YouTube2Text

I Tested PewDiePie's FREE ChatGPT - Odysseus AI (Setup + Guide) — Transcript

by Zinho Automates · 2,818 words · 395 segments · language en · Watch on YouTube

Full transcript

  1. 0:00PewDiePie just dropped his free open
  2. 0:02source AI workspace for your computer
  3. 0:04and it looks and it feels like Claude or
  4. 0:07even ChatGPT. We'll pair it with Ollama
  5. 0:10that runs the AI model on your hardware.
  6. 0:12Odysseus runs everything that you see
  7. 0:15[music] and touch. Your chats, your
  8. 0:16research, your documents, your notes.
  9. 0:18You can load almost any open model out
  10. 0:21there. For anyone using the basics of
  11. 0:23ChatGPT or Claude, you can now do that
  12. 0:26for free. And the best part about all of
  13. 0:28this is that you are not renting it.
  14. 0:30There's no subscription involved. You
  15. 0:32are not signing into anybody's cloud and
  16. 0:35the things that you actually type in, it
  17. 0:37stays on your machine. And in today's
  18. 0:40video, we'll set up Odysseus AI. We are
  19. 0:42going to test its features and we are
  20. 0:43going to see if it's actually worth
  21. 0:45time. And super quick, if you get stuck
  22. 0:47anywhere, then our free to join school
  23. 0:49community has every command and
  24. 0:51troubleshooting steps right there
  25. 0:53waiting for you. The link is down in the
  26. 0:55description below. Go and check it out.
  27. 0:57And now,
  28. 0:58let's check out Odysseus. Now, the first
  29. 1:01thing that we need to do is we need to
  30. 1:02open up our terminal and I know that
  31. 1:04word makes some people so nervous. So,
  32. 1:07let me just kill that right now.
  33. 1:08Terminal is just a text box where you
  34. 1:11type in instructions to your Mac or your
  35. 1:13PC and that's basically it. You are
  36. 1:16talking to your computer in a slightly
  37. 1:19different way, okay? So, the first thing
  38. 1:22that we need to do is we need to install
  39. 1:23Homebrew. And if you've never heard of
  40. 1:26it, just think of it as the App Store
  41. 1:29but for command line tools. It makes
  42. 1:32installing everything else is so much
  43. 1:34more simple. Now, head over to brew.sh
  44. 1:37and then grab the install command. And
  45. 1:39then you're going to paste it into
  46. 1:40terminal and then you're going to hit
  47. 1:41return and it'll take just like a couple
  48. 1:43of minutes and then it's going to ask
  49. 1:45for your password along the way. And
  50. 1:47now, once Homebrew is done, all you have
  51. 1:49to do is just run the command on the
  52. 1:51screen and then it installs Git, Python,
  53. 1:53and Tmux all in just one shot. Now, just
  54. 1:55a quick note, I'm not going to be saying
  55. 1:57all of the commands out loud in the
  56. 1:59video because you actually don't need to
  57. 2:01type all of this in from scratch. All of
  58. 2:03the commands are down in the Free School
  59. 2:05community. All you have to do is just go
  60. 2:06over there, copy and paste them, and
  61. 2:08then you can just follow along with the
  62. 2:10video, okay? Remember, link is down in
  63. 2:12the description. Now, quick breakdown so
  64. 2:14that you know what you've installed. Git
  65. 2:17is what we'll use to download Odysseus.
  66. 2:20Python is what Odysseus actually runs
  67. 2:23on, and then tmux is basically a small
  68. 2:26helper that Odysseus uses for some of
  69. 2:29the background tasks. Now, when it
  70. 2:31finishes, you are good, okay? Just keep
  71. 2:33going, and now we're going to download
  72. 2:35Odysseus itself, and the way we do that
  73. 2:38is called cloning, which I know sounds
  74. 2:41pretty technical, but it just means that
  75. 2:42we are copying the entire project from
  76. 2:44the internet onto your computer. Then
  77. 2:47you're going to run the two lines that's
  78. 2:48on screen, and the first one copies the
  79. 2:50whole project onto your machine, and the
  80. 2:52second moves you into that new folder.
  81. 2:54And that part is extremely important
  82. 2:56because every command after this runs
  83. 2:59from inside it. Now, just take a second
  84. 3:01to let that sink in. PewDiePie's code,
  85. 3:03the interface, all of it, is now sitting
  86. 3:06on your machine, and it was that easy to
  87. 3:08set everything up. Okay, so now I want
  88. 3:11you to stop because before you launch
  89. 3:14anything, there's one thing that you
  90. 3:16have to do first. And almost every
  91. 3:18tutorial about this has skipped it so
  92. 3:20far, which I mean, to me, is exactly why
  93. 3:22people end up in the comments just
  94. 3:25swearing the app is broken, and some
  95. 3:27people even saying that it just hangs
  96. 3:28forever and it never loads. Okay, so
  97. 3:31here's what's actually happening. The
  98. 3:33very first time Odysseus starts, it
  99. 3:35needs to download something called an
  100. 3:37embedding model. It's a small file that
  101. 3:40helps the AI understand and search text.
  102. 3:43Now, if your internet hiccups even
  103. 3:45slightly during the first launch, the
  104. 3:47download isn't going to stall, [music]
  105. 3:48and then the whole app just freezes
  106. 3:51mid-startup. Then your browser shows
  107. 3:53nothing and you just think that you
  108. 3:55broke something. When really it's just
  109. 3:56waiting on one file. Now the fix is
  110. 3:58simple. All we have to do is download
  111. 4:00the file ourselves before we launch the
  112. 4:02app. The commands are on screen right
  113. 4:04now. So just run the first three to set
  114. 4:06up the environment and then the last one
  115. 4:08to grab and cache the model. And that
  116. 4:11last one's basically a mouthful. So do
  117. 4:14yourself a favor, just go over to the
  118. 4:15free school community and then go and
  119. 4:17copy it over there and then you just
  120. 4:19paste it in, okay? It's going to save
  121. 4:20you a bunch of time when it comes to
  122. 4:21typing. Now you'll see a progress bar
  123. 4:23and then it prints cached, okay? And if
  124. 4:26your connection drops partly through,
  125. 4:29then it's pretty much a no big deal.
  126. 4:31Just press the up arrow and then hit
  127. 4:33return and then it picks up right where
  128. 4:35it left off. Now that one step just
  129. 4:38saved you a ton of frustration. So just
  130. 4:40going to say it right now, you're
  131. 4:41welcome. Now let's launch it, okay? So
  132. 4:45you're going to run the start script
  133. 4:46that's on screen right now and then
  134. 4:48you'll see the terminal start moving.
  135. 4:50Now the lines of text scrolling by,
  136. 4:52there's checks passing, the database is
  137. 4:54setting up. Then it will stop and it
  138. 4:55will ask you to create your admin
  139. 4:57account. It asks for username. So just
  140. 5:00type one or hit return to keep the
  141. 5:03default. And then it's going to ask for
  142. 5:05a password and then, okay, now another
  143. 5:08one, a heads up.
  144. 5:10When you type the password, nothing
  145. 5:12shows on screen. There's no dots, no
  146. 5:14stars and that is pretty normal. That's
  147. 5:16how terminal hides it. Just type it in
  148. 5:19and then hit return and then write that
  149. 5:22username and that password down to make
  150. 5:24sure that you always have it because you
  151. 5:26are going to need it every time you open
  152. 5:28the workspace. Now it prints setup
  153. 5:30complete and then it starts the server.
  154. 5:32Now because we cached that model in step
  155. 5:34three, it boots in a few seconds and
  156. 5:36then it opens your browser automatically
  157. 5:38to localhost colon [music] 7860.
  158. 5:42Then you're going to log in with the
  159. 5:44account that you just made and this
  160. 5:47this is now the moment, okay?
  161. 5:49This is your AI workspace. It's a
  162. 5:51running on your PC and it's not a server
  163. 5:55in another country, it's not a
  164. 5:56subscription, it's now yours. And I
  165. 5:59mean, if you just look at that for a
  166. 6:00second, you can see the navigation down
  167. 6:02the left, the chat window in the middle,
  168. 6:04and the model picker sitting at the
  169. 6:06bottom right. That's empty right now,
  170. 6:08but we are about to fix that. So, right
  171. 6:11now the workspace is up, but there's
  172. 6:13nothing behind it yet. There's no model
  173. 6:16to actually [music] answer you. Now,
  174. 6:18that's what Ollama is for. Go to
  175. 6:21ollama.com and then download the
  176. 6:24installer and then you're going to run
  177. 6:25it. It takes about a minute to finish
  178. 6:27and then when it's done, you'll spot a
  179. 6:29small icon up in your menu bar at the
  180. 6:32top of the screen. And when the Ollama
  181. 6:35is there, then Ollama is up and now it's
  182. 6:38ready to load models. Now, we pull our
  183. 6:41first model, the small M-135,
  184. 6:46and I want you to start with the
  185. 6:48smallest ones out there.
  186. 6:50Now, this is just to prove that the
  187. 6:52whole system works before we upgrade.
  188. 6:55And then we are going to run the command
  189. 6:57on screen. Now, that's a tiny model.
  190. 6:59It's about 19 MB, so it's going to
  191. 7:01download pretty fast. And [music] when
  192. 7:03it finishes, you've got something for
  193. 7:04the workspace to talk to. Now, we
  194. 7:07introduce the two programs to each
  195. 7:09other. Ollama's running, Odysseus is
  196. 7:11running, but they don't know about each
  197. 7:14other yet. And here's how you connect
  198. 7:16them. It's not buried in some settings
  199. 7:18menu, so that's at least one thing you
  200. 7:20don't have to worry about. It happens
  201. 7:22right in the chat. All you do is you
  202. 7:24click the chat box and then you type the
  203. 7:26setup command on screen and then
  204. 7:28Odysseus shows you your options. Now,
  205. 7:30for a local model like ours, it tells
  206. 7:33you to paste your endpoint address
  207. 7:35straight into the chat. So, that's what
  208. 7:38we do. We drop in the address on screen
  209. 7:40and we're going to drop that into the
  210. 7:41chat and then we are going to hit enter,
  211. 7:44and that address just points to Ollama
  212. 7:46running on your computer. It looks like
  213. 7:49a web link, but it actually never leaves
  214. 7:51your machine. Now, here's the mistake
  215. 7:54that gets almost everyone the first
  216. 7:56time. If you paste that before you've
  217. 7:59pulled the model, Odysseus comes back
  218. 8:01with no Ollama models found, and people
  219. 8:05basically just panic after seeing that.
  220. 8:06They think something broke, but the fix
  221. 8:09is exactly what we already did in the
  222. 8:11last step. You have to pull a model
  223. 8:13first. Now, on our side, we did that, so
  224. 8:16we are fine. Just paste that address in,
  225. 8:18and this time it comes back with found
  226. 8:21one model, and then it drops your state
  227. 8:23into a fresh chat with it selected. Now,
  228. 8:26if you ever need to switch to it
  229. 8:27manually, then that select model
  230. 8:28drop-down at the bottom right is where
  231. 8:30every model that you've pulled shows up.
  232. 8:33Now, just type your first message, and
  233. 8:35then watch. And that reply, it didn't
  234. 8:38travel anywhere. There's no API call,
  235. 8:40there's no round trip to OpenAI or
  236. 8:43Anthropic. Your machine generated every
  237. 8:45word of it. Great idea, and it's
  238. 8:47offline. Now, that tiny model, small
  239. 8:50135, was just to prove that the pipe
  240. 8:52works. It's honestly too small to be
  241. 8:55useful day-to-day, so let's grab
  242. 8:57something with some material range to
  243. 8:59it. So, I personally go with Qwen 2.5 7B
  244. 9:02if your PC's got the memory for it. It's
  245. 9:06sharp, and it's released under the
  246. 9:08Apache 2.0 license, which means it's
  247. 9:11genuinely free to use for anything,
  248. 9:14including your work. Now, on a lighter
  249. 9:16machine, drop down to Qwen 2.5 1.5B. It
  250. 9:21has the same license, a smaller
  251. 9:22footprint, and still a big step up from
  252. 9:25where we started. And then basically
  253. 9:26just pull whichever one fits. The
  254. 9:28commands are on screen right now. And I
  255. 9:30mean, now, let me just answer the
  256. 9:31question that I know you must be asking.
  257. 9:34Is this actually free? Well, the answer
  258. 9:36to that is yes. Not a tile, not a fee
  259. 9:39for now before we drop a price on it.
  260. 9:41This is free. They are open weight
  261. 9:44models with open licenses, so you
  262. 9:46download them once and then you run them
  263. 9:48as much as you want. There's no per
  264. 9:50message cost, no monthly bill. The only
  265. 9:52actual real cost is that bit of extra
  266. 9:54power that your machine draws while it's
  267. 9:56thinking. Now, once it finishes
  268. 9:58downloading, just go back to Odysseus
  269. 10:00and then open that select model drop
  270. 10:02down at the bottom and then just pick
  271. 10:03the model that you just pulled. And
  272. 10:05that's a no re-setup, no
  273. 10:07re-configuration, just pick it and then
  274. 10:09just go.
  275. 10:11Now, let me show you what you just
  276. 10:13built. Because once you see it all
  277. 10:15together, you realize how wild it is
  278. 10:18that this thing basically costs no money
  279. 10:21whatsoever. Now, first, the chat. It
  280. 10:24works like you'd expect. You type and
  281. 10:26then it responds, you have a
  282. 10:27conversation, except this one is
  283. 10:29private. Then you just run it with your
  284. 10:31Wi-Fi off and it still answers. Nobody
  285. 10:34else is reading it, no one is training
  286. 10:36on it, no one is keeping a copy on their
  287. 10:38servers. The only place it's saved is on
  288. 10:41your machine. And then there's deep
  289. 10:43research. You simply just give it a
  290. 10:44topic and this is the one feature that
  291. 10:46does reach out to the web. It pulls in
  292. 10:48sources, it reads them and it writes you
  293. 10:50up a report with citations. Now, simply
  294. 10:52just type on something that you actually
  295. 10:54care about and then watch it work
  296. 10:56through the whole thing and hand you
  297. 10:57something that you'd have spent an
  298. 11:00entire afternoon on yourself. Then
  299. 11:03there's also documents. You open a doc
  300. 11:05right inside Odysseus and the AI helps
  301. 11:08you as you write. It's basically like
  302. 11:10having a writing partner sitting right
  303. 11:12beside you, except you are the one
  304. 11:15that's driving everything. And then
  305. 11:17there's memory. Now, this is the one
  306. 11:18that pays off over time because it holds
  307. 11:21onto context. It stops feeling generic.
  308. 11:25So, just tell it about a project once
  309. 11:27and it will still know that exact
  310. 11:29project next week. The more you work in
  311. 11:31it, the more it fits the way you
  312. 11:34actually think and operate. And here's
  313. 11:36the thing, apart from deep research and
  314. 11:38reaching out for sources, all of that
  315. 11:40ran right on your machine for free
  316. 11:43without handing your work to anyone.
  317. 11:47Now, that's what you built today. And
  318. 11:49then there's tasks. This is where it
  319. 11:51starts doing things without you. You can
  320. 11:52set up scheduled prompts, research
  321. 11:54answers, or actions, and it just does
  322. 11:57them on its own. Have it run a research
  323. 12:00report on a topic every week. You set a
  324. 12:02trigger, a schedule, an event, a
  325. 12:04webhook, and then you just pick what it
  326. 12:06should do, and then it just runs in the
  327. 12:08background. It's kind of like giving
  328. 12:09your AI a to-do list, and then having it
  329. 12:12actually will follow through. Okay, so
  330. 12:14let me knock out the three issues that
  331. 12:16most people hit, and this is going to
  332. 12:17help your build go more smooth.
  333. 12:20If the app won't load or it just hangs,
  334. 12:23it's usually that embedding model from
  335. 12:25step three that's getting interrupted.
  336. 12:27So, just rerun that cache command, and
  337. 12:30then run the start script again, and
  338. 12:32then you're done. Now, if you see the
  339. 12:34command not found when you try an Ollama
  340. 12:37command, then it almost always just
  341. 12:39means that Ollama isn't installed or
  342. 12:42isn't running. So, then something just
  343. 12:44check for Ollama in your menu bar. Now,
  344. 12:46no Ollama, no model, just open Ollama
  345. 12:48first. And if you get a port already in
  346. 12:53use error, then something else is
  347. 12:55sitting on the port Ollama wants. So,
  348. 12:58the command that's appearing on screen
  349. 12:59right now, that clears it. And then
  350. 13:01simply just start it again. Now, I've
  351. 13:03dropped all of this down in the free
  352. 13:05school community. Go and check it out,
  353. 13:07and everything that we covered in
  354. 13:09today's video, all of it is down there,
  355. 13:11okay? So, the link is down in the
  356. 13:13description below. Now, simply just take
  357. 13:16a step back, and then look at what
  358. 13:17you've got. The model, the interface,
  359. 13:20and your chat history. All of it lives
  360. 13:23on just one [music] machine. It's yours.
  361. 13:26There's no bill showing up every month.
  362. 13:28There's no company sitting in the middle
  363. 13:30and nobody setting the rules about what
  364. 13:33you can ask the thing that you are
  365. 13:34running locally. And this is just the
  366. 13:37starting line. From here, you can pull
  367. 13:39bigger models as you get comfortable.
  368. 13:41You can set it up to run tasks for you
  369. 13:45automatically. You can wire it into your
  370. 13:47own tools and files and even your
  371. 13:49workflows. And if you want the full
  372. 13:52command list from today plus the
  373. 13:54troubleshooting guide and you want a
  374. 13:56place to ask some important questions,
  375. 13:59then the free school community is where
  376. 14:00you are going to want to be. The link is
  377. 14:03down in the description below. Go and
  378. 14:04check it out. There's a bunch of people
  379. 14:06that is currently there all learning
  380. 14:08from each other and you could join them.
  381. 14:10Okay. So now, if you made it till the
  382. 14:12end of this video, I'm pretty sure that
  383. 14:13you had a lot of fun. So, give this
  384. 14:15video a like, subscribe to the channel,
  385. 14:17and comment down below. What do you
  386. 14:19think the future holds for Odysseus? Do
  387. 14:21you think it's going to change
  388. 14:23everything? Or do you think it's going
  389. 14:24to burn up pretty soon? Let me know down
  390. 14:26in the comments down below. And also,
  391. 14:28check out one of the videos popping up
  392. 14:29on the screen right now because I
  393. 14:30guarantee you you're going to learn
  394. 14:32something cool and I will catch you on
  395. 14:35the next one.

About this transcript

This page contains the full transcript of I Tested PewDiePie's FREE ChatGPT - Odysseus AI (Setup + Guide) by Zinho Automates, generated from the public captions YouTube serves with the video. The transcript has 2,818 words across 395 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.