what is up You Beautiful People welcome to this free course called The Ultimate Guide to prompt engineering and chat GPT in this entire training I'll give you a brief introduction to Jack GPT and how exactly you can write the requests the questions that you asked to the GPT to get the maximum output of it after a lot of popular demand I figured this is something that you might need so here I am so let's get started so in this particular training we'll be covering these contents which is chapter number one introduction to chat GPT we'll
learn a little bit about it strengths and weaknesses of Charity GPT what it can do what it cannot do I'll give you some tips for success and to get the most out of chargpt3 because that's the most the free version chapter number four is when I will show you how to verify the information that you just gathered out of chat gbt number five will be the introduction to prompt engineering probably why you are here number six prompt defining number seven prompt refining chapter 8 I'll give you three additional tips to take your prompting to the
next level and chapter nine is when I'll give you some bonuses that you can take home and use strategypt to become an expert marketer or a businessman right so without any further Ado let's get started with the chapter number one which is Introduction to child CPT so welcome to chapter number one introduction to chat GPT in this chapter we'll learn a little bit about the background of this GPT thing why is this called GPT and what is this AI thing how did it did it evolve all those things a little bit of a boring stuff
but it will be important for you to observe and notice why you are getting the kind of results that you are getting right now right so we're talking about GPT GPD stands for generative pre-trained Transformer language and it is designed to understand and generate human-like text based on information it has learned from various sources like Wikipedia common cruel so many things I'll tell you more about those and the GPT right now the the one that is developed by this company open AI is you know strained a massive number of data sets to make uh this
beautiful AI product which understands what we have to say to it so it just means that it has already learned much before that you can even interact with as a result it can generate meaningful responses based on the knowledge it has acquired during its training so here is how the AI systems work to be honest so we train a system to you know understand some input this input can be a question a query and analyze it reason with it understand what exactly the customer means and give you the right answer right to understand analyze and
give you the right answer it has to have some prior knowledge right which is as good as asking let's say you have a employee at your home and let's say you have a personal assistance and you tell your personal assistant hey grab me the Apple right or let's say grab me my phone or Grammy a phone or whatever let's call it grabbing an apple right and this apple might mean an Apple phone and it can also mean an apple the fruit right so if they don't know or they don't have that reasoning capability to understand
what exactly do you mean by this they will probably mess up right so this is exactly what it is this language model is trained on massive amount of data so that it can understand what you are telling and give you the right suggestion of the right output or the right response so let's also talk about large language models so Chinese GPD is some sort of an extrapolation of this NLP or I would say machine learning technique called as the natural language processing which means they train the AIS or the machine to learn how humans talk
right that's how the natural language processing power comes in and among these there is a certain kind of model called as the large language model large language models essentially means that the the AI has been trained on massive amounts of data and especially in this case Text data and they are trained on that particular data so that they can predict what word comes next that is exactly why chat GPD is also you know the the core system the core model behind training this AI this chart GPT AI is called the large language model training well
no surprise but it was discovered that the more data you feed to the system the better it learns and the better is the quality of the output right that is what has happened and chat GPT which the free version is called GPT 3 or 3.5 it has been trained on 175 billion parameters right I'll uh I'll send you the link of the paper maybe drop in the comment I'll uh you know send the link there ah this was trained on 570 gigabytes of text and its predecessor gpd2 was over 100 times smaller and at 1.5
billion parameters this was the uh this was a paper published by this Stanford or stands for University human centered AI uh newsletter and this has been trained on massive data like common crawl webtext to books one by two Wikipedia English examples of the program coded in CSS JavaScript python now llms the way they work they predict the next word that can come after a particular query so if I say hey bring me the whatever right based on the context it will give the next word but at this level of data input that we have given
to charity GPD it can do so in a mind-bending scale so this ability allows them to write paragraphs and entire pages of content there are people writing entire books there are people writing blog posts and whatnot using just change CPT and that is the magic behind it the next thing that I want to talk about is chart GPT versus Google a lot of the people keep on arguing hey strategy Google killer and you know how is it different so let me just clarify that particular thing you see chat GPT is an artificial Intel agents and
it's some sort of I would say a personal assistant I I see charity GPD as literally a personal assistant and it's it's as of like you give it a particular context it understands that and gives you the right output this is like as good as having an employee or an intern that can like listen to your command now it's then you can have a conversation with this GPT and the kind of input that you get is very very personalized to just your particular problem or the query right and it's very good to get detailed answers
descriptions have a little bit of context behind it and rather I would say a faster generation of solutions that is why chat GP is so so popular because imagine Google Google is more like a library if you go into the library you can ask the library and hey I want this particular help what do you suggest the library and tell you know what uh section 2 il-3 row number five marketing book name go ahead and check read that book page number three you probably get your answer that is exactly how Google works you know the
mission statement of Google to organize world's information that's what it does so it basically gives you direction to the books or the library however it doesn't give you the exact answer nowadays it is obviously improving charity PPT understands your query it's like that math tutor let's say who can understand what is your doubt and give you the right answer tell you that you know what you should do this and this should give you the solution so it's just like that right I hope that is a clear-cut answer to this particular question which is like how
is chat GPT different from Google right and why there is such an arms race for AI is because well you get information so so faster with an AI that a lot of people started using it and when open AI made Chad GPT public what happened was everybody started using it and it gathered a massive number of user base which is why Microsoft showed interest in GPD and started you know just bought it completely for I think 10 billion dollar if I am not wrong now the beautiful part about it is like that this AI can
be integrated with the search engines to give you even more personalized recommendation one more thing that happened is because with Google a lot of the uh I would say a lot of the revenue came from Google AdSense and people would search for something they would see ads they will see such results and they would Google would charge money for the ad impression and everything but here people are getting results right there so they might just ask charging period and about the query and Google might just feel left out and they don't earn anything hence Google
is like having that competition that they also want to come up with a AI tool which can basically do the same thing which is why they have launched Bard and it is not yet launched in India at the time of this recording and if it is at all launched at some point in the future obviously will be the first ones to use it right so yeah that's about uh you know how is chart GPT different from Google the next chapter we will discuss the strengths and weaknesses of chat GPT and how you can use it
for your own benefit right so let's get into chapter number two where I'll be talking about strengths and weaknesses of chat GPT so without any further Ado let's get started so here's the reason why I love chat GPT the first thing is personal station like I mentioned in the last chapter that the solutions that you get is very very personalized based on the query that you give like for example if I tell hey I'm a 28 year old male I am 76 kgs of weight and my height is this my age I just told you
and I give me like a seven day workout routine home workout routine through which I can lose another 4 kgs and also give me a South Indian diet plan that should go with it keep it you'll understand and give you that response imagine if you just had to do the same with Google you'll probably have to go ahead and research about you know what are the common household exercises I could do then research about the nutrition question the kind of food that I should eat when I'm in Karnataka although all these things right however with
GPT the kind of language has been trained on it will give you the answer that much faster that much more personalized to your specific needs right that's the beauty of it although I'd be careful when it comes to health related things I'll talk about that in a little bit of time second is ease of use it's like super super easy easy to use I mean the interface cannot be like more simpler just go to chat Dot openai.com and you pretty much got this input like you're pretty much sorted just select the model type in something
and you get results out of it right it cannot be more simpler this is why it is you know so so popular that there is no complications like other AI tools like you have to go ahead and figure out okay vsl key there is a different tool there is a tool for writing blog posts this is like too much right and third is contextual understanding it is very very good at understanding the context behind the query and give you the right response so if I just tell you hey act like a strict manager sending a
email to the employer regarding their bad performance it will definitely take on that particular Persona and write a certain kind of response right that is exactly why people love it a lot so you see right here dear employee I'm writing to express my disappointment with the recent performance as a valued member of our team I mean this is pretty good right it has given a pretty nice long email that a manager can send so this is exactly why it understands that particular concept and gives the right response and that is why we love it uh
fourth reason why I love chat repeat is because of its decision making progress it is very very Advanced when it comes to like deciding things although I would not completely depend on charge GPD to give you know take decisions for me the reason why I'm saying this is because ultimately we are humans and what I would suggest is you treat GPT as like your personal assistant you know you just take the recommendation but don't Implement on it without having some critical thinking around it I hope that makes sense right and fifth obviously time saving I
kept mentioning this that a lot of the time you just have to ask a question you get the answer as opposed to like researching on Google hours and hours you just get the data in minutes that's the most beautiful thing about chat GPT right but that said there are multiple things that I personally don't like about GPT and there are a lot of problems in it right so for example the first problem is the quality of answers depends on the quality of directions that you get now charges ability to provide helpful answer relies on how
you ask questions or give instruction it's as good as saying that I mean let's take an example if you had a personal assistant well let's say you had an intern and you told them that he hey you know what uh Grammy and apple they might either think it's an apple the fruit apple the phone if he still manages to understand that you have asked me the Apple phone he might not know where exactly the phone is but if you tell me your intern or your personal assistant said that you know what grab my iPhone I
think I have kept it near the lounge next to the sofa and that was the flower was near to it then they'll have a direction where to go right so you have to be very very detailed when you are asking questions to get the right kind of response and that is obviously I would classify that as a smaller problem because if you do not give GPT the right instructions it is still not smart enough to understand what exactly you are asking for next it does provide non-toxic response and while it is good it does avoid
some and not safe for work kind of responses but because of which sometimes the answers becomes very insensitive and irrational right now here's one more thing the answers that you get in GPT three or even four sometimes are not always correct you see the hard limit of knowledge of chat GPT is at like September of 2021 if I am not wrong it has been trained on the data that was published before that it's not after that so many times what happens is the information that you get is limited to that particular date and not the
any kind of information that is published after 2021 so sometimes if you ask it for some codes and stuff if the you know kind of the platform itself is updated you may not get the same answer so always double check for any critical error that you might get with GPT itself right now many times what happens is the reasoning capability of gptc or even GPT 4 is not the best of all for example in one of the future lessons I'll show you uh how I use GPD to give out the data on meal plans of
of an entire week so that I lose weight and it is not the most perfect response that I had expected out of it right so sometimes it is not the best when it comes to complex reasoning but it is good at like simple reason like hey what should I do if I'm hungry and I want to lose weight how do I control my Hunger all those things it can help but when it comes to a little bit of a more complex reasoning thought process that is not the best language model of all as of now
right in future it might really change GPT 4 definitely has come come very very long uh you know gpd4 definitely has come out very well and it's much more advanced than chat GPT 3 itself all right welcome to chapter number three tips for success in getting the most out of chat GPT 3 so right here we'll talk about how you can get the maximum results out of deeply some tips and then we'll move into front engineering and all that stuff right so first thing first like we mentioned in the last class that the quality of
the output depends on the quality of the input and that is exactly why you have to be very very specific and very clear in your requests you have to provide context draft a detailed paragraph on what exactly you want out on GPT and the better you put in your the input the better will be the output of it right and it's as simple as that so be very very specific don't just write hey what workout should I do rather what would be a suitable workout for a middle aged man who wants to lose weight and
is on very rich Indian diet right be very specific and it will give you the right kind of response here second be patient and understanding because at the end of the day this is not a human that can understand and help you out this is an AI that has a very limited set of data to work with so all the responses may not be the most perfect ones so what you need to do is to have an open mind and do refine The Prompt I'll show you the process of defining and defining prompts in the
future but do consider changing and iterating and giving out a very good I would say giving a very good input query here third use gpt3 as supplement not a replacement like most of the people say that you know what it might replace us in our jobs and what not well it will not right it is it's as good as that intern or that you know personal assistant it might replace them but still it is not physical it's still digital it can only give you that kind of information I would not depend on chat GP to
give me a complete let's say a product in itself even when I suggest people writing blog post and stuff I still tell them that you should go ahead and you know revisit the entire thing use a little bit of your subject matter expertise and then you go ahead and publish a blog or do whatever right your own expertise it should be at the core of every single problem statement that you give into charity and you should verify every single information that you get out of chat CPT right that I hope that makes sense on always
when it comes to Health Medical and legal professions always always will consult some expert not take its word for granted right and do not make that mistake there might be some lacunas it might create and that might land you in trouble in the future so I would not recommend that at all finally have a lot of fun and be creative on your input uh sometimes you might want to have fun ask chapter GP to write emojis on your blogs ask GPT to like you know connect you with some uh gifs with the blogs that you
are writing all those things you can do and it's really really fun thing to have right and that is all about it chapter number four is where we'll be discussing how exactly you should verify the information that you have gathered with chat GPT all right so let's move on to the next chapter all right welcome to chapter number four on how to verify the information that you get out of chat GPT well the information that you normally get out of chat GPT may not be always the best information out there so you need to use
your brains a little to see whether this is something that you are you want right so there are the first thing is like ask chat GPT for specific information backed by sources and sometimes I have seen like the link that you generate that work approximately 50 to 60 percent of the time not always so if I were to give you an example the most significant contributions of Albert Einstein to Quantum physics let's see so it has definitely added a lot of the contributions uh we'll just ask them give me the a pro create sources for
each of these points so you go ahead and simply ask them so it has given some uh you know some links so go ahead and check this links properly but if you see a lot of the links don't work like see the space doesn't exist this kind of exists but I think if you click here you'll get some notes and all this doesn't exist four zero four error this is one more and this is one more so some information does come some doesn't so in my experience what I've seen is like roughly 50 to 60
percent links do work else don't if you're writing a research paper using GPT do be careful on the information that it gives out right uh use tools for verification like Google Wikipedia libraries and database consult experts in the field and do look for and do use a little bit of your critical thinking skills to understand whether this is the right thing or not many times you might get some nonsensical response from GPT so use your brain understand whether this is the right thing to do and then implement it use it as a supplement not a
replacement just like I mentioned right investigate any point that might seem questionable or uncertain because the last thing that we want is to have wrong data in our hands and do cross check information through other means or methods through this right so I hope this helps from the next chapter will be going into the more into interesting subjects this is like prompt engineering so this is where exactly we'll figure out how to write kick cast prompt so that you get wonderful results right all right welcome to chapter number five which is called the introduction to
prompt engineering and here we have two things right if I have to Define what is prompt engineering well prompt engineering is designing and refining the input text which is also known as The Prompt uh to guide the AI model like charging between genetic specific and relevant responses example here when I when we put like can you share the links to the references or what are the biggest best exercises to you know best ways to exercise your heart all of these are called prompts and the more accurate the prompts are is the better the results you
get right this process is crucial for effectively utilizing AI models and ensuring that they produce the desired output and like for example instead of telling hey what are the best ways to exercise your heart if I tell it that um what are the best home workouts or RDO SQL health of people suffering from diabetes let's put this it will give a much accurate response here right so the more efficient you are in writing this particular prompts a better kind of results you get out of it now good prompts will help you with this they'll make
sure that the AI understands the commands and provides relevant responses they help reduce the chances of getting weird ass responses and they do allow you to harness the full potential of AI models by guiding them through specific tasks or problem solving right and how does a good prompt look like is something that we'll be discussing in the future but in general there are two parts to prompt engineering part number one is prompt defining part number two is called prompt refining in the future I'll in the future lessons I'll talk about each of these particular things
so yeah let's start with the first one which is chapter number six prompt defining so how do you define a prompt well defining on drafting the perfect prompt has three particular angles which is one Clarity and specificity to context and background information three desired output so I'll tell you a little bit more about the these first when I say Clarity and specificity an effective prompt should be clear and B specific it should not be vague like hey give me like 10 exercises to do right you have to be specific give me 10 exercises to do
and you know to improve my muscle mass and reduce weight here itself if I say what are uh how to work out what's the best way to exercise yeah let's see we do this right if I just write something like this the very first thing it says consult with the doctors or certified fitness professional right so let's say if I do this exercise um you know to improve my cardio vascular health at the age of 55. if I do this it can be like it's very specific right earlier the first response was consult with your
doctor or a certificate certified fitness professional here well start slowly and gradually increasing intensity aim for at least 150 minutes incorporate both cardiovascular and strength training exercises listen to your body all these things so the better your prompt is the more information that you put in the prompt the better it becomes right and it's very very important to provide context or the background information if you do this your prompting will go to the next level so let me go ahead and show you this right I recently turned 30. I have some belly fat and I
weigh let's say 78 kgs my height is 5 feet 9 inch or 9.5 9 inch yeah and uh ask I want to lose weight give me a weekly exercise plan weekly I would say 60 Minute exercise plan including or I would say work out plan including cardio uh including cardio I go to a gym nearby which has basic equipments or which I would say only weight training equipments like barbells and dumbbells um barbells tumbles and and machines to do pull ups and I would say what is it called dips pull-ups and depths um let's go
there let's see what happens all right wow we have full body workout 45 minutes cardio workout 60 Minutes you see it it's still not ready yet okay full body workout 45 minute I should do this squats bench press dumbbell rows overhead press cardio workout da da and I think day four it is giving me a rest day and day seven arrest today it is giving me uh a good uh one week uh plan so it's pretty cool right I actually kind of followed this um uh South Indian male so just I we igh 77 kgs
my height is 5.9 inch my age is 28 suggest me uh meal plan let's say South Indian and so that I lose wait 272 kgs so let's check what kind of results we get out of it yeah we got this right uh breakfast to idlis one boiled eggs I love this right it has given some sort of result so yeah so when you give a little bit of context on what exactly uh who you are and what kind of results you should get out of it and it gives the right kind of uh data out
of it right third is desired output format now I'll use the same prompt but I will change this a little bit better I want this male meal plan to be in a tabular form and should be a weekly meal plan and it should follow intermittent fasting uh I want to follow intermittent fasting give me a weekly meal plan for the same represent data and a tabular format so let's put this thing in let's see what happens all right so it has given some sort of data here but it says as an AI language model I
cannot provide personalized medical advice okay um okay it did have some restrictions but okay it did give some kind of result like 10 AM to 8 PM this is definitely not an intermittent fasting meal plan so let's say please okay meal plan that complies with intermittent fasting let's put this okay so now it's little bit better uh meal one meal to meal three but it did not give me any times so eight hours and fasting for remaining 16 hours blah blah blah I knew about this but it did not give me a fasting window right
obviously there is a little bit of uh if I put them give me also a fasting window the times when I eat it will give that as well but having Clarity and specificity context and background information and the kind of output that I want it has definitely brought in a lot of good amount of data we can improve on it of course later on but these are the three basic things that you want on top of it you can add some expert Persona like hey act as a fitness Fitness professional and give me this advice
or you can add an initial verb that we already write me this thing length and output type how long of a thing that I want whether it should be in a blog format should it be in a tabular format all those specific tasks and objectives what kind of desired outcome I want uh the particular tone in voice if I say right in the voice of let's say if I say um right in the voice of Elon Musk maybe we can create something yeah so it has kind of changed so tone and voice you can change
target audience you can change let's say write it for middle-aged main above 45 it it will give a different response out of it data and references if you want it to give you the right kind of data and reference it will give it out so other consideration you can talk about you're going to consider follow-up prompts if you got a particular result you can write one more from below that say tell me that you know write this in the form of a human it will change you can purpose you know change the purpose of the
goal of the prompt you can change that perceived difficulty of the prompt you can say write me and explain to me as if I am a five-year-old kid it will change subject matter based on you know you can ask them to act as you know a geology scientist and was expert in this particular subject and give you the right kind of response you can put in keywords especially when you are writing blog posts you can tell hey this is my main keyword and this should appear at least three times in this entire blog post and
they will it will definitely write it in that way ah you can input related sources suggested you can ask it for suggested sources you can ask it for additional nodes or you can like tell them hey it's a draft response or finalize the response all those things you can do with chat GPT itself right so this is what we call the as the prompt defining and the next uh lesson is going to be as interesting which is called The Prompt refining right and I'll tell you the thought process behind prompt refined so that you can
use it for your own business all that all right welcome to chapter number seven which is called The Prompt refining now that we have defined a prompt it's also important to like you know make it even better so what happens is whenever you are writing a prompt for example I have asked for this prompt and I changed this prompt to let's say you know let's say this is a new prompt that I'm putting in where I said hey you know what I'm a South Indian male this is what it is and give me a weekly
meal plan here right and let's say I also want to add on top of it hey give me um also suggest me weekly home workouts for losing weight and put the data and add tabular format along with meals so if I add this it talks about day uh you know and the meal plan and we have just added a new prompt to get this data instead my thought process is like why not just use this prompt along with this particular prompt which you can refine by clicking on this button and you know click on submit
and it will give you meals and let me see it has also given some home Accord plan if I say um [Music] same table with the meals it should also contain at what time I should work out let's see so this is called defining so here what he has done is like day one it has given me some times but it has not suggested okay it has not suggested uh the workout plan in the same table both work out and meal schedules at in the same table let's see if it works yeah here what you
are doing now it it's fine it looks good right Monday 12 P.M to at least 12 p.m caspas you have to work out Tuesday at 12 pm you work out maybe after workout it is asking us to eat it has taken up to Sunday so what we have just done is basically we kept on trying multiple prompts so that with one single prompt we get the most refined outcome out there right so this one prompt I can copy and save it as a five file and use it over and over again instead of constantly adding
follow-up prompts to make it better so that's exactly what prompt refining is here you analyze the ai's response that's the step one look for issues like some vagueness some inaccuracies and step number two you identifies what exactly can I improve so that I get a better result out of it right and third once you have revised it you make an adjustment add some prompts to you know to make it even better more crisper more sharper good context and step number four is when you revise the prompt see if it works the best right and you
can repeat the process to keep checking whether it works or not if I were to you know change it even better a little bit better right um let us say act as a fitness an instructor and I am your client and if I change this and boom we have the table and every time the table is different so I might want to change this and by the way once the GPD has stopped creating results just put continue and it will continue the next part yeah so this is exactly how you go ahead and refine the
particular prompt and repeat the process until you get the right result right and now let's say I go ahead and put in the input saying hey write me a blog post on how to do intermittent fasting so if you can read the language intermediate fasting has been gaining popularity as a way to improve health lose weight it looks very third person right I honestly don't like it so if I want to refine The Prompt I just click on this particular thing and I want to make it a little bit better let's say I say write
a you know 1000 or 2000 word blog post on how to do intermittent fasting and right now the language is not that great so if I say write an active voice and first person singular now this will make a little bit of sense so if I click on Save and submit all right if you are interested in trying intermediate fasting there are several thing you should know here are the steps I take when doing intermediate housing so definitely it is first person and singular right but it is still not great yet I'd be like let's
say I say start the introduction with uh emotionally with an emotionally engaging story uh personal story about how I lost weight uh with intermittent fasting and then engage reader throughout the blog with simple Indian examples write as if you are explaining to a 12th grade kid let's say I put this all right see intermediate fasting has been a life changing experience for me both physically and emotionally I struggled with my weight for years and tried various diet but nothing seemed to work it wasn't until I discovered this technique and I was finally saw results in
this blog post I'm going to share this see I love it now now you can see now if I hit continue it will continue this particular blog right step six step seven step eight and boom so we have a pretty nice Long blog so if I say write intermittent fasting for Indians right it in a listical istical format it will go ahead and start writing right understanding this start slowly choose the right food all right so it's pretty good right so this strategy what I just did you can see the number of iterations I did
with this particular prompt is called prompt refining so what did I do I analyzed the First Response it didn't look good I realized that the first key area of info Improvement was like it was not written in active voice it was like very vague bookish language so we wanted it to write an active voice so that's that third you we need to revise our prompt so we checked this was pretty cool we can go ahead with it and kept adding more and more details in the prompt itself so that it looks or feels much better
right next we tested the revised prompt and it seems like it's a pretty good uh you know blog post was a pretty good result so it has given us a cool summary and test the device from and repeat the process so okay so the previous format was actually pretty good right in a step by step format if I just do this now I figured out okay this prompt is the best front out there then I can go ahead and save this particular prompt later so that I don't have to keep on adding multiple follow-up prompts
here so I'll put continue if I say every step should have a emotionally engaging story it can add the story as well to make the blog even more better so this is the process of prompt refining the idea behind this is you go ahead and create the perfect prompt that works just for you and keep using it over and over and over again for uh you know your particular purposes so pretty slick huh in chapter number eight I'll talk about three tips that you should you can use to take your prompt engineering to the next
level all right chapter eight three tips to take you from the engineering to the next level here I'll give you three awesome tips that you can use to make it so much better that you know will go crazy the first one is ask explicit questions so just like you look at the example here that we used here write a 2000 word blog post on how to do this do this and this should be the introduction write as if you are working here and write in a step-by-step format very clear very direct which is why the
body that we got is a pretty good body so if I go ahead and add more input here let's say I say write uh and and informal tone and use emojis to make things fun and engage the reader throughout the blog if I just do this it will definitely add some emojis pretty much in every paragraph it is adding but pretty cool huh so when I added it it should write in an information an informal tone I it definitely you know added a little bit of magic here so pretty cool you can also ask GPT
to give you the response as if you are a 50 year old 12th class Cad and basically take on a personal for yourself so that you get the right kind of response example you know write as if you are explaining to a 12th grade kid this was an analogy so if you write change it to let's say uh fifth grade kid right so it will be even better hey there have you ever heard of intermediate fasting it's a weight loss strategy that has been gaining popularity in each recent years I know what you mean what
you might be thinking fasting that sounds scary let me tell you it doesn't have to be I'm an Indian just like you and I have successfully lost weight throughout intermediate fasting today I'm going to tell you everything that you need to know about this and I want to check if this kind of passes the grade grading test by Hemingway app just Google Hemmingway editor and we'll open this up and I'll remove this and paste this it says grade seven now okay so it's definitely not exactly grade fifth but it's pretty good I highlight I personally
like it let's say I take this entire thing and say I think it says it's grade 7 because there are a lot of sentences who are easy to read um three adverbs meeting the goal of three advances um four of the 16 sentences are hard to read one of the sentences should be you know very hard to read so let me say uh none okay make sure that none of the Sun tens says and more than I would say a little bit okay so how many words is this this is probably 10 words one two
three four five six seven eight nine ten eleven twelve thirteen I would say let's say none of the sentences are more than 10 words I do this all right now this will make things fun let me see if we can have good results out of it okay this is definitely not a 2 000 word blog post but let's just go ahead and use it to see uh what kind of grading level we are in see we got down to grade number four this is fantastic but still we got some sentences that are hard to read
maybe it was not possible to get them under 10 words but yeah so this was a pretty slick trick to make this blog even more readable and you can use this trick to make your blogs readable by AI detectors as well so AI we have content at Scales AI detector which is I I use it a lot it's called content at scale dot AI if I go ahead and check this AI for Content it says 80 percent highly likely to be human so pretty good we kind of did it right so it took up to
step number okay it took the entire thing but yeah when you have a very good prompt you get very good results that's that's what I keep saying so now we also added include examples and analogies specify response format or structure guide the AI how to structure the response for example list five benefits of meditation along with a short explanation for each I can do that ah let's say list five benefits of intermittent fasting explaining the examples of each and 100 words okay he says hey there kiddos okay so we got five examples but we did
not add heading tags I can go ahead and change it and make it even better so I hope you got some idea on how exactly you can tweak around the prompts to get very high quality detailed results out of it right this is an example of a very good prompt the more detailed it is the better as you can see it says act as an expert cold email copywriter and I am your client I am a marketing agency owner who only works with SAS companies create five different personalized code outage email that I should send
to my target audience so that I can get more booked appointments to deliver my services the subject line should be emotionally engaging so my target audience opens my email it should be no more than three words it should be a pattern interrupt or an open loop this is a copywriting term the body should be written actively and compelling enough to get me an email response it should be no more than 100 words it should highlight my target audience pain point and add a one line case study to show that we are worthy of having a
conversation having a soft CTA asking the prospect to reply to my mail the writing style is professional yet fun to read but approachable and targeted towards this target audience add a cool signature email signature that is fun to read and then I left blank spaces to mention my target audience the pain points and the case study and have a look how it works here I have actually added my core offer is I helps as Founders to rank on Google within four months I'll remove the audience pain points let's see and case study says that I
have ranked uh the most difficult for Astra in the first page of Google I would say for my last client last client I used I ranked very difficult keywords Within a couple weeks and achieved thousand plus quality visitors so yeah Revelation Revolution is your SAS is is getting enough visibility Skyrocket is a success take your sex to new heights want to rack on Google oh wow I love this great great subject line beautiful so if I say does your ad suck so that can be a good subject line as well but yeah as you can
see this particular prompt is possibly the most uh important prompt when it comes to like cold emailing so I very clearly mentioned what exactly I want and give me a couple of offers so let's say um I removed the case study here's my offer pain point and and should clarify why I'd be a great fit for them save and submit oh wow hi there I came across your company and saw that you're offering amazing SAS solution but I also notice that your website is not ranking high enough on Google an expert in SEO I can
help you rank in a website in just four months I have health okay so it has added some sort of credibility okay let's I say I don't have any case studies to offer let's have a cool experiment here let's see what happens uh do you have the expertise and proven track record of success okay it is not great I'll just use Chad gpt4 to see what kind of responses I get here let's say I say body should be no more than 50 words make it even more challenging for our body here quick win alert struggling
with your SAS company's visibility on Google my marketing agency specializes in helping SAS Founders rank in just four months let's connect and discuss how I can help you boost your online presence interested I will use this this is pretty fun it's a respect child marketing magician in my company this this looks really really fun so I'll use it so gbt4 is much much better uh if I can say right I personally love GPT 4 more than any other kind of AIS so time for some bonuses so welcome back to chapter nine I'm so happy that
you have completed till date till this particular Point what I've been doing is I have been publishing a lot of content around chat GPT and for free and uh this have all these are covering many many use cases around chat GPT right so here are a couple of bonuses that you can take a look and use it for your own business or your marketing purposes the first one is how to use strategypt for YouTube growth right so this video will show you how exactly you can use it to create thumbnails how you can use chat
GB to create your YouTube headlines description all of those and uh I mean there are two videos one is on description and one is on thumbnails I'll give you the links to that in the description description below and then I'll also show you how to create a buyer's Persona with chat GPD whatever marketing campaign you run you need a proper proper buyer Persona so this particular video will allow you to create your own buyer Persona with Eerie details with very very large details so that you can use it for a marketing campaign building websites building
landing pages all sorts of things right next if you are somebody who wants to build a high ticket funnel this video will help you craft everything from scratch this includes writing headlines landing page copies vsl copies the soundtrack you should be using in the vsl the Facebook ad copies and everything in there which you will require to build a complete High ticket funnel so I hope this one helps you a lot next uh the next video the next bonus is creating a video sales letter with charged gpt4 it is very very smart and it can
help you craft your own personal video sales letter that you can use for an appointment generation funnel so this is all for you link is in the description or below this particular video um it I will also give you how exactly you can create a grand slam offer how exact and not just that there are multiple bonuses like how you can create your own offer for an agency or a coaching business how can create your own AI app and uh how you to write undetectable AI content with chat GPT so that you can rank your
blogs you know very very fast and produce content very very fast right these are my selfish ways to make you like me and I hope that these help you a lot in return I only ask for one subscribe and one like so that this helps you a lot right