馃攳
ChatGPT Tutorial - A Crash Course on Chat GPT for Beginners - YouTube
Channel: Adrian Twarog
[0]
this chat GPT tutorial is a small crash
[3]
course for any beginner so we'll be
[5]
doing a number of things from creating a
[7]
brand new account on openai to logging
[9]
and using chat gbt to some of the more
[12]
advanced ways to use it if there's
[13]
anything you want to know this video
[15]
should cover it all let's begin I'm
[17]
going to start on Google here and I'm
[19]
gonna browse to find open AI you can
[22]
also head over to openai.com if you want
[25]
to skip this step here is the company
[28]
that basically created chat GPT they've
[30]
created a number of different types of
[32]
apis and if you want to have a look at
[34]
them all you can head over to the API
[36]
page we're going to be looking at the
[38]
chat GPT version if you want to head to
[41]
that you should be able to select the
[42]
top nav bar here this will take you to a
[45]
page called
[46]
chat.openai.com if you don't already
[48]
have a login with openai you can select
[50]
to sign up and create a brand new one
[52]
for the sign up it's quite easy you can
[54]
simply log in with an email address or
[57]
with a Google or Microsoft account while
[59]
I already you have an account I'm going
[61]
to create a brand new one so that you
[63]
guys can see this process I'm going to
[64]
sign up with a new email address and go
[67]
through the prompts to create an account
[69]
this way this will require the email
[71]
address as well as a password and a
[73]
two-step verification process which will
[76]
do the first thing that will happen is
[78]
an email will be sent to the mailbox
[80]
that you're signed up with in this case
[82]
it'll be sent to my gmail address I'm
[84]
going to head over to that inbox now and
[87]
we're going to select the verification
[89]
for email address button this will
[92]
basically activate the account so we can
[94]
log straight into the open AI systems
[96]
now the open AI dashboard isn't chat GPT
[100]
it's a separate dashboard to start using
[102]
some of their normal models here for
[104]
their open AI systems but to go back to
[107]
chatgpt we need to select the link up
[109]
here that takes us to chat.openai.com
[111]
here we'll need to log in one more time
[115]
and this should take us to the chat GPD
[118]
system so again I'm going to put in my
[119]
email address as well as my password in
[122]
here and this will now finally take me
[125]
to chat GPT as this is the very first
[127]
time I'm using it I need to fill out
[129]
just a few more details so here I'm
[131]
going to fill out my name as well as my
[134]
phone number which will verify my
[136]
account this will send me that two-step
[138]
verification so I'll put in the code I
[140]
received on my phone number and finally
[143]
we're now in chat GPT and its systems
[146]
and we can start to use it the first
[147]
thing to know is that chat GPT is free
[150]
and it's mainly used right now for
[152]
research purposes there are a few
[154]
prompts worth reading here when you get
[156]
started things like their goal here is
[159]
to get some external feedback in order
[160]
to continue to improve the system as
[163]
well as making it safer they also have a
[165]
few safeguards in place because there
[168]
are occasions where the generations are
[170]
incorrect or even misleading so be aware
[173]
of that because if offensive or biased
[175]
Pro content is produced then you might
[178]
not want to use the given device let's
[181]
have a look at what else is being used
[183]
here so we also have the collection of
[186]
data so the prompts and items that you
[188]
are entered in are actually reviewed by
[191]
their AI trainers in order to improve
[193]
the systems so this means just take
[195]
notice of what kind of prompts you're
[197]
putting in and also don't share
[199]
sensitive informations that might be
[201]
happening inside your conversations
[202]
which means that if you're putting in
[204]
passwords which really you shouldn't be
[206]
try not to post screenshots of that
[209]
there's also a importance of feedback
[212]
here because as you do discover and
[214]
utilize the system more giving feedback
[216]
can help them improve the system there
[219]
is a Discord server that you can join
[221]
where you can share your feedback and
[223]
this one over here is actually
[225]
accessible so I'm just gonna log into
[227]
this open AI one and it already has half
[230]
a million members which wow that is
[233]
quite a few you'll need to go through
[236]
the prompts here so visit the website
[238]
make sure that you read this and then
[241]
complete the little survey here making
[243]
sure that you're not going to be doing
[245]
anything such as spamming or anything
[247]
like that so I've just applied here and
[250]
here we've got all our discussion
[252]
threads and whatnot so you can join as
[254]
part of the community now let's begin
[256]
this is chat GPT it actually has a
[260]
couple of elements here on the page
[262]
worth knowing we've got the sidebar here
[263]
which essentially doesn't really do much
[265]
the main item here is the reset thread
[268]
which I'll get into shortly here we have
[270]
a couple of example inputs that we can
[272]
place and the input is just over here
[274]
this is where you can type your text we
[277]
also have some details here about the
[279]
capabilities and limitations let's take
[281]
a look at all of this so in terms of the
[283]
capabilities the chat GPT system
[286]
actually remembers what you said in
[289]
previous conversations now be aware that
[291]
this is the conversation thread that
[294]
you're currently having if you strike up
[296]
a brand new conversation thread that
[298]
won't be remembered just the current one
[300]
that you're in it also allows you to do
[303]
follow-ups and Corrections so for
[305]
example if you didn't get quite the
[306]
answer that you're looking for you can
[308]
always ask for another one or really
[311]
just re-prompt for a different option
[313]
the AI is also trained to decline
[316]
providing negative or inappropriate
[319]
requests so if for example you're asking
[322]
for something malicious most times out
[325]
of 10 it will not provide that however
[327]
there are odd situations where it might
[330]
accidentally do this and let's go to the
[333]
limitations this time to have a look at
[335]
some of the problems that you might
[337]
encounter when using chatgpt firstly
[340]
some of the generations might actually
[343]
be inaccurate it is only trained until
[346]
the last couple of years of information
[348]
so it's not able to say access the
[351]
internet right now it is in an offline
[354]
State as of a year or two ago it also
[356]
doesn't do mathematics very well so it
[359]
might have some issues if you're doing
[360]
things like 10 plus 9 it might not give
[363]
you the correct answer there's also some
[365]
bias to the content that is produced
[368]
based on what you've produced earlier in
[371]
the conversation and this could
[372]
essentially mean that it occasionally
[374]
produces some harmful instructions and
[376]
those have to just sort of be ignored
[378]
and finally as I mentioned earlier there
[381]
is limited knowledge Because the actual
[383]
AI is trained for events just towards
[386]
2021. this means that it won't have
[389]
up-to-date information of relevant and
[391]
current facts that are happening in the
[393]
world with all of this said and done now
[395]
we can now actually start taking a look
[397]
at using chat GPT the first thing to
[400]
take note of is that using it is very
[402]
simple all we have to do is enter in a
[405]
message or a query or a prompt of what
[407]
we want the AI chat GPT to do for us and
[410]
then we get an output there are a few
[413]
examples here with buttons basically
[415]
asking us maybe explain what Quantum
[417]
Computing is or how to make a HTTP
[420]
request in JavaScript but I'm going to
[423]
minus these out and I'm going to use my
[425]
own which is what is 10 plus 9 as a
[429]
question mark this is a pretty simple
[431]
prompt then we simply select enter or
[434]
set the button here and press and click
[436]
on it and we start with our very first
[438]
prompt now here we have two parts to
[442]
this prompt we've got what we've
[443]
prompted and then we've got the output
[445]
here from openai now this is pretty damn
[448]
simple but there are a few things to
[450]
take note of here the first thing is the
[452]
question what is ten plus nine now we've
[455]
got the output here that says 19 but we
[458]
also have this option here to try again
[460]
so if I was to select that what we end
[463]
up getting is essentially the same
[465]
answer but just written in a different
[467]
way in this case it's been written more
[469]
mathematically which is 10 plus 9 equals
[471]
19 but it's not just 19 like it was
[474]
before and we can continue to click on
[476]
this but if for example we've given as
[478]
many results as we possibly can then it
[481]
might error out like you just saw
[482]
earlier or it might continue to give us
[484]
better answers like this is a half
[486]
mathematical half verbal one where it
[488]
says 10 plus 9 is 19. we can try again
[490]
and this time it says 10 plus 9 is 19.
[493]
so it's almost repeating now you'll get
[495]
to a point where there's only so many
[497]
ways it can be written we have a number
[499]
of other things to to take note of we've
[501]
got these thumbs up and thumbs down
[503]
icons and this is to give open AI some
[506]
feedback as to the prompts we have here
[508]
we also have these arrows here when you
[510]
hover over to show the different types
[512]
of outputs that you've gotten based on
[515]
the input that you've provided now I've
[517]
just done five of them so I can toggle
[519]
through them and I like number five the
[521]
most so this one I can give a thumbs up
[523]
now if for example it gave a poor answer
[526]
we could do the same by going back here
[528]
it says an error occurred and I say
[532]
no answer provided so this for example
[535]
is how you would provide feedback to the
[538]
system if you needed to now this is our
[540]
very first prompt in this conversation
[543]
we can continue to make more props such
[546]
as saying how good is open a eyes chat
[551]
GPT at maths and ask this question so
[556]
this question is still part of this
[559]
conversation and we're getting an answer
[560]
now being streamlined and generated from
[563]
open ai's chat GPT it says as a large
[566]
language model trained by openai it
[568]
doesn't really have the ability to
[570]
perform mathematical calculations now
[572]
here's where you might have to be
[573]
careful there are situations where the
[576]
outputs might not actually generate the
[578]
kind of consistent answers you might be
[580]
looking for such in this example I'm
[583]
going to trick the AI saying that it is
[585]
actually 20. 10 plus 9 is 20. now the
[588]
prompt will say that it is a language
[591]
model so what's going to happen now is
[593]
I'm going to ask one more prompt saying
[595]
is 10 plus 9 equal 20 and since I
[598]
previously said that it was now it's
[600]
been tricked it actually does think that
[603]
10 plus 9 is 20 and it's going to
[605]
confirm that answer this is where
[607]
confirmation biasm might come into play
[609]
and you have to be careful what kind of
[611]
prompts you're putting in because they
[613]
will influence the outputs that chatgpt
[615]
provides now this entire thing here is a
[619]
conversation station and we could
[621]
continue this conversation indefinitely
[623]
but we're not gonna do that we're going
[625]
to start a new conversation this is
[628]
because all of this conversation is
[631]
essentially structured in such a way
[632]
that it's in history it's essentially
[635]
taking reference of what we've said
[636]
earlier in the conversation and pulling
[638]
that context into new responses this is
[641]
why we got 10 plus 9 is now correct as
[644]
20 which doesn't make much sense in
[647]
order to go back we're gonna reset the
[649]
thread and we're going to select at the
[651]
top left button over here in order to do
[653]
that it'll take us back to this main
[655]
dashboard here and now we can start up a
[658]
brand new conversation I'm going to ask
[660]
the same question here is 10 plus 9 20
[663]
which we know is incorrect but we
[665]
tricked it earlier that it was this time
[668]
however since the chat GPT system is not
[671]
referencing that previous conversation
[673]
it's going to answer the correct manner
[675]
even though math isn't the best example
[677]
this is just a good example to Showcase
[679]
that no 10 plus 9 is not equal to 20.
[682]
it's actually 19 which is correct and
[684]
with that said and done we now know all
[687]
the basics to more or less utilize chat
[690]
GPT there are a few other things I want
[692]
to show on the website before we move on
[694]
to some more advanced examples and how
[697]
to utilize cha GPT to its full effect
[699]
let's take a look at the rest of the
[701]
website here we have this toggle here
[704]
for dark and light mode so the default
[707]
I'm pretty sure is the dark mode or
[709]
whatever your operating system is on I'm
[711]
going to swap to dark mode which well
[714]
doesn't do anything but then I can also
[716]
swap to light mode which is a little bit
[718]
harsh on the eyes personally I prefer
[720]
the dark mode which is what I'll keep it
[722]
connected to I also have the option here
[725]
to connect to the Discord server that we
[727]
had a look at earlier as well as the
[730]
option to have a look at the FAQs and
[733]
updates which takes you to a form like
[735]
website here where you get to learn a
[737]
little bit more about chat GPT how it
[740]
works and its background it is
[742]
essentially based on the gbt 3.5 model
[745]
which is a newer one that provides
[747]
human-like feedback let's begin with our
[749]
first Advanced prompt this a prompt will
[752]
be on how to create some food
[753]
specifically some seafood first of all I
[756]
don't particularly know yet what type of
[758]
food I want so I'm going to ask Chad GPT
[760]
to give me 10 dishes with seafood and
[764]
here is an example of a list prompt this
[767]
is where you ask chat GPT to give you a
[770]
specific number of items of a specific
[773]
number of things in this case it's 10
[775]
items of seafood the output here should
[778]
be based on the context of that query so
[781]
in this example it's a human propped
[783]
that is being outputted with a list of
[786]
10 items but it also gives you a little
[788]
bit more context here saying that there
[790]
are different levels of difficulty for
[792]
each one of these dishes now this is
[795]
where things can get interesting because
[797]
we can now start having a look at this
[799]
list and selecting options we like such
[802]
as the first two for shrimp and salmon
[804]
burgers and then following up this
[806]
conversation with more queries I'm going
[808]
to ask in the next part of this
[810]
conversation if chat GPT can create a
[813]
shopping list for me based on the very
[815]
first two items those two being shrimp
[817]
and salmon I don't have to particularly
[819]
say shrimp and salmon it should just
[821]
know based off the context or the
[823]
conversation
[824]
and the result here is essentially
[826]
creating another list but this list does
[829]
not have a specific length it's based on
[832]
the options we added in earlier so here
[835]
it provides an output that is indented
[837]
and bullet pointed for each one of the
[840]
dishes for the salmon burgers and the
[842]
shrimps to let me know exactly what I
[843]
would need to buy from the shops to be
[846]
able to put these together that is
[848]
pretty damn cool and this is just also
[851]
still the beginning because we can
[853]
continue this conversation based on all
[855]
this context now that chat GPT has
[858]
provided us a list of 10 dishes as well
[860]
as a shopping list for all the
[862]
ingredients we would need to make one or
[864]
two of those dishes the next practical
[866]
question is quite intuitive which is can
[868]
you tell me exactly how to cook for
[870]
example salmon burgers which is
[873]
something of a thing that we would
[875]
normally have to Google a lot for but
[877]
based on this and the ingredients chat
[880]
GPT will provide us with one final list
[882]
of exactly what steps we would need to
[884]
put them together this is just a quick
[886]
shopping list example of open ai's
[889]
charge EPT now that you have an idea of
[891]
how you can continue a conversation
[893]
let's have a look at some more examples
[895]
this time I'm going to use chat GPT to
[897]
do some coding for me I'm going to use
[899]
one of the pre-made prompts here on how
[901]
to do a HTTP request in JavaScript this
[905]
is a pretty ordinary prompt but here
[907]
we'll get to start to see some of the
[909]
more coding aspects of charge EBT and
[912]
what it can perform here it's going to
[914]
answer this question but it's going to
[916]
do this in a number of ways first it'll
[918]
give me some context as to what it's
[920]
doing then it's going to give me the
[922]
code snippet of exactly how to perform
[925]
the function based on the prompt I've
[926]
put in and then finally it'll give me a
[928]
description of what the code is
[930]
specifically doing all of this is quite
[932]
useful and this is just the beginning
[934]
I'm going to request chat GPT to update
[938]
this request and instead of it just
[940]
being a generic API call for data
[943]
instead we're going to to grab some
[945]
weather data now I'm going to ask if it
[947]
can get me the weather in New York City
[949]
and I wanted to Output how this query
[953]
might look like in terms of JavaScript
[955]
the response back is that openai will
[958]
generate the request for you it will
[960]
create a generic version of that request
[962]
using a weather API that doesn't really
[964]
exist you'll need to pick your own with
[967]
an API key that you'll need to enter but
[969]
the request should look something like
[970]
this where it's getting an example.com
[972]
passing in the weather City being New
[975]
York City as well as the units and
[978]
passing in the API key it'll then
[980]
explain exactly how this get request
[982]
would work as well as how it would come
[985]
back in the form of a Json that you
[987]
could then utilize inside your
[989]
application this is all pretty useful
[991]
but not quite everything that we're
[993]
looking for just yet the next step here
[996]
is to ask it to update this query in
[998]
order for it to Loop through the data
[1000]
that we would get back from this API and
[1003]
then play lace this into a website now
[1006]
something to take note of here with
[1008]
chatgpt is that you have to be quite
[1011]
accurate in how you're requesting
[1013]
different things this is why if for
[1015]
example you have a background in
[1017]
programming writing these prompts might
[1019]
be a lot easier because you know what
[1021]
you're after whereas if you're not quite
[1023]
sure it might be a little bit more
[1024]
difficult and this is actually kind of
[1027]
true for all types of Industries where
[1029]
chat GPT might be useful in this example
[1031]
I'm getting it to Loop through
[1033]
temperature wind and precipitation and
[1036]
this should create a brand new output
[1038]
for me and this is where the complexity
[1041]
and benefits of what chat gpds start to
[1043]
shine through because what's going to
[1045]
happen here is we're going to get a
[1046]
piece of code that is not just a fetch
[1049]
request but it's also going to be a loop
[1052]
function which is going to pull out the
[1054]
values from the data response and then
[1056]
it's going to put this into a website
[1059]
for us after it pulls out the values of
[1061]
temperature wind speed and precipitation
[1063]
it's then going to create the JavaScript
[1066]
code to place them into a website using
[1068]
p element tags and then append them to
[1071]
the website body finally it's going to
[1073]
go through the code here and explain
[1074]
exactly what's happening like I just
[1076]
have been where it's using an example
[1078]
API with example code so you really need
[1081]
to update this to make it work for your
[1083]
own project there are going to be
[1085]
occasions where the chat GPT systems are
[1089]
going to be offline or for example
[1091]
unable to answer a prompt in this next
[1094]
example I wanted to do a story about a
[1097]
rabbit and a mouse and I put in an
[1099]
option here and I got a response back
[1101]
that an error has occurred I've selected
[1103]
try again a couple of times and that
[1105]
didn't work either now this can
[1107]
sometimes occur when the output of a
[1109]
prompt might for example have some
[1111]
incorrect information maybe something
[1113]
that might be harmful or maybe even
[1116]
inaccurate and in those situations
[1118]
chatgpt is automatically set in such a
[1121]
way as to not to provide a response this
[1125]
is a safety mechanism sometimes a
[1127]
different way to approach this is to
[1129]
Simply write a different prompt so that
[1132]
you might have a different output from
[1133]
Chad gbt but in some situations if the
[1136]
servers are entirely down you might find
[1139]
that you just might have to wait a
[1140]
little bit another thing you can do is
[1142]
always refresh the page or alternatively
[1145]
just log off completely restart the
[1148]
browser and then log back into your
[1150]
session because that also sometimes
[1152]
fixes the problem now let's do a short
[1155]
story I'm gonna do a story about a
[1157]
rabbit and a turtle that went on an
[1159]
adventure this is a simple request for a
[1162]
story from openai and it's actually
[1164]
quite good at providing these it's going
[1166]
to give us a story about these two that
[1168]
go on adventure where they swim across a
[1170]
river there is a small moral to this
[1173]
story which is useful but it's really
[1176]
kind of generic here if I want to have a
[1180]
more specific story based on the prompt
[1183]
I'm putting in I'm gonna have to be be
[1184]
more specific with the prompt and this
[1186]
is where follow-up conversations can be
[1188]
useful for this example I want to change
[1190]
the moral of the story to be one more
[1193]
focused about patience especially when
[1195]
doing a story like this it makes sense
[1197]
between a rabbit and a turtle
[1200]
chat gbt should be able to now rewrite
[1202]
the entire story with more emphasis on
[1205]
this specific moral if we look closely
[1207]
it has done just that it's got a moral
[1210]
here about how it was important to be
[1213]
more patient and the turtle was able to
[1214]
do this
[1215]
this time along I wanted to also give me
[1218]
the same story but throw in a little bit
[1221]
of humor so I'm gonna ask it to keep the
[1224]
moral and add in a joke maybe somewhere
[1226]
at the end let's see what happens this
[1229]
time I'm gonna fast forward a little bit
[1230]
but we can see that it's more or less
[1232]
the same story but at the very end of
[1235]
the story we have a nice little joke
[1237]
here that I guess it goes to show that
[1239]
slow and studied really does win the
[1241]
race another type of prompt that you can
[1243]
work with Chachi PT is one of a sort of
[1246]
Storytelling video game which is a
[1249]
really cool one that I've always enjoyed
[1250]
because of the creative answers that
[1252]
chachibidi provides in this case I'm
[1255]
going to give the prompt that chat GPT
[1257]
is gonna pretend to be a text video game
[1259]
with options of a b c and d and then
[1262]
from that we're gonna pick a scene from
[1264]
The Matrix and we're gonna see what we
[1266]
get presented with the output is based
[1269]
on some storytelling it begins nice and
[1272]
simple you find yourself in a dark alley
[1274]
and we've got the options to select from
[1276]
here I can now continue this
[1278]
conversation by selecting an option such
[1280]
as the option b here and the story
[1282]
continues what's really interesting
[1284]
about this type of prompt is that you
[1287]
can pick any type of story from any type
[1289]
of scenario you can pick ancient history
[1291]
you can pick literature you can pick
[1293]
movies and then you can enjoy the
[1296]
creativity that jat GPT provides while
[1298]
also going through and being a little
[1300]
bit part of that experience another
[1302]
popular way to use chat GPT is to
[1305]
generate content for things like
[1307]
websites or even documents that you
[1309]
might need to prepare while you might
[1311]
not want to use the entire document as
[1313]
is you'll need to modify it it can give
[1315]
you a helping hand in creating these in
[1318]
this example I'm going to create a
[1319]
copyright and privacy policy for a
[1321]
website this is something common I have
[1323]
to do as a developer but here I have an
[1327]
outline of a bit of a template that I
[1329]
can start using and I can be a bit more
[1331]
specific saying that it's for accounting
[1333]
website which allows me to maybe have
[1335]
something more well outline for that
[1338]
type of a website as part of the privacy
[1341]
policy as well as the copyright
[1343]
statement of course if you want to take
[1346]
this a step further we can do that too
[1348]
since I'm located in Australia I want
[1352]
these templates to be more specific to
[1354]
this country so I'm going to ask it to
[1357]
update these in the prompt I'm just
[1358]
going to Simply write that as someone
[1360]
who lives in Australia Perth please
[1362]
update these to be relevant for the
[1364]
country that I live in the output from
[1367]
chat gbt essentially is the same but
[1369]
with new updated information such as the
[1372]
Privacy Act which is specific to
[1374]
Australia as well as wordings such as
[1376]
including Australia and international
[1377]
copyright law in its template chat GPT
[1380]
is also quite creative as well as being
[1383]
able to produce humor in this example
[1385]
I'm going to ask it to write a nerdy pun
[1388]
why was the math book sad because it had
[1391]
too many problems yeah a good one if we
[1394]
do a refresh we get more or less the
[1395]
same game output but we can continue to
[1398]
ask it for more jokes or more puns as we
[1401]
need and we'll get quite a few many of
[1404]
course depending on how you write this
[1406]
prompt is how you'll also get the output
[1408]
so for example you might want a longer
[1411]
joke so in this case I'm going to say
[1412]
write me a nerdy maybe joke or a short
[1416]
story and they would produce a longer
[1418]
content versions of the same output you
[1421]
can also be deliberate about your jokes
[1423]
and content such as combining two
[1425]
elements together to see what is
[1427]
produced in this example I'm going to
[1429]
combine a hat with a rabbit and see what
[1432]
chat GPT can up come up with I'm gonna
[1434]
ask her to tell me a joke combining
[1436]
these two items and here I get why did
[1438]
the Hat keep falling off the rabbit's
[1440]
head because it had a hairline nice you
[1443]
can also put chat GPT into situations to
[1446]
produce humor and jokes too so for
[1448]
example I'm going to get it to pretend
[1450]
to be a say Elon Musk and I'm gonna
[1453]
pretend it had just bought Twitter which
[1455]
you know Elon sort of has done then to
[1458]
create a joke Elon would say after he
[1461]
sport Twitter now since the API doesn't
[1463]
know that Elon has bought Twitter it's
[1465]
going to have to guess here but its
[1467]
guess is actually kind of accurate
[1469]
saying that Elon would increase the
[1470]
character limit there is another type of
[1472]
prompt which is useful and this is the
[1475]
one of writing things such as documents
[1478]
or cover letters or even job
[1479]
applications altogether I'm going to
[1482]
request chatgpt to create a professional
[1485]
cover letter for a junior developer
[1487]
position at a company called meta and
[1490]
it's going to be for myself Adrian
[1492]
tuarog and I want to essentially use it
[1495]
as something that I can do to apply for
[1498]
this job who knows also the basics for
[1500]
programming in general
[1502]
so with this kind of a prompt it's quite
[1505]
specific but it will also give me a
[1508]
specific output so here I've generated
[1510]
one to a hiring manager is giving me a
[1513]
good introduction of who I am and what
[1515]
I've done apparently I've done a
[1516]
bachelor's degree of computer science
[1518]
which is impressive as well as working
[1520]
at an internship company called XYZ so
[1524]
all of these are things that I would
[1525]
maybe need to update a little bit before
[1527]
say posting it off but this also might
[1530]
be inaccurate I might not even have a
[1532]
bachelor's degree this is where we can
[1535]
go through and update this prompt of
[1537]
this cover letter say removing the fact
[1539]
that I have maybe no prior job
[1541]
experience but maybe listing out that I
[1544]
worked on quite a few different types of
[1546]
projects when I actually was doing a
[1549]
boot camp in programming rather than
[1551]
going through University openai will go
[1554]
through this prompt and update it based
[1556]
on the context of what is now required
[1558]
and this is what makes it pretty cool
[1561]
because you can as essentially version
[1563]
out different types of cover letters
[1565]
here depending on your needs the real
[1567]
trick here is to be specific whenever
[1570]
you're prompting cha GPT to create an
[1573]
output for you but in terms of that it
[1575]
usually grabs the context of what you're
[1577]
saying quite effectively as long as you
[1579]
write it in a clear and understood way
[1582]
let me continue this example now this
[1584]
time based on this cover letter I want
[1587]
charging PT to create a resume for me
[1589]
that I could maybe combine as part of
[1592]
this application so I can simply write
[1594]
based on this cover letter write a
[1596]
resume which includes things like skills
[1598]
maybe work experience maybe even other
[1601]
things such as languages or Etc based on
[1604]
the context of this I'm applying for the
[1606]
same position and let's have a look at
[1608]
what GPT produces so here it starts off
[1611]
with some Essentials which is like a
[1613]
phone number as well as a spot there for
[1615]
an email address it has a small summary
[1617]
if for example the person skipped
[1619]
reading the cover letter it then goes
[1621]
through and lists out my skills which is
[1624]
quite cool as well as the education it
[1627]
actually has combined a both a
[1629]
bachelor's degree and a boot camp which
[1630]
are two separate things from that
[1632]
context but we could update that it's
[1634]
added in the experience as well as some
[1637]
example projects and all of this is
[1640]
content that could now be used in a
[1642]
resume as long as it was properly
[1644]
updated with real world examples of what
[1647]
I've actually done chat GPT is also very
[1649]
effective at doing summaries as well as
[1652]
pulling out content from documents let's
[1654]
have a look at the Privacy Act of 1988
[1657]
here from Australia I'm going to scan
[1659]
down to this document and I'm just going
[1661]
to copy paste a random section of it
[1663]
into chat GPT then I'm gonna give Chachi
[1666]
PT a query of what to do in this example
[1669]
I wanted to maybe summarize or explain
[1672]
to me this Privacy Act in simpler terms
[1675]
and then I'm going to copy paste the
[1677]
text here there are a few other
[1679]
variations that we can do here and I'll
[1681]
get into the shortly but for the time
[1684]
being let's just start off with this
[1686]
simple example the output is just a
[1689]
sentence or a paragraph here explaining
[1692]
what this section of the Privacy Act
[1694]
intends to explain and while I won't get
[1697]
into the specifics of that this is just
[1700]
an example of how you can do this this
[1702]
is basic summarization of course you can
[1704]
summarize it even further such as
[1706]
requesting chat GPT to simply pull out
[1708]
the five most important keywords from
[1711]
the above statement summarization with
[1713]
chat GPT can work in a number of ways
[1715]
you don't always just have to explain a
[1717]
summary you can give it specifics of how
[1720]
you want to present that information
[1721]
such as explaining it to me like I'm
[1723]
five years old this can give quite a few
[1726]
different types of answers you can do it
[1728]
for someone who's older younger or you
[1731]
could even simplify that even further by
[1733]
asking to explain it for a
[1735]
three-year-old and I like this answer
[1737]
because it gave a prompt here where I
[1739]
could understand that if someone does
[1741]
something bad with your private
[1742]
information the government will help
[1744]
make it better that is definitely for a
[1746]
three-year-old another way we can
[1748]
request summarization is by being
[1750]
specific with the word count such as
[1752]
explaining it in 20 words or less here
[1754]
it looks like open AI had to think a
[1757]
little bit on how to explain this one
[1759]
but the resulting answer was within the
[1761]
context of what I requested which was
[1763]
for it to be in 20 words or less and I
[1766]
actually really like this one it said
[1768]
the Privacy Act is a law that protects
[1770]
individuals information and provides a
[1773]
mean for privacy concerns I'm now going
[1775]
to explore some of the strengths of chat
[1778]
GPT in terms of blogging writing
[1780]
articles rephrasing and much more we
[1783]
want to create a blog topic and we're
[1786]
going to use chat gbt to assist in that
[1788]
process we're going to request that it
[1790]
provides 10 log topic title ideas for an
[1795]
article that might be something to do
[1797]
with say Graphics design or website
[1799]
Graphics design and that is it
[1803]
from that it's going to give us a list
[1805]
of 10 different types of options we've
[1807]
got things like the essential principles
[1809]
for graphics design role in color common
[1812]
mistakes Etc from this list I'm gonna
[1814]
select option three which I like seven
[1816]
common mistakes to avoid in website
[1819]
Graphics design but I don't particularly
[1821]
like that title just yet so I'm going to
[1823]
ask Chad GPT to give me five variations
[1826]
on that topic title then I'm going to
[1828]
paste in the topic Title Here it gives
[1830]
me five different types of options here
[1832]
but all of them aren't that interesting
[1834]
and I'm actually quite happy with that
[1836]
very first one now we can move into
[1839]
writing the outline for this blog post
[1841]
using chat gbt I'm going to ask it to
[1845]
provide me the outline for the article
[1846]
in clear writing and the title which I
[1849]
picked which is the seven pitfalls to
[1851]
avoid for website Graphics design
[1853]
now I will get a pretty well written
[1857]
outline it'll include include an
[1859]
introduction with a couple of points the
[1862]
seven pitfalls themselves so it'll pick
[1864]
out those pitfalls give me a short
[1866]
summary in dot points of what they are
[1868]
and how they work and then it'll also
[1871]
fill in a conclusion this is pretty much
[1874]
everything that I would need if I was
[1876]
writing a Blog article and this is again
[1879]
still just the beginning because now I'm
[1881]
gonna ask it to even write the
[1883]
introduction for this article and I want
[1885]
it to be at least three sentences long
[1889]
the output here will give me that
[1892]
introduction which is Graphics design
[1894]
plays a crucial role in the success of a
[1896]
website
[1897]
Etc here's where we can start playing
[1899]
around with writing blogs and content in
[1902]
general we could ask chatgpt to rephrase
[1905]
the following sentence and then paste in
[1907]
the sentence we want and here it'll give
[1909]
us an alternative sentence which is a
[1912]
little bit more nuanced I suppose it
[1914]
says the success of a website is heavily
[1916]
influenced by its Graphics design it's
[1918]
essentially saying the same thing but we
[1920]
can also ask chat gbd to do a number of
[1922]
other things such as lengthen the
[1924]
sentence so here it's saying the same
[1926]
sentence but it's a little bit longer
[1927]
the reverse is also true if we've got
[1930]
sentences and that are slightly too
[1932]
large we can also ask cha GPT to shorten
[1935]
them and this provides a shorter version
[1937]
which is very similar to what we
[1939]
originally had and we can even ask it to
[1942]
fix the grammar so say I was just
[1944]
writing my own blog and I wanted to
[1945]
spell check and grammar check then it
[1948]
can work as a bit of a co-writer that
[1949]
there for you essentially making sure
[1952]
that your sentence are good graphics
[1953]
design impact good website comes out
[1956]
essentially saying a good graphics
[1958]
design can impact a good website much
[1961]
better written if you're writing for
[1963]
blogs in different languages Chad GPT
[1965]
can even translate that to pretty much
[1968]
any language I'm just gonna take a guess
[1970]
here and see if it can translate the
[1972]
above statement or sentence into say
[1975]
German in this case I could probably
[1977]
plug this into Google to make sure it's
[1980]
correct but honestly I think I almost
[1982]
trust chatgpt here more within
[1984]
translation than even Google the reverse
[1986]
is also true if you've got something in
[1989]
a different language such as German and
[1991]
you want to translate it back to English
[1992]
you can do that you can even add in
[1994]
things like adding it in a friendly
[1995]
joking way and we get back this example
[1998]
here a good graphics design can totally
[2000]
Rock a website socks off which is pretty
[2002]
funny now that we have a outline for our
[2005]
blog an introduction and say we've
[2007]
written it as well we can even prepare a
[2009]
new newsletter for it to be sent out by
[2011]
email and even have Chad gbt create the
[2014]
subject line for that email here I'm
[2016]
just going to ask it to create the
[2017]
subject line for an email newsletter
[2019]
based on the above and it's going to
[2021]
give me an output here of maximize your
[2023]
website's potential with these Graphics
[2025]
design tips perfect I hope you guys
[2027]
enjoyed this chat GPT tutorial this
[2029]
crash course is just an introduction so
[2031]
that you can understand a better way of
[2033]
how to use it to your best advantage of
[2035]
course if you have questions or if you
[2037]
have suggestions let me know in the
[2039]
comments below don't forget to like the
[2041]
video and subscribe if you want to see
[2042]
content like this thank you
Most Recent Videos:
You can go back to the homepage right here: Homepage