to do
here. Brain dump and reverse prompt.
What are those two things? Brain dumping
is first telling your agent everything
about yourself. Your career, your
ambitions, your goals, your interests,
your hobbies, everything about yourself,
you brain dump it in here. Now, let me
show you. So, here's a quick one about
me. I am Alex Finn, an entrepreneur and
content creator. I have 200,000 subs on
YouTube, 450 on X. I'm working on a new
startup, Henry Intelligent Machines,
which will help people build their own
businesses. I am super into AI. All my
content's about AI updates. I love AI
agents and like Openclaw and Hermes. I
also enjoy Boston sports like the
Celtics and Patriots. I live stream
Monday, Wednesday, Friday, 11:00 a.m.
Pacific. You brain dump everything about
yourself. This is the first message
you're sending, right? It allow the
agent to create a whole bunch of
important memories about you. Oh, Hermes
has an incredible memory system. So, you
brain dump it in, Hermes learns
everything about you, and you're good to
go. You should be more detailed than
this. I just did it kind of light for
demo purposes, but you should go in and
put in as much as humanly possible about
yourself, cuz it will create all of
those memories. Next, we are going to
reverse prompt. If you've ever watched
one of my videos before, you know what
reverse prompting is. I'm the biggest
reverse prompting proponent on planet
Earth. What does that mean? That means
instead of you prompting the agent, you
have the agent prompt you. You go, "What
are things we can do? Based on what you
know about me, what should we be doing
here?" Right? It's asking the agent
questions and seeing what they think.
Cuz remember, these agents are smarter
than us. This This is super
intelligence, right? So, here's what
we're going to do. Based on what you
know about me, what are the best
workflows we can implement? How could
you improve my life? Ask me any
questions you have for me. This is a
reverse prompt. The issue people have is
they go into these agents and they go,
"Uh I don't have any use cases. I quit."
Right? You should never be like that. If
you don't know what to do, if you don't
have any use cases, you go to your agent
and ask, "Hey, what can you do for me?"
Right? Based on what you know about me,
what can you do? The issue is a lot of
people, they just copy other people's
use cases, which are fine. I mean, I'm
going to give you a few of my own right
after this, and you should definitely
copy any you find interesting. But, the
most important thing here is you should
be using use cases that work for you,
that are custom for you, that your agent
does based on what knows about you. So,
you want to do this reverse prompt, and
it gets you custom use cases built
specifically for you that will improve
your life. I promise, if you do those
two things, brain dump and reverse
prompt, you're going to walk away with a
whole set of automations that will make
your life so much better. So, do that
exercise. Next, I want to go over three
use cases: a beginner use case, an
intermediate use case, and an advanced
use case. Let's go through those three
right now. Use case number one, and I
will put all these prompts down below so
you can copy and paste them and get set
up right away. This first use case is
for discovering and using and staying on
top of new AI tools. There are many,
many AI tools out there, most of which
you're not going to use, but you should
be experimenting and using new ones
every day. So, here's how this is going
to work. Every morning, tell me the
latest AI news with an emphasis on new
AI tools I've never used. Recommend a
new AI tool I need to use based on my
workflows. You're going to hit enter on
that. What's going to happen when you
hit enter on this, it's going to set up
a cron job for every single morning.
Then what's going to happen is when that
cron job triggers, your agent's going to
go, it's going to look at your
workflows, it's going to look at AI
tools you've used with it before, it's
going to look at different things you've
done with AI, and it's going to go and
find new and popular AI tools that it
will recommend based on your workflows.
Now, every morning you will discover new
AI tools to use that will improve your
life because they're custom based on
what you do. This is one of the
advantages of Hermes Agent, and this is
one of the things you should be doing is
taking advantage of its incredible
memory. It knows everything about you.
It knows everything you do, and now it
can make you custom recommendations
based on what it knows about you, and it
can help you stay on top of those latest
AI tools. So, this is huge. You want to
use this. I'm going to send this to you
down below in the description. Copy,
paste that, hit enter, you're good to
go. That's the beginner one. Here's an
intermediate one. Here's one I think
will be helpful for absolutely everyone
watching this video, and that's the
proactivity check-in. I think one of the
number one questions I get is, "How do I
make my agent more proactive, Alex? How
do I make it do things without me having
to ask?" Well, this is one of the ways
to do it, and this is a super powerful
method. Again, prompt down below, the
proactive check-in. Here's the prompt
down below. I want a proactive check-in.
Every morning, ask what my one big
priority is that day. Based on that,
update your memories about me and make a
recommendation to me on how you can help
my life or insights on things I can do
better. Again, we're taking advantage of
the memories. We're taking advantage of
the self-improvement and the skills.
Every morning, it's going to check in on
me. I'm going to give it my one big
priority that day and then it's going to
do recommendations for me. It's going to
recommend workflows so it can automate
those tasks for me or it's just going to
give me insights on my life that will
improve what I'm doing. This is huge and
this will make your agent way more
proactive. Again, this will set a cron
job for every morning. For those who
don't know what cron jobs are, they're
just scheduled triggers for your agent
that wakes them up. It'll schedule it
for every morning and help your agent be
way more proactive. So let's do a more
advanced one. This is really cool. So
this is using the new hype frames skill.
For those who don't know, there's tons
of new AI video creation skills coming
out for these agents that make it really
powerful for your agent to make videos.
There are many different reasons you'd
want to make AI generated motion videos.
For me, I'm doing it so I can make more
content so I can put out more helpful
educational graphics on my social media.
But for you, there could be many
different reasons you do this whether
you're getting into content, whether you
just want to help people out. You want
to use hype frames. The other big one is
Remotion. I think hype frames is pretty
cool. And what this will do is it will
have your agent build you out motion
graphic videos that you can use for any
reason. So use the hype frames skill to
build me a 60-second short on why people
should be using the Hermes agent. Make
the screen changes quick, visually
beautiful and eye-catching. Actually,
I'm going to make this one 30 seconds.
Make it a little bit quicker. I'm going
to hit enter. We're going to see what
this video looks like. I promise you
it's going to look cool. But there's so
many things you can be doing here with
these new skills. You want to stay on
top of the new skills. That could be
another kind of proactive thing you have
it do every morning is have it search
for other skills other Hermes agents
have and recommend it to you for how to
learn. See, this is amazing. It didn't
have the hype frames skill installed so
it went search for the hype frames
skill, found it. Now it's installing it.
It's hyperframes. Apologies,
hyperframes. I had a typo in there. Now,
it's going to install it and get it set
up and working and now we're going to
see a really cool video it's going to
generate for us. All right, boom, here
we go. That was actually pretty quick.
That took like 5 minutes.
Uh let's see what we got here. So, this
is a fully generated short. I can put
this on YouTube if I want. Here we go.
This is the future. This is your AI. It
runs itself. Look at this. So, it gives
you nice motion graphics, really nice
colors. I like the way things go in and
out. Now, would I just post this raw?
No, but maybe now I put a voiceover on
it. Maybe I put my face in the bottom
corner. And now, instead of having to
spend thousands of dollars on an editor
or hire a team, my Hermes agent just did
it for me and saved me thousands of
dollars. You now have unlimited control
to build any sort of videos or motion
graphics you want. You don't need to be
a professional anymore. The gatekeeping
has been taken away, which is really
amazing. That was the more advanced use
case. What I'm thinking about doing in
the next few days is doing like a video
where it's like 10 use cases on Hermes.
Just give you a whole gamut use case
focused video. I think that would go
hard. Let me know if you want that down
below as well. Again, if you learn
anything, leave a like, subscribe, turn
on notifications, super super critical.
Vibe Coding Academy doing another live
boot camp on Friday on Hermes agent.
Feel free to join, ask me questions,
chat with me, whatever you want. Link
for that is down below as well. Hope
this was helpful. I absolutely love
making these videos for you guys. It is
an honor you would even spend 30 seconds
watching my videos of me. I do this out
of love, so thank you so much for
watching these. Hope it was helpful and
I will see you in the next video.
Ask follow-up questions or revisit key timestamps.
This video presents the 'Hermes Agent' as a more reliable and well-curated alternative to the Open Claw AI agent. The host explains the benefits of switching to Hermes, including its self-improving capabilities, support for multi-agent 'swarms,' and its focus on stability. The video covers the setup process using Telegram, discusses model recommendations (such as using Claude Opus or GPT for varying budgets), and provides practical, life-improving use cases like brain dumping, proactive check-ins, and using automated motion graphic tools.
Videos recently processed by our community