Explore the capabilities of the revolutionary DeepSeek AI chatbot in this comprehensive demo. We test DeepSeek's ability to answer complex science questions, summarize job postings from Australia and Hong Kong, and provide detailed summaries of chapters from 'Alice's Adventures in Wonderland' and 'Meditations' by Marcus Aurelius. Additionally, we compare job advice given by DeepSeek's default model and the advanced DeepThink (R1) model. Learn how DeepSeek, a Chinese AI company, is changing the landscape with its open-source approach and cost-effective training methods.
---
Read our DeepSeek guide 👇
https://watsontechworld.com/deepseek-ai-chatbot-demo/
---
DeepSeek website: https://www.deepseek.com/
---
Get my coloring book, Watson the Time-Travelling Dog: Dogs and Dinosaurs Adventure Coloring Book 1 🐕🦖🎨👇
https://watsoncoloringbooks.com/book1-locations/
Join Watson and his friends on an exciting time-travel adventure filled with dinosaurs and fun! Perfect for kids and coloring book enthusiasts alike.
---
Video Chapters:
00:00 Intro about DeepSeek
01:30 Asking about terraforming Venus
01:48 Asking about the deep biosphere
02:30 Dealing with an output: markdown and more
03:35 Summarizing a job posting in Australia
04:43 Summarizing a job posting in Hong Kong
05:37 Discussion about benefits of AI
05:57 Summarizing a chapter from Alice's Adventures in Wonderland
07:35 Summarizing a chapter from Meditations by Marcus Aurelius
09:03 Asking for job advice with default model
11:43 Using DeepThink (R1) with same question about job advice
13:57 Conclusion and review of video
---
#DeepSeek #AIChatbot #ArtificialIntelligence #OpenSourceAI #DeepSeekAI #AIDemo #AI #chatbot #R1 #LLM #OpenSource
---
Read our DeepSeek guide 👇
https://watsontechworld.com/deepseek-ai-chatbot-demo/
---
DeepSeek website: https://www.deepseek.com/
---
Get my coloring book, Watson the Time-Travelling Dog: Dogs and Dinosaurs Adventure Coloring Book 1 🐕🦖🎨👇
https://watsoncoloringbooks.com/book1-locations/
Join Watson and his friends on an exciting time-travel adventure filled with dinosaurs and fun! Perfect for kids and coloring book enthusiasts alike.
---
Video Chapters:
00:00 Intro about DeepSeek
01:30 Asking about terraforming Venus
01:48 Asking about the deep biosphere
02:30 Dealing with an output: markdown and more
03:35 Summarizing a job posting in Australia
04:43 Summarizing a job posting in Hong Kong
05:37 Discussion about benefits of AI
05:57 Summarizing a chapter from Alice's Adventures in Wonderland
07:35 Summarizing a chapter from Meditations by Marcus Aurelius
09:03 Asking for job advice with default model
11:43 Using DeepThink (R1) with same question about job advice
13:57 Conclusion and review of video
---
#DeepSeek #AIChatbot #ArtificialIntelligence #OpenSourceAI #DeepSeekAI #AIDemo #AI #chatbot #R1 #LLM #OpenSource
Category
🤖
TechTranscript
00:00Hi everyone. In this video I'm going to do an overview and demo of DeepSeq.
00:05And specifically I'm going to talk about DeepSeq.com.
00:09In recent months the company DeepSeq has revolutionized the AI world.
00:16And one of the very unique things about DeepSeq is most of their models are mostly open source.
00:27And also the code is open source.
00:31And you can check out their GitHub page and also they have a hugging face page if you want to check out specifically downloading the models and also the code.
00:42And to go to DeepSeq you would just go to DeepSeq.com.
00:46And specifically for this video I'm just going to talk about the chatbot DeepSeq.com.
00:51And DeepSeq.com is kind of like ChatGPT.com or like Google Gemini in that it's a chatbot and it's super smart.
01:01Again so you can go to DeepSeq.com and you might see it in Chinese.
01:06If you want to change it to English you can click on an English button in the upper right.
01:10And you can make an account and recently making an account has become much easier.
01:18And then once you make an account you'll see a blank sort of a blank canvas.
01:23But this is I asked it two questions and I just wanted to show you some what this might look like.
01:29I asked it a question teach me how humans might terraform the planet Venus.
01:35Okay this is a rather it's an interesting science question but it's the answer is very complicated.
01:43But I think this gave a very great answer.
01:47And I also asked it about the deep biosphere.
01:53So this is about life that lives very deep either in the oceans or very deep in the earth.
01:58So for example the deepest known life has been detected about 5 kilometers below continents and over 10 kilometers below the ocean.
02:15So I think that gave a really great answer.
02:18And in this video I'm going to go over some interesting questions.
02:22For example summarizing things and others.
02:25So I wanted to go over this question here.
02:30The one about terraforming Venus.
02:32If you want to actually copy the output.
02:36Okay so if you want to copy the output.
02:39You can click on the copy button like here.
02:45And it's going to give it as markdown code like that.
02:48Which this is markdown code.
02:49It's kind of like HTML code but a little bit different.
02:53And you can use a website for example markdown2html.com
02:58And if you paste it in like this then it will give it in something that's more familiar like formatted HTML code.
03:08And if you want to put it into for example Google Docs.
03:12You can make a Google Doc and then paste that output.
03:16I think that's very beautiful.
03:18Beautifully formatted.
03:21And now I'm going to show you how to ask it some other questions.
03:24So okay so briefly again you can copy the output like this.
03:31You can regenerate it and you can give feedback if you want like like dislike.
03:35Now what I want to do is show you how to summarize a job posting.
03:40So for example let's say you search for data analyst and you're searching in Australia for example.
03:46There's several different job postings.
03:49But let's say you wanted to search this one for some reason.
03:52Senior data analyst.
03:54So what we can do is copy all this text and give it to DeepSeq.
04:01It would say summarize the following job post.
04:08And I'll come right now this is what it looks like.
04:12I'll come back when it's done.
04:15Okay so that actually took less than 10 seconds.
04:19I think it wasn't not too bad.
04:21So it talked about where it's located.
04:23That's Brisbane.
04:24That's Queensland.
04:26And it gives information about salary.
04:30And the role is that the Crime and Corruption Commission seeks a senior data analyst.
04:35They want these skills.
04:37SQL, PowerBy, Tableau.
04:39So you can use DeepSeq to summarize jobs.
04:44And then let's say for example you wanted to search for a job in Hong Kong.
04:47This is yet another one.
04:50This is a data analyst job.
04:51Let's say you wanted to copy this.
04:52And let's summarize this one now.
04:59Summarize the following job post.
05:04Now I want to show you the full output so you see what it looks like actually running it.
05:10So it's pretty fast I think.
05:19Even just one year ago AI would have likely been much slower.
05:24And this seems really fast.
05:25I think it's great.
05:26So both the Australia job and the Hong Kong job, I think the prompts did a great job of summarizing them.
05:37And one of the great benefits of AI is that it can massively speed up your productivity.
05:42Massively improve your speed and effectiveness.
05:48And massively increase your productivity.
05:52And then of course if you wanted to copy this you can just click on copy.
05:56Okay, now let's say that you want to copy, excuse me, you want to summarize a chapter from a book for example.
06:05Let's say Alice in Wonderland.
06:06This book is in the public domain.
06:10I'm on gutenberg.org.
06:12And let's say we want to summarize the first chapter.
06:16It's possible DeepSeq has already been trained on this because this is one of the most famous books in the world.
06:22But let's just pretend that it isn't.
06:24We can say summarize the following book chapter.
06:29But I want to include the misspelling to see how smart it is.
06:34It probably will be able to just move on and not have to worry about the misspelling.
06:41So I copied the chapter one of Alice in Wonderland and I said summarize the following book chapter.
06:52Let's see how this is going to go.
06:53And given how fast it was last time, I think it's okay to show you the whole thing running now.
07:06So bored and restless while sitting with her sister, young Alice spots a white rabbit wearing a waistcoat and clutching a pocket watch.
07:14And I like that it gives the themes and moments and notable quotes.
07:19So would you like an analysis of specific symbols or connections to later events?
07:30So it's probably already trained on this book, but I think it gave a really great answer.
07:34Okay, now I want to summarize another chapter.
07:40Again, this is something it may have already been trained on, but this is a famous book.
07:45Meditations by Marcus Aurelius.
07:48He was a Roman emperor.
07:49And he's considered one of the most famous Stoic philosophers.
07:54So what I want to do is summarize the first book.
07:57That's the first part of Meditations.
08:02There's some commentary on it, but I'm not interested in the commentary.
08:06To say, summarize this chapter in a book.
08:12So I'm giving it Meditations.
08:13Meditations.
08:15Again, it might have already been trained on it, but I'm still interested to know how it can summarize.
08:24Summary of Meditations, book one, Debts and Lessons.
08:29Okay, so I guess he's already been trained on it, but I still think this is a great answer.
08:35So it talks about Marcus.
08:38And this book was kind of written for Marcus himself.
08:41I don't know that he necessarily thought anyone was going to read it, or at least he would have gotten so famous for it.
08:50So he talks about teachers and mentors and his family members, and a lot of it is just kind of notes for himself.
08:58It's a moral autobiography, and I think this is a great answer.
09:04Okay, now I want to ask it a question to give some advice.
09:10So previously, it's summarizing two different books, and it gave some answer to some science questions, and it summarized some jobs.
09:20Now I want to ask it to give some tips for a job, getting a job.
09:25Now I'm saying data analyst here, but you could substitute almost any job.
09:29So I'll read the prompt for you.
09:33Give tips for how to get a data analyst role within three months.
09:37Can one use the same resume for every employer, or is it better to customize the resume for each job?
09:43I want to be efficient with time, but also maximize results.
09:47Also, is it better to apply to many jobs on sites like Indeed and LinkedIn, or only apply to jobs I find interesting?
09:54Give me the best tips.
09:55So for this question, I'm going to ask it twice.
09:59First, I'm going to use the default model, and then in a moment, I'm going to use DeepThink R1, and this is considered kind of their smartest model, at least on deepseek.com.
10:11So first, let's run it with the default model, and then we'll use the more powerful model.
10:17So again, I'm asking it for tips for, and this was maybe a bit of a complicated question, and even an expert sort of advisor might not even give tips as good as this.
10:34So let's see.
10:38So use a hybrid approach.
10:40For example, resume tips.
10:41Use a hybrid approach.
10:42Create a master resume with all skills and experiences.
10:45Customize the top 25%.
10:48Okay, that is excellent advice, and I've read a book that gave similar advice.
10:53Format.
10:53Clean.
10:54One page.
10:55ATS friendly.
10:55No graphics.
11:00Keyword.
11:01Optimize your headline.
11:02I think that's correct.
11:03You have to build a portfolio.
11:06So you have different projects.
11:09SQL, Python, R, Bytools.
11:14So this is for data visualization.
11:17Excel.
11:18This is a spreadsheet.
11:21It doesn't have to be Excel.
11:22It could be Google Sheets or something like that.
11:27And yeah, I think this gives really great advice.
11:29So for example, have a, I like the advice I gave, have a master resume and customize the top 25%.
11:40Yeah, this is really great advice.
11:43And now what I want to do again is I'm going to ask the same question, but I'm going to ask the smarter model.
11:50So it says here, think before responding to solve reasoning problems.
11:56So it's going to think better, whatever that means.
12:01And I think it's probably going to give an even better answer.
12:03So let's see.
12:04So notice, it's probably going to say it's thinking.
12:10Okay, so this is taking a little bit longer, but it's also a smarter model.
12:15Okay, so let's see here.
12:16It says it's thinking.
12:18Okay, let's tackle this query.
12:19The user wants to handle, wants to land a analyst role in three months.
12:24They're asking about resume customization and where to apply.
12:28So I think this is really interesting output.
12:33Now this is in the output we're actually going to copy.
12:36We're going to, it's, you'll see in a moment.
12:40So here's a time efficient results driven plan.
12:45Resume strategy.
12:46Hybrid approach, save time, maximize input.
12:49Do not use the same resume for each job, but start from scratch, but don't start from scratch each time.
12:55Yeah, I think that's a great tip.
12:58And apply to both types of jobs, but split your efforts.
13:01So 60% targeted applications, companies you like.
13:0540% high volume applications.
13:07That is an interesting strategy.
13:10Filter for posted less than seven days.
13:12Auto fill tools.
13:14Aim for 10 to 15 quick applications a day.
13:16Okay, I think that's great.
13:18That's really great.
13:19And then it talks about making a killer portfolio.
13:23I think that's important.
13:24Crush interviews.
13:26Prepare for these.
13:29SQL and Python.
13:32Work on mock interviews.
13:34Network efficiency.
13:35Okay, this is really, really great output.
13:38And even let's, let's copy it and put it in the Google Doc.
13:46So that's what it looks like.
13:49I think that's really great output.
13:54And yeah, I'll give it a like.
13:55I think that's a great answer.
13:58So let's do a recap now.
14:00I'm just going to do a review for this video.
14:03So I started, I first, I asked it a question, terraforming Venus.
14:07That's a complicated question, a science question.
14:10I think it gave a great answer.
14:12I asked another science question, teaching me about the deep biosphere.
14:15That's about life.
14:16That's either super deep in the ocean or below the crust.
14:22So it's an interesting topic, in my opinion.
14:25I had it summarize a job post in Australia.
14:29I had it summarize a job post in Hong Kong.
14:34And I had it summarize two chapters, a chapter from two different books.
14:40So I'll summarize two chapters.
14:42And then I had it give me some resume and job advice using their default model.
14:48And then I asked it using R1, their smarter model.
14:51Okay, if you enjoyed this video, I'd appreciate it if you could subscribe to or follow the channel.
14:56And see you in the next one.