- 5
- 8 925
Boris Nikolaev
United States
Приєднався 4 вер 2013
Welcome to my channel! I’m a professor of entrepreneurship who likes to help students, entrepreneurs, and educators unlock the power of AI tools in both their work and studies.
My goal is simple: to share practical, real-world use cases that will save you time, boost productivity, and make learning more engaging. I’m new to video creation, so I’d love to hear your feedback-there’s always room for improvement, and I’m excited to grow with you!
So, ready to share the AI journey with me? Let’s explore together how these tools can truly make a difference in education and beyond.
My goal is simple: to share practical, real-world use cases that will save you time, boost productivity, and make learning more engaging. I’m new to video creation, so I’d love to hear your feedback-there’s always room for improvement, and I’m excited to grow with you!
So, ready to share the AI journey with me? Let’s explore together how these tools can truly make a difference in education and beyond.
Clone Yourself into a Fantasy Character with Flux AI (easy & awesome!)
Learn how to fine-tune Flux.ai, a powerful open-source image generation model, to create personalized AI art.
This tutorial covers step-by-step guide using replicate.com/
0:00 Introduction
1:21 What is Flux.ai?
2:06 What is LORA?
2:25 Using Replicate.com
2:42 Training Your Model
4:22 Generating Pictures (with some samples)
5:33 A Few Tips
6:53 Conclusion
Cost: ~$2 for model training, ~$2 per 50 images generated.
Ideal for personal projects, business branding, or creative projects.
Use responsibly.
Meta Prompt: "Create a detailed, vivid description of me as a Star Wars character, following these guidelines: 1. Start with my trigger word 'BORIS_NIKOLAEV' to activate my custom model. 2. Describe the character's appearance, including clothing, weapons, and any unique features, using specific and colorful language. 3. Incorporate Star Wars-specific elements, such as lightsabers, droids, or alien species. 4. Specify the setting, whether it's a familiar Star Wars location or a new, imaginative environment. 5. Include details about the lighting, atmosphere, and perspective of the scene. 6. Mention a specific artistic style or technique, such as 'cinematic realism' or 'concept art illustration'. 7. Add technical specifications like 'high-contrast lighting' or 'shallow depth of field'. 8. Structure the description to flow from character details to setting and finally to artistic style. 9. Aim for a concise yet comprehensive description of about 60 words. 10. Use vivid, concrete language to ensure predictable results. Please generate this description, keeping in mind that it will be used as a prompt for an AI image generation model."
My Substack (if you prefer reading): tinyurl.com/9tny6hr7
Please note: The FLUX.1 [dev] model used in this video is open-sourced with a non-commercial license. It is intended solely for research, education, and personal projects. Commercial use of this model or its outputs is not permitted under the current license terms. Enjoy responsibly!
This tutorial covers step-by-step guide using replicate.com/
0:00 Introduction
1:21 What is Flux.ai?
2:06 What is LORA?
2:25 Using Replicate.com
2:42 Training Your Model
4:22 Generating Pictures (with some samples)
5:33 A Few Tips
6:53 Conclusion
Cost: ~$2 for model training, ~$2 per 50 images generated.
Ideal for personal projects, business branding, or creative projects.
Use responsibly.
Meta Prompt: "Create a detailed, vivid description of me as a Star Wars character, following these guidelines: 1. Start with my trigger word 'BORIS_NIKOLAEV' to activate my custom model. 2. Describe the character's appearance, including clothing, weapons, and any unique features, using specific and colorful language. 3. Incorporate Star Wars-specific elements, such as lightsabers, droids, or alien species. 4. Specify the setting, whether it's a familiar Star Wars location or a new, imaginative environment. 5. Include details about the lighting, atmosphere, and perspective of the scene. 6. Mention a specific artistic style or technique, such as 'cinematic realism' or 'concept art illustration'. 7. Add technical specifications like 'high-contrast lighting' or 'shallow depth of field'. 8. Structure the description to flow from character details to setting and finally to artistic style. 9. Aim for a concise yet comprehensive description of about 60 words. 10. Use vivid, concrete language to ensure predictable results. Please generate this description, keeping in mind that it will be used as a prompt for an AI image generation model."
My Substack (if you prefer reading): tinyurl.com/9tny6hr7
Please note: The FLUX.1 [dev] model used in this video is open-sourced with a non-commercial license. It is intended solely for research, education, and personal projects. Commercial use of this model or its outputs is not permitted under the current license terms. Enjoy responsibly!
Переглядів: 2 877
Відео
Systematic Literature Review with Elicit: 4 Practical Use Cases + Limitations
Переглядів 4437 днів тому
In this video, I will show you how to use Elicit, an AI-powered tool, to speed up your literature review process. You will learn how to: 0:00 Introduction 1:13 Limitations of traditional search 1:58 Semantic search 2:46 Rapidly discover relevant papers using Elicit 4:39 Extract custom data from your papers 6:53 Export your literature review 7:53 Identifying gaps, themes, and common concepts in ...
Crush School with Notebook LM: 5 Powerful Use Cases for Students
Переглядів 6 тис.19 днів тому
In this tutorial, I am going to show you five use cases of google's research assistant NotebookLM that can help you study more efficiently! We will go over: Summarizing notes and clarifying complex concepts Generating a practice quiz Creating a podcast Building a study guide Preparing for essay questions All prompts and references are available in my substack posts: tinyurl.com/nhptsbcf Thank y...
ChatGPT for Data Analysis
Переглядів 337Місяць тому
In this tutorial, we are going to cover how to use ChatGPT for data analysis. You can find a written version of this guide on my Substack: tinyurl.com/svdaz92x We will use data from the World Happiness Report: tinyurl.com/5n99hrnj You will learn how to: - Create visualizations such as line charts, bar graphs, heatmaps, and more. - Perform time series analysis to track global happiness trends - ...
Reducing AI Hallucinations & Staying Awake at the Wheel
Переглядів 178Місяць тому
AI hallucinations are a fundamental feature of Large Language Models (LLMs). In essence, hallucination is the core function of LLMs - they generate plausible continuations based on learned patterns. This capability often results in outputs that are both useful and engaging. In many applications, particularly in creative and knowledge-intensive professions, non-deterministic answers frequently h...
Grok is censored
Thanks for pointing this out! It seems to be far less censored than other models which is causing a lot of controversy. But you are right that there is *some* censorship.
@@borisnikolaev1982 Hands down, it is the best as far as being censored as compared to others.
It's terrible I'd rather use sdxl
Based on the ELO leaderboard, flux is better! But maybe you should make a video to compare the two? Flux worked great for me. Thanks
@@borisnikolaev1982 the photos you've created are terrible compared to what sdxl can produce.
@@W-meme I find the video and information it conveys to be enormously helpful. It makes it all so accessible to those of us that don't have time to work through more complex setups. And, the images he created look great to my eye.
A beautiful and inspiring video !🙂
Already free to clone yourself in any setting free via meta messenger, ig, WA etc
Thanks for sharing! I wonder if fine-tuning Flux still gives you more flexibility since you can clone anything.
You forgot One thing! PDF & Documents take a while to upload if you dont have an proper internet. I recommend either taking screenshots and Uploading to Gemini or Any Ai Image Recognition App& Send those Screenshots to those Apps and tell it extract the text and either copy paste them or use Google Docs for this example. I think even NotebookLM Could Images (i am not sure)
Great demo! Would it be possible to make a model with two dogs in order to create images with consistent characters, i.e. the dogs, in different environments and play roles?
Great question! I only tried the model cloning a single person (object) at a time. I wonder if you train a model with dog 1 and then train the trained model with dog 2 if this will work.
Flux 1 dev model (the one being used here) is for NON COMMERCIAL USE.
Great point! Noted in the description of the video.
Elon musk has nothing to do with Flux being uncensored. Don’t give that idiot any credit for the work of others.
Amazing!
Thank you for the excellent video. Can I use the trained model in a completely private setting? or I should use entirely offline after downloading it?
Thank you! Yes, just select your model to be "Private" when training it! (this is what I do). I ran all my models online (through the replicate.com website). While you can technically download the model, running it locally is more complicated. The website has really intuitive interface and you are basically purchasing compute time! Hope this helps!
@@borisnikolaev1982 thanks!
love it! :-)
🔥🔥🔥
Very original thumbnail, your face + pointing. The #1 sign of mediocrity on UA-cam. Don’t consider for a second how looking like EVERYONE else makes you a lemming moron not a marketing genius 😂
Very original thumbnail, your face + pointing. The #1 sign of mediocrity on UA-cam. Don’t consider for a second how looking like EVERYONE else makes you a lemming moron not a marketing genius 😂
Professor with annoying students here. Thanks man!
Great video!!!
This is mind-blowing! You're saving lives here hahaha😂
haha
Super interesting and useful tool! Thanks a bunch!!!
Thank you for the help.
Great job 🎉 …!
Thank you! 😃
Very helpful video !
I like your podcast, Bobi and find it interesting🙂
Go Boris! Wooooo!
I’ve been using NotebookLM for months and really enjoying it. You made a very nice overview of the product. Good luck with your channel.
Very informative video - this is a game changer for me in terms of learning new information efficiently and in finding new ways to learn. Thank you!
Боби, много добро представяне.