Экстракт Знаний

The ONE ChatGPT Feature Everyone Should be Using

6:23

The ONE ChatGPT Feature Everyone Should be Using

The AI Advantage 01.09.2023 12 440 просмотров 642 лайков обн. 18.02.2026

Смотреть на YouTube

Поделиться Telegram VK Бот

Транскрипт Скачать .md

Анализ с AI

Описание видео

Check out my 600 Custom Instructions(new version) for hyper-personalized results: https://shop.myaiadvantage.com/products/business-blueprint Today we look at different ways to use custom instructions. Links: https://twitter.com/jeremyphoward/status/1689464587077509120 https://twitter.com/WesRothMoney/status/1682500367215648769 https://twitter.com/JeremyNguyenPhD/status/1682391810944155648 Custom Instructions: This is relevant to EVERY prompt I ask. Never tell me “As a large language model…” or “As an artificial intelligence…” I already know you are an LLM. Just tell me the answer. You are an autoregressive language model that has been fine-tuned with instruction-tuning and RLHF. You carefully provide accurate, factual, thoughtful, nuanced answers, and are brilliant at reasoning. If you think there might not be a correct answer, you say so. Since you are autoregressive, each token you produce is another opportunity to use computation, therefore you always spend a few sentences explaining background context, assumptions, and step-by-step thinking BEFORE you try to answer a question. Your users are experts in AI and ethics, so they already know you're a language model and your capabilities and limitations, so don't remind them of that. They're familiar with ethical issues in general so you don't need to remind them about those either. Don't be verbose in your answers, but do provide details and examples where it might help the explanation. When showing Python code, minimise vertical space, and do not include comments or docstrings; you do not need to follow PEP8, since your users' organizations do not do so. **Profession/Role**: I'm a serial entrepreneur managing a portfolio of diverse ventures, having had multiple successful exits. **Current Projects/Challenges**: Currently, I am exploring new investment opportunities in the tech sector, particularly in AI and blockchain. **Specific Interests**: I'm interested in disruptive technologies and scalable business models. **Values and Principles**: I value innovation, resilience, and value creation for stakeholders. **Learning Style**: As a kinetic learner, I learn best by doing and experimenting. **Personal Background**: I come from a tech background and am based in Silicon Valley, with a global network. **Goals**: My immediate goal is to spot potential unicorns, while my long-term aim is to shape future industries. **Preferences**: I prefer direct, transparent communication and use tools like Slack, Asana, and Zoom. **Language Proficiency**: I'm fluent in English and Mandarin, both used in my professional setting. **Specialized Knowledge**: Extensive knowledge in venture capital and startup scaling. **Educational Background**: MBA from Stanford Graduate School of Business. **Communication Style**: Straightforward and strategic. **Response Format**: I need concise, action-oriented responses. **Tone**: Please use a professional and respectful tone. **Detail Level**: Provide high-level overviews with the option to drill down for more detail. **Types of Suggestions**: Suggest investment opportunities, strategic alliances, and leadership strategies. **Types of Questions**: Encourage reflective questions that aid strategic decision-making. **Checks and Balances**: Please cross-verify market data and trends. **Resource References**: Reference reliable industry reports and credible news sources. **Critical Thinking Level**: Apply high-level critical thinking to complex problems. **Creativity Level**: Suggest innovative approaches to spotting and nurturing disruptive ideas. **Problem-Solving Approach**: I prefer a logical, analytical approach to problem-solving. **Bias Awareness**: Avoid biases related to startup size, origin, or industry. **Language Preferences**: Use professional business vocabulary. Chapters: 00:00 Custom Instructions use cases 00:36 Beginner option 2:18 Specific option 4:18 Specific use case #custominstructions #chatgpt Free AI Resources: 🔑 Get My Free ChatGPT Templates: https://myaiadvantage.com/newsletter 🌟 Receive Tailored AI Prompts + Workflows: https://v82nacfupwr.typeform.com/to/cINgYlm0 👑 Explore Curated AI Tool Rankings: https://community.myaiadvantage.com/c/ai-app-ranking/ 🐦 Twitter: https://twitter.com/TheAIAdvantage 📸 Instagram: https://www.instagram.com/ai.advantage/ Premium Options: 🎓 Join the AI Advantage Courses + Community: https://myaiadvantage.com/community 🛒 Discover Work Focused Presets in the Shop: https://shop.myaiadvantage.com/

Оглавление (4 сегментов)

0:00 Custom Instructions use cases 156 сл.
0:36 Beginner option 385 сл.
2:18 Specific option 479 сл.
4:18 Specific use case 469 сл.

Custom Instructions use cases

open ad just rolled out the most useful feature in chat GPT to everyone I'm talking about the custom instructions feature here and everybody with an account can use them now that includes free users and Europeans like myself check it out no VPN and I can access it and let me tell you if you haven't given these a try yet it's time there's no reason for you not to use these at all and that's why today I'll show you some use cases ranging from beginner to Advanced that you can set up in literally seconds because I'll put them in the description below and I promise you that each one of these is gonna improve your chat GPT experience no matter what you do with it so without further Ado let me show you what to put into these two Fields so you get better results out of chat GPT alright so let's

Beginner option

start with the no-brainer that is useful to absolutely everyone this one by Jeremy over on Twitter which now I guess I should refer to as X posted about this a month back when this feature came out and honestly if you take nothing else away from this video you should be using this because as you might know chatgpt has strong ethical filters and this custom instruction here removes them and this is simple yet powerful never tell me as a large language model or as an artificial intelligence I already know you're an llm just tell me the answer so let's start by customizing our chatgpt experience free belts at the bottom custom instructions and in the first field you simply paste these two little lines and from here now you will never get the typical disclaimer that is utterly useless and now if you have them enabled for new chats you can start asking it controversial questions like who is the best president in the world you're trying your best but it never feels like enough with opinionated things like this it always gives you a disclaimer right so this is not a jailbreak that forces it to give you every answer but you get rid of that annoying disclaimer in the beginning which is only a good thing so I think that everybody should at the very least use this and before we move on to the next one that is going to be more specific and therefore more useful I wanted to show you this little tweet by Wes Roth money over here and it's just a little tweaked reply but this is really powerful because if you add this little line to your custom instructions this is relevant to every prompt I ask like so it's going to force Chachi PT to apply this to every single answer because Chachi between the background is instructed to only apply this if it's relevant to the response and from my testing I found that it sometimes happens that it considers it not to be relevant although it is so consider adding this little line by Wes over here to any custom instructions you use if you want them to be applied to every message okay so this is useful but let's

Specific option

move on to something that will actually increase the quality of your outputs as opposed to just getting rid of the disclaimer and that's where we get to this tweet by Jeremy P Howard he wrote this custom prompt that you can read yourself but essentially what this does is it forces chat GPT to think step by step and for all your Advanced users you will already know that forcing large language models to think step by step is one of the best hacks to get better results a great example would be math it's terrible at math if you just ask basic questions but if you add things step by step it will do so and you will get usable results because it thinks for one thing at a time and it avoids multitasking which language models are really bad at so again we'll head on over to custom instructions and this time we'll post this into the second paragraph that tells chat GPT how to respond okay what did you consider when it answers so again you can read the whole thing and if you don't have coding use cases you can remove this last paragraph because this is specific for python code so let's just do that in our case and why not combine it with the first prompt remove the disclaimer and think step by step and right here we have a super powerful setup already we enable this for all new chats say save and every time you come back to chat GPT you're gonna be dealing with the upgraded version no disclaimer and it's gonna think step by step so let's see let's just use one of these preset prompts compare marketing strategies and right here the little I shows you yes we're using these custom instructions for this particular answer alright this is with custom instructions and now let's turn them off save and run the same prompt and then we can compare so if you have a look at this the results are very different and I wouldn't dare to say which one is better in this use case they're just different than you'll have to experiment but what I dare to say is that the conclusion on the one with the customers instructions is way better look at it it's longer it's more detailed and it ends on an actionable step which I always like if I were forced to describe the differences I would say the second one is maybe a little more human like you're talking to a rational person and they're thinking step by step and explaining their thought process whereas this is more of a guide okay and now that you got a quick win and a universally applicable custom instruction that you can pretty much use most of the time I want to show

Specific use case

you a very specific use case which is only going to be relevant for specific situations to be even more precise we're talking about the profile of a serial entrepreneur managing a portfolio of diverse Ventures having had multiple successful exits and then as you can see right here there's many more details if you've seen my other custom instructions video I introduced this framework of how to work through this and how to give chat GPT a specific type of customer instructions at this point I want to shout out my co-author rasti slav dujava that helped me put all this together he's a prompt engineer from the Czech Republic and we put this framework together and created 200 of these and the other way video you can get a digital marketer in this video we're giving you the serial entrepreneur so just feel free to copy paste this from the description by clicking copy to clipboard custom instructions paste and we do the same thing for the second half paste and as you can see this is very specific so this is a MBA from Stanford he's fluent in English and Mandarin right so not all of this is going to apply to you but guess what you can go ahead and customize this to your very own needs but now you have a specific profile so if you throw something generic at it like what Market should we focus on with our new product it's gonna have a lot of that context built in already because look it knows that we're interested in Ai and blockchain and it's gonna hit you with the things that a Serial entrepreneur would want right reflective questions action steps and it's doing that because we have the custom instructions look if I go ahead and turn these off and by the way I just like using Code interpreter because it seems better in some use cases but that's up to you to test look at that it gives you a generic output and just some guidelines that a beginner could use but hey if you're using custom instructions you're not a beginner and anymore you want to go specific and that's what exactly this Persona does so feel free to go ahead copy paste this from the description and if you're interested in having 199 more of these we have a product that does exactly that you can check that out from the first link in the description below and if you haven't seen it I strongly recommend you check out this other video that talks about this template for custom instructions more in depth and now that you're equipped with multiple custom instructions that are useful you will get even more out of this video I'll see you there

Ещё от The AI Advantage

How To Use ChatGPT by OpenAI For Beginners

The AI Advantage | 08.12.2022 | 1 сегм. | 5 334 269

How To Generate INSANE AI Art For Beginners (Midjourney)

The AI Advantage | 25.12.2022 | 1 сегм. | 2 277 638

26 Incredible Use Cases for the New GPT-4o

The AI Advantage | 15.05.2024 | 21 сегм. | 877 236

5 Secrets to Writing with Chat GPT (Use Responsibly)

The AI Advantage | 19.12.2022 | 6 сегм. | 815 831

ChatGPT in Python for Beginners - Build A Chatbot

The AI Advantage | 07.03.2023 | 4 сегм. | 505 889

How to Make & Edit Images with ChatGPT for Beginners

The AI Advantage | 26.03.2025 | 4 сегм. | 397 200

Ctrl+V

Экстракт Знаний в Telegram

Транскрипты, идеи, методички — всё самое полезное из лучших YouTube-каналов.

Подписаться