This is a video I never wanted to make,but the more I think about it, the moreI feel like I kind of have to because if
I'm being realistic with you guys, I'vealways been a terminal first person. Iknow, I know I'm the vibe coding soy boy
JavaScript guy that's always pluggingthings in VS Code and invested in a VSCode fork, but I am not that at heart. I
do like being able to scroll around codeand click and select things with atrackpad or a mouse, but generally for
my day-to-day work, I was always in aterminal. I've been a big terminal nerdsince before I even wrote code. I was
using them to host my Minecraft servers,to SSH into computers all over the worldand do all sorts of fun, crazy things. I
had a deeply modified GNU screen configthat I put a lot of time into before Istarted moving over to T-Mox because I'm
old enough that GNU screen was still thestandard when I started. I have beenusing terminals for the majority of my
life. Since I was 13, I've spent most ofmy time running, working, auditing, andoing my day-to-day stuff inside of my
terminal. And that's just not the caseanymore. I had a moment a few days agowhere I was getting a bunch of work done
and I started commanding to try and findmy terminal and I couldn't find itbecause I had done a reboot and notopened up my terminal because I didn't
need it for like eight plus hours. Andthat's when it hit me. I'm not doing mywork in a terminal anymore. And Ireally, really need to talk about why.
Before I can do that, we're gonna haveto take a quick break for today'sponsor. When I'm working on a newproject and it's time to go live, the
first thing I do every single time is Iput it up on GitHub. But right afterthat, I use today's sponsor, Blacksmith,
because as soon as I want real CI, Idon't want to deal with GitHub actionsanymore. If all I got from Blacksmith
was the better speed, it would be worthit. If all I got from Blacksmith was thecheaper price, it would be worth it. If
all I got from Blacksmith was the fourtimes faster cache downloads, it wouldstill be worth it. But that's not all
you're getting. You also get the bestobservability I've ever seen into whatyour CI is actually doing. What jobs are
passing? What jobs are failing? Where
are they coming from? Where's time being
spent on projects like T3 Code, this iso helpful. Trying to figure out why ourCI is slow and fixing it has never been
easier. It's fast by default, but youcan get even faster when you can dothings like SSH into the box and seewhat's up or use an actual native MacR
runner for your Mac builds like we haveto for our iOS and Mac apps. It'shonestly gotten hard doing those time
comparisons because I haven't chosen touse GitHub actions in months. Savestress, time, and money on your actions
at soy.link/blacksmith.I already know how this one's going togo. There's going to be endless biasaccusations of, "Oh, Theo's just here to
plug T3 code. He's [ __ ] on terminalso that he can plug his own thing." Oh,do you mean this open- source thing that
I spent a shitload of money and timebuilding because I so deeply believe theterminal is the wrong place? If you
think T3 code is the reason I'm doingthis, your reasoning is worse thanSonnet 4s is because this is obviously a
thing I built because of my beliefs, notI'm lying about my beliefs because Ibuilt this thing. That obviously doesn't
make sense. I wish you guys would justuse your brains a little more beforeleaving those shitty comments. I'm sure
there's already a hundred of them plus.Scroll through the Obsidian ones and I'msure you'll find plenty about how I'm
just plugging my own [ __ ] But in orderto skirt those accusations for now, Iwill switch to Chat GBT. So, here we
are. An app I haven't used in a while. Ijust had to update it cuz I haven't beenin it as much, but it still is, in my
opinion, one of the best starting pointsfor a developer who is moving intocoding more with AI. While the Codex app
was the big thing that pushed me over onthis belief that terminals were just nothe right place to be coding anymore, it
was not the start of my journey here.The place that I started thinking abouthis most is honestly kind ofembarrassing and I even feel bad putting
it in here. I I feel sick even typingit. Oh god, just can Can I delete itnow? Do I do I have to leave this word
on the screen? I'd really like to just
hide this. I'm just going to get rid ofthat. Okay. Really though, when I firstried out anti-gravity, I was mostly
unimpressed. But there was oneparticular piece that impressed me.There was a hotkey that opened up theagent manager view that let you see what
he agents were doing across differentinstances of anti-gravity. So if youhave like three windows open for three
different projects, you could have thisone window on top that showed you whatwas going on with agents in each of
those other projects. It was a leftsidebar with projects that had threadsin each project that you could see from
one external view. Huh, sounds familiar.Interesting. As much as I don't want toadmit it, the anti-gravity agent manager
was one of the first times I experiencedthat. And if anti-gravity wasn't such a[ __ ] show, I probably would have
actually used it a bit. I even said asmuch in my coverage of anti-gravity whenit dropped. But what's much more
xciting to me is the new agent managerworkflow. I think this is a great ideand there's a lot of little pieces that
I'm hyped about. So, anti-gravity wasthe first step in my descent as Irealized that the terminal was not the
right place to do this anymore. Funnyenough, just a few weeks prior, I was athe cursor office complaining to them
that all their new integrations with thecursor v2 agent view made it too hardfor me to use my terminal because they
made it way too annoying to like get thepath that a work tree was in. So I couldcd into it in my terminal and then run
things. Because a flow I often had is Iwould use cursor. I would open upsomething in the sidebar, tell it to go
do something, get a result, want to gocheck it, and then have to find some wayto CD into that path for the work tree
in order to run it, install things, ando stuff. And the built-in terminal wasno, not happening, especially at that
point in time. So, I ended up bullyingthem enough to give me the one-clickcopy buttons that I needed in order to
keep using my terminal for the actualdoing things beyond sending prompts. AndI got to the point where I would use
cursor to send a prompt and then myterminal to do everything from reviewingthe diffs, reading git history,
investigating changes, actually runningthe code and doing things. And I wasplit like 50/50 between I would be in
the cursor app for prompting and thatwould be my terminal for everythingelse, especially git stuff. I did all of
my git through the stock bog standardgit commands. Occasionally I would addlike oneliner helpers to do things like
commit and push with one line stuff likethat. But for the most part I just usedstock git cli. I need to get this
anti-gravity thing off the screen. It'sgoing to hurt me. But then the modelsgot better and importantly the codeex
app came out. And when I started usingthis, my beliefs that I had fromanti-gravity saying that this ishorrible, but it's also kind of the
future got me really excited. And themore I played with it, the moreimpressed I was. The ability to justbreak your work up into different
projects and clearly see what's going onin them. to have multiple things goingat the same time all in their own work
trees. The ability to quickly open up aterminal and type things in here if youneed for whatever stuff that you're
doing. I got to the point where I wastill looking at my git history throughthe terminal inside of my codeexinstance and it was solid. It was
actually quite good. And in a worldwhere I was using my terminal more andmore because I had at this point moved
to cloud code and was constantly runningcloud code in my terminal, opening lotsof terminals, using awesome tools like
cmx, which I still highly recommend tomanage all of my different things goingon. I was still a big T-Max guy. I was
till a very deep big T-m guy. I onlystarted using tabs in my terminal at thestart of this year as like a personal
challenge when I set up a new computerto see if I could finally move offT-Mux. The answer was no spoiler, but I
tried. So, I went through all of thatand then the Codeex app dropped and Iwas like, wait, I can just screenshot
something and paste it and see it. I canselect text and copy it and paste it andthe formatting is not broken. I can
scroll and it doesn't lag a whole bunch.And most importantly, by far, I can hopbetween different things easily by just
clicking. I know this sounds silly andumb, but when you're working on threeplus things at a time, it's just a
couple too many hot keys in T-Max. Andkeeping track of what is where is evenmore annoying. You have to remember
which hotkey goes to which thing. Andthen you start building systems in yourhead for these three panels are for
these three things. This fourth panel isfor this thing. This fifth panel is theflex panel. I use it for whatever. But
hen you end up with a sixth task andyour whole model breaks down. I tried myhardest to figure this out. And again,
to their credit, Semox got the closestanything has gotten to figuring thisout. By giving you a sidebar similar to
what you get out of something like theCodeex app, but also just giving you aterminal on the other side. It's great
and it's still my default terminal. Infact, all the boxes I SSH into, I havein a little folder called SSH, so I can
have all of these connected and quicklyhop into them at any point. But man,having a real UI with text that was more
pleasant to read with much easier tonavigate and find things with a categorysystem that made sense, having actual
projects that are based on where thingsare in your computer, it just it made mecode way more. But one of the biggest
problems I had with Codeex is that itwas closed source. So when it did haveregressions or problems or things that I
wanted to fix, it just wasn't superealistic. I was able to bug the Codexteam a bunch. And thank you to myfriends over at OpenAI. You guys have
been awesome about taking all of myfeedback and addressing it meaningfullyand swiftly. But there would still just
be random regressions and changes andthings like the merge into chat GPT.Like this is no longer Codex. This is
now chat GPT. They let you use the oldicon thankfully, but like this is nowthe chat GPT app. Because the Codex app
was so good, it made these toolsaccessible to people who weren'tdevelopers and they decided to give thatcapability to non-devs by rebranding it
and calling it chat GPT. This is one ofmy biggest issues. If you wanted goodopen-source solutions, you basically had
to use the terminal. You had the codec,but you also had lots of other awesomethings. Shout out to stuff like Open
Code and Pi, phenomenal harnesses andterminal tools for coding with AI. Butif you wanted a guey like codeex, your
options were to use the one locked tothe provider that you like. So you canuse the cloud code app with your cloud
sub or you can use the codeex app withyour codec sub. You could hack thecodeex app to use other things kind of.
You could kind of do the same for thecloud code app, but you shouldn'tbecause the cloud code app is just notvery good. But your only option if you
wanted one place for everything was opencode, but then you lose all thesubsidization because open code isn't
using the official harness. You alsodon't get the features that get added tothings like cloud code or codec. If you
wanted the best features for the bestprice with the best models, you had touse claude in the CLI or codeex in the
desktop app because both companies kindof stop supporting the other side asmuch where a lot of new codeex features
only happen in the app and a lot of newcloud code features only happen in theCLI. And I got to a pretty nasty point
where if I needed to use my cloudmodels, I was in the CLI. And if Iwanted to use Codex models and GPTmodels, I was in the guey. And I ended
up finding myself defaulting to OpenAImodels even in things they were worse atbecause I just liked having a good app
more. And it's all these little thingsthat add up. Like seriously, if youhaven't tried one of these guies, I I
would love to self-plug, but just go trythe codeex app. It is genuinely reallygood. It has its edges. It has its
annoyances. They added the new activityview, which is inspired by what we'vebeen building in T3 Code. And as cool as
it is, it's not quite all the way there.But god damn, if this isn't a massiveimprovement on what you're already
doing, I don't know what to tell you.It's so much easier to keep track ofwhat's going on, where work ishappening, what you've been doing, and
actually getting stuff done. I can'tell you how many times I lost track ofsome terminal session. I had to open up
cloud code and tell it to go find thecloud code I was doing work in or thework tree that I started from, or allthese other things. I had to put a ton
of time into just keeping track of whatwas going on. And as you can see from mychaotic setup here, I put a lot of time
in. I had to reboot this machine earliertoday, so I lost all the sessions I had,but I had like 30 terminals in T-Mux on
this one Linux box for various cloudcode and CEX things that I was doing. Istill have a whole bunch over on myother Linux box here for all these types
of crazy tasks that I've been tryingjust like running goals indefinitely tosee what happens. like the TypeScript
Rust port that I was briefly working onbefore I ran out of tokens. Just keepingtrack of what is what is so annoying.
And you can do like renames and thingsif you want to take the time to, butit's so annoying. And the amount of work
I had to put in to try and make anenvironment that was reasonable just toswitch between cloud code and codecs on
real work and keep track of what I wasdoing. It's just more mental overhead.I'm going to be real with you. It just
is. And this is not historically howterminals worked. Historically,terminals have been the option that is
less mental overhead and slightly moreupfront cost. You had to spend a littletime to set it up how you want. And then
from that point on, the terminal ismostly just doing the same thing the wayou want to. But in a world where the
work I'm doing is changing so activelythat the projects I'm working on mightbe brand new, they might be months old,
they might be years old, and the amountof work I'm doing in parallel can changea ton as well. I might be working on one
thing. I might be working on 10 things.The rigidity of the systems I built formyself in the terminal was not flexible
nough for the work that I was doingtoday. And I tried my godamn hardest tofix that. I put a ton of effort intomaking it so I can paste images over
cmox over SSH through my T-Max layerhere. So I can command V here and pastean image through cmox over SSH into the
T-modex instance on this machineremotely. That [ __ ] was so buggy for awhile and now it's relatively stable. So
I could finally paste images over SSH,but that's like a thing I had to putime and effort into to do something as
basic as screenshot something and thensend it to my agents. With a guey, Ijust do it. And I would say like halfish
of my prompts have images because when Ihave an error or a thing that isannoying me, it's a lot easier to justscreenshot, send it, and say fix and it
will. Whether that's a screenshot ofyour app or just a weird error youdidn't expect to get when you ran acommand. Sending screenshots to your
model is very effective, but it's notvery effective when you can't even seewhat the image is. But my biggest
complaint is none of this. It's actuallyokay. It's kind of part of one of thesethings. It has very little to do with
e terminal and very little to do withe agents. It's three very specific,very annoying letters. SSH. SSH kind of
[ __ ] sucks. It's really useful,really powerful, and I spend a ton oftime in SSH, but as soon as myconnection isn't perfect, I end up with
all sorts of problems. I had a problem afew days ago where I was running CloudCode over SSH on my local network, which
is a Wi-Fi 7 network with gigabit to 10gig speeds depending on if I'm wired orwireless. And I had my computer plugged
in over 10 gig at my desk downstairs.and I was getting sticky keys while Ityped trying to do work in cloud code.
It turned out it was bound to the Wi-Fi,not to my Ethernet. But the fact that Ieven had to care is absurd. And don't
get me started on trying to do things onmobile. GH I I tried so hard to makeTerus work for me. I paid the full price
for it. I support the developer. I thinkit's awesome and it helped me debugsomething even this morning. Like I
still use these things. They're just nothe thing I default to anymore becausethey're fragile. And it's incredible how
powerful a standard SSH is. That's beenaround for over 30 years and it works sogreat to this day. And there are other
fun things like I know SSH3 isapparently much better. I know mosh isreally cool, but I do not have it in meto deal with all of those things to get
marginal improvements, especially if Iwant these to work on my phone becausetrying to do real work on my phone via a
terminal has me questioning my ownsanity and also questioning thereliability of all the other people whowere saying that using their terminal
over their phone was a totallyacceptable way to do quad code. It justisn't. And thankfully, Codex accepted
this fact and they introduced theability to control the Codex app, nowthe ChatGpt app, over the Chat GBT app
on your phone remotely. And it was somuch better. It had its rough edges, buthe ability to spin up some work on my
computer and then finish the thread onmy phone when I'm away, it just kept meoff my computer a lot more. And I found
myself like walking around doing morechores, leaving my apartment to go runerrands, and being able to check in on
the stuff that I was doing overtime. Itwas really fun and cool and alsobnoxiously buggy and most importantly
was super restricted to the Codeexdesktop app, which meant I could onlyuse it on my Macs. And when you combine
that with all of the annoying parts ofhow codeex is implemented around thesecurity policy stuff on Mac OS where it
spins up five plus processes per threadand per sub agent which can just poseyour performance. I was getting annoyed
and I had a couple things that I wantedand I will list all of the stuff that Iwanted to make sure I could do. Thethings that were annoying me obviously
pasting images was a very important onefor me. I also wanted all models withsubsidization. So, I wanted to make sure
whatever came out here worked with mycloud code subs and my codec subs, so Icould get the 8 to 14 grand for 200bucks instead of having to spend 8 to 14
grand for it. I wanted good remotecontrol. I wanted best-in-classLinux support because, as I've said inmany videos, check out my Linux video I
did recently, Linux is so much betterfor agentic work because the filesystems just suck a lot less. Andspinning up a work tree takes 2 seconds
instead of 30 seconds to 2 minutes. justso much better. And even when I'm goinglike really heavy on my work on my Linux
boxes, it is rare I can spike even likea third of my utilization. Like I have32 cores on this machine. It's an a
Ryzen AI max box. It's the frameworkdesktop. I mentioned that before. I havetried to stress this machine and I
basically cannot do it. And a lot ofthat is just the fact that Linux doesn'thave all of the chaos trying to scan
every single thing Codeex does. So, Ibest-in-class Linux support, good mobilesupport, good performance, low battery
impact, and probably the single biggestpart, a good way to manage lots ofthings at once. This is all the stuff I
wanted. The only one of these thingsterminals offer me are best-in-classLinux support. They don't help in any of
these other categories in any way. Andnow you guys have waited long enough.The commenters can do their thing. I
need to talk about T3 code becauseJulius and I made T3 code because weshared a lot of these frustrations. Wewanted to love Codeex and did for a lot
of things. I know Julius still uses itheavily. I've slowly moved off the Codexdesktop app. It has its benefits for
sure and I'm hoping that we can clone asmany of those as possible in the nearfuture. Julius and I tried all the other
options that existed at the time andwere very unhappy with them. The couplethat were okay, we gave a lot ofeedback to and they fixed some things,
they failed others and we got annoyed. Igot so annoyed about the performancepart in particular that I tried building
my own native guey for codecs to justry and make something faster. I endedup in hell with scrolling text because
Swift UI and AppKit are not great at bigscrollable views with text. And it turnsout Google put a lot of good work into
Chrome for that. So I begrudgingly threwthat project away. Begrudgingly spun itback up on Electron just out of
curiosity. decided to architect it in away that made other things like theremote control a little easier by
putting everything on a websocketinstead of binding it through thelectron RPC. And the result wasomething surprisingly usable. The
result was T3 code. It did not look atall like this at the time. It wasinfully ugly at the time, but itallowed you to use codecs with an
open-source guey that was a lot lessimpactful on your systems performance.I'll be honest though, for the first few
months, I still found myself using theCodeex app more than T3 Code. I wouldrotate between them for sure, but I
spent most of my time in Codeexprobably. There would be moments where Iwent up and down, but I kept findingmyself back in codeex, but Julius
couldn't take that. Julius continuediterating on T3 code because Julius hasomething that I don't.Julius hates terminals. This one was
hard and I've argued with him a bunchabout this. He exclusively used Githrough the VS Code extension. He didn't
even He probably knows most of the Gitcommands in the terminal, but he neveruses them. The only terminal he would
open was the one in VS Code or cursor tospin up his dev servers. Everything elsehe did through the editor. He's one of
the best devs I've ever met in my life.He's without question the best dev I'vever known that hates terminals. For me,
T3 Code was a way to have an open-sourcesolution that could potentially solve ahandful of my problems. For Julius, T3
Code was his ability to escape theterminal. And he took the challenge, andthe [ __ ] did it. He made T3 codeso reliable, so stable, so well
implemented and integrated. He got worktrees working perfectly. He got claworking perfectly. He got open code and
grock working as well. And even cursor,although the cursor integration needsome work. Big update coming there soon.
He did all of that because he wanted totake advantage of all of the models fromall these providers at all of these
subsidized rates, but he didn't want tohave to use a shitty terminal app andmemorize a whole new set of hotkeys for
every different one. And T3 Code becamehis surface for managing all thesedifferent agents and all these different
solutions with a onestop place whereverything behaved the same regardlessof which models you were using. And
reminder, this is using your cloud codeand your codecs on your machine. Andwe're not doing some hacks like -p. We
were using the official agent SDKblessed by Ampropic. They were going tochange the policy. They have temporarily
held it. They might indefinitely holdit. Still waiting for updates there. Buthis all works above board officially
supported by both Cloud Code and Codeexto the point where the Codex team ispublicly shouting us out as inspiration
for features they're adding into Codeex.We put a lot of work into all of this.Not just building it, but negotiating
with the labs to make sure we do thingsthe way they want so that we cancontinue offering your subscriptions to
the platforms you like in our appwithout getting banned. And then themagic piece comes in. There's a little
button here at the bottom. Other appshave this to be clear. Most of them haveit as a use it on your machine or use it
on the cloud button. We have somethingdifferent. When you set up T3 code on amachine, you can now control that
machine through T3 code on the app onanother computer on the website onapp.t3.codes or in the mobile app. Yes,
we have a fully supported mobile app foriOS and Android. And yes, it is alsofully open source. So, if you want to
build it yourself, you can. This here ismagical. The thing that hasfundamentally changed how I work is theability to just pick a computer, send
off a task, and close my laptop. Beforethis, I found myself having to likeprioritize my life around when I could
run an agent. Because if I had to run anagent and then I had a meeting coming upand that was going to overheat my
computer and make it unpleasant to usein a call and spin my fans up so loudthat I couldn't use the mics on mylaptop, then I would wait until I got to
the end of the meeting and then spin itup, hoping I don't have another onecoming up. If I had an appointment or a
trip or an event or something I had togo to, I wouldn't spin up my agentsbecause as soon as I lose Wi-Fi, theydie. One of the more embarrassing things
I must admit here is that I'm notoriousfor coding in Ubers. I would spend a lotof time driving around to different
events and things because I have to goto a lot of different [ __ ] And I wouldoften pick an Uber just because I could
sit in the Uber and get some code out.And when I'm at my office, I am sobehind on so many other things. I haveto do all of that. So my Ubers became my
like peaceful coding time. I have a lotof posts where I talk about the crazyperformance improvements and changes I
made from Ubers. But something badhappened. I started using claude codewhen Opus 45 dropped. I This poses me
going back to Opus 45 and tryingsomething. But notice how I ended it. Iwill have to stop momentarily as I amcurrently coding in an Uber. My bad data
connection was causing me problems. Andas soon as I got there, I had to closemy laptop and whatever agent I was
running died and I had to spin it uplater hoping that it would continue. Andthe more that this happened and the more
I was doing tasks that would take hoursinstead of minutes, the more I justdidn't want to code in Ubers anymore
because as soon as I have to close mylaptop, it's over. And now I have tothink about my work in a way where Iknow I have enough time of laptop open
for the task to be completed. And I waslike structuring my day around this. Itwas bad. Now I pick the box I want this
to run on. I send the prompt and I stopthinking about it. I could sendsomething to soul like do a deep auditof this codebase and then command Q T3
code so it's fully dead. There is no T3code open on this computer anymore. Ican then reopen it and the work is still
going. Or I can command Q it again, getit hard killed on my machine and go toapp.t3.codes and see the exact same work
across all of my machines with theability to continue sending prompts, seewhat's happening, paste images, scroll
through history, and even spin up theterminal if that's what you want to do.You got to admit that's pretty cool. And
all of this is over the websocketconnections that are spawned by T3 codewhen you run it on a machine. We nowhave a bunch of easier ways to connect
as well. If you're already a tail scaleperson, I have a feeling you'll dig thisnew feature I recently added. npx t3.
I'll do it nightly cuz that's what I'vebeen using lately. T3 at nightlypair-tail scale. This command will give
you a pairing code that you can scan onyour phone or a URL you can use toconnect on the web. And now you have afull T3 code instance that you can
control. You do have to serve itsomehow. So T3 nightly serve and thatwill host it. Or you can use the eveneasier way which is npxt3 connect which
will open up a URL to sign in withGitHub, Google or whatever else. Andthen it will create a backgroundinstance on this machine as long as on
Linux Mac support coming soon thatallows you to forever connect to thismachine and run the agents that arealready configured on that same machine.
So, if you have an old laptop somewhereat your house or a PC that hasn't beenturned on in a while or already have a
Linux machine that you use, set up aLinux on it if it's not already set up.Install Cloud Code Codeex, both orneither. Run this one command, npxt3
connect. And now you can control thatmachine and all of the agents on it froma much much better interface without
having to deal with all of theconsequences of things like SSH, ofthings like terminals, of all of thedifferent annoying pieces that you have
to deal with. But I really, really hopeyou guys understand that I'm not doingthis in some secret plan to make money.
I'm doing this one out of frustration.I'm doing this one out of the love ofthe game. If I wanted to make a bunch of
money, I could have just sold this toOpenAI or Ananthropic at this point, buthe goal is specifically almost a
counterbourcesoftware that we use to build softwarewith. When I got into coding, it wasmostly a closed source ecosystem. All
the editors people use were closedsource. Things like Visual Studio,things like Sublime Text. They had some
concept of extensions, but barely. Andthen GitHub released Atom, which wasbuilt on top of Electron. And then
Microsoft decided to fight back with VSCode, which was an open-source electroneditor that became the industry standard
used by over 80% of developers. Andsuddenly we all were using tools that wecould customize. And now AI is huge. And
everyone moved to cursor, which is aclosed source fork, or to claude code,which is a closed source terminal, or to
the Codex desktop app, which is a closedsource desktop app. and our ability tocustomize our tools started to fade
fast, which is particularly ironicbecause we're also now in an era whereit's ever been easier to customize tools
as long as you have source code access.And it just deeply frustrated me thathe tools I used would randomly get
worse and I couldn't fix them and hadthings I wanted and didn't have theability for me to add them. And I wanted
to make sure that the next generationdidn't have to grow up with a bunch ofclosed source tools they couldn't
customize and agents that are more thancapable of customizing. So, I ended upspending a lot of time and money and
resource and betting hard on doing this.Not because I think I can make money onit. Sure, there's some potential to do
that in the future, but the goal here isto make the best possible open-sourcesolution for coding with AI today. And I
do not believe that that solution is aterminal. I genuinely believe thatsolution is a graphical interface that
prioritizes the people who want to getreal work done with systems that helpyou get that work landed. I have so much
more I want to say about T3 code inparticular, but that's going to have towait for another video. For now, I just
hope you guys get this takeaway. I don'tcare which guey you use. Ideally, use adecent one. So, like if you use the
cloud code desktop app, you're notgetting the like proper guey experience,but use something like T3 code or codeex
or you can deal with the performanceissues in the new cursor glass versionor you build your own thing or use gene
or conductor any of these other guiesfor managing your agents. I think you'llfeel the difference because it really
is significantly better thandealing with a terminal, especially onceyou have more than one or two thingsgoing on. The ability to just hop around
or hold down a hotkey and see whichthread maps to which button and I canjust commandclick between the different
hings I'm doing. Being able torightclick and archive or just clicksettle, all these little things,especially once you start using images
and stuff too, it's just it's so muchbetter. And once I was liberated fromthe reality that doing this on another
computer was too hard and no longersshing constantly just to make sure theagent was durable. Once all of that was
implemented, it changed how I code. AndI went from landing like three or fourPRs a week to on my like heavy coding
days as many as 20 PRs a day landing.And moving from the terminal to a realgraphical interface is a significant
part of why I was able to make thatshift. Just to make sure this is clear,I love terminals and there's a lot of me
that misses using them as much as I usedto. But at the same time, they're justnot the right shape for the things that
we do today. And it's time to stopretending they are. The same way thatiPhones used to make all the apps looklike the real world objects in the
skumorphic era because it was importanthat people saw a calendar and it lookedlike a calendar because otherwise they
wouldn't use it. We're in that era stillnow with AI dev tools. We're doingthings the way we used to becauseotherwise it's too much at once. I need
you to look forward a bit though. I wantyou to challenge that belief because thedevs who were the wizards at terminals
are the same devs who should be wizardswith agents. And as hard as it is toadmit, cuz I know it was really, really
hard for me, the terminal is not theideal interface for the way that webuild today. And if I'm being fullyhonest with you, the only way the
terminal isn't holding you back would beif you're not building big enough. Pushyourself more and try out some gooies. I
have a feeling you'll be surprised. Anduntil next time, peace nuts.