[0:01] Right now, there are two big reasons why
[0:03] people are hesitating to use open cloud.
[0:06] Number one, it is not free. You're
[0:08] paying for every token, every task,
[0:11] every agent use. That's money out of
[0:13] your pocket.
[0:15] And number two, every time you send
[0:16] something to open cloud, it goes out on
[0:19] someone else's server. Whether it's
[0:21] ChatGPT or Anthropic, your data leaves
[0:24] your computer, your contacts, personal
[0:25] projects, all of that goes out on the
[0:28] internet.
[0:29] What if I tell you there's a way to use
[0:31] open cloud without
[0:33] using any token and your data stays on
[0:36] your computer? So, today, let me
[0:38] introduce you how to use Ollama to power
[0:42] your open cloud with a local model. This
[0:45] is completely free and your data is 100%
[0:47] safe. So, how does this actually work?
[0:51] There are two pieces. The first piece is
[0:53] Ollama.
[0:54] Think of it as a platform that runs AI
[0:57] models locally on your computer. It
[1:00] downloads the model, hosts it, and keeps
[1:02] it running in the background. The second
[1:04] piece, obviously, is open cloud. So,
[1:06] normally, open cloud goes out to a
[1:09] provider like Anthropic for [music] its
[1:11] intelligence.
[1:12] But, when you point it at Ollama
[1:14] instead,
[1:16] it calls for local model for
[1:17] intelligence. This way, no internet is
[1:20] involved, your data is completely safe.
[1:23] So, the very first thing to do is just
[1:25] to download Ollama. Head down to the
[1:27] Ollama official website,
[1:29] ollama.com/download,
[1:31] and uh download the installation file.
[1:33] And after doing that, just install
[1:35] Ollama. Okay, now we open up this
[1:38] application, Ollama, and uh we have a
[1:41] fun task of picking a model which your
[1:44] open cloud will be powered on.
[1:46] So, just go to Ollama
[1:48] official page, and these are a list of
[1:51] models you can download to your
[1:53] computer,
[1:54] which can then be used to power open
[1:56] cloud. But, before you pick a model,
[1:58] there's one thing you need to
[1:59] understand.
[2:00] The most important thing about the model
[2:02] is the parameters. The more parameter a
[2:05] model has, the smarter it is. But, the
[2:07] trade-off is it also takes up more space
[2:10] on your computer and needs more
[2:11] computing power. So, if you have a
[2:13] regular laptop, you probably want to
[2:16] start with something small.
[2:18] For this tutorial, I'm going to pick up
[2:20] a very popular model that performs
[2:23] relatively well. It's called Kimmy K
[2:25] 2.5. Click on this copy button and then
[2:29] open Open up a terminal,
[2:32] paste in [music] this command, run it.
[2:36] It will say something like, "If your
[2:38] browser did not open, navigate to this."
[2:40] So, copy this URL,
[2:43] go back to your browser
[2:45] >> [music]
[2:45] >> and paste it.
[2:47] Connect.
[2:49] Go back to terminal, paste in this
[2:51] command again, and hit [music] enter.
[2:54] Connecting to Kimmy 2.5. Send a message.
[2:56] Just say hi.
[3:00] Okay, so this means your Kimmy model has
[3:03] been installed on your computer. Go to
[3:06] Ollama.
[3:07] You should be able to find that model,
[3:09] Kimmy K 2.5.
[3:12] The next thing we want to do is to
[3:13] install and launch open cloud via
[3:16] Ollama.
[3:17] To do it, also very straightforward,
[3:20] head down to this link
[3:22] and copy this command, Ollama launch
[3:25] open cloud.
[3:27] Copy it, go back to terminal,
[3:30] paste it, and hit enter.
[3:34] Open cloud is not installed. Install
[3:35] with npm. Say yes. Uh it will ask you
[3:38] which model do you want to use to power
[3:40] open cloud. So, here, just press enter
[3:43] on Kimmy 2.5.
[3:46] Proceed. Yes.
[3:48] And I understand the risk. Continue.
[3:50] Yes.
[3:51] So, if you're seeing this, that means
[3:54] your open cloud has been successfully
[3:57] installed. And sometimes, if you get hit
[3:59] with an error message, that means your
[4:01] node version is not up to date. What you
[4:05] need to do is to just update node. And
[4:08] the command to do that is
[4:11] nvm install node.
[4:13] So, right now, your open cloud is up and
[4:15] running. If you want to see the
[4:16] dashboard, scroll to this section. This
[4:20] is the URL for your open cloud
[4:22] dashboard. Copy the URL,
[4:25] go to a browser, paste it.
[4:29] And here you go. You can configure your
[4:31] channel, your skills, and everything in
[4:35] here.
[4:36] You want If you want to test it out, you
[4:37] can simply go back to the terminal and
[4:39] say, "Who are you?"
[4:44] Because we're running this on a local
[4:45] model,
[4:46] it's going to take a lot of time for it
[4:48] to answer your questions.
[4:50] So, to be honest with you, local models
[4:52] are definitely not as powerful as
[4:55] ChatGPT or cloud.
[4:57] If you give it a complex, multi-step
[4:59] task, it will struggle. The answers are
[5:02] generally shorter, less nuanced, and
[5:04] sometimes just worse. But, this is free
[5:07] and uh it keeps your data safe. So, make
[5:10] that trade-off and see if you want to
[5:12] use open cloud this way.