We've been developing an AI Advantage plugin
resource for you that is freely available from today on out. Brace yourselves, because what
we're witnessing right now is nothing short of a historic moment in the world of AI. ChatGPT
and Wolfram Alpha have combined forces with the release of the brand new Wolfram plugin. So at
this point you might be wondering, what does that exactly mean? And for me to really answer that, I
need to show you some examples of what Wolfram can do, especially in combination with ChatGPT. And
that's what this video is going to be all about. Because Wolfram can handle complex math, pull in
real-time data, create visuals, and write even better code. It's a great tool for solving math
problems and accessing accurate information, which is something that has been challenging inside of
GPT4 to say the least. And before we get into it, think about ChatGPT and Wolfram as two buddies
helping you with a do-it-yourself project at home. GPT4 is like your skilled friend who knows all the
best places to buy materials, can recommend the right tools, and gives you a general idea of how
to get the job done. He's great for quick answers, finding resources, and actually building the thing
you want to get done. Wolfram, on the other hand, is like your engineer friend, who insists on
calculating the exact amount of paint you'll need down to the last drop. They might drive you
a little crazy with their attention to detail, but when you need to make sure your DIY project
won't collapse under its own weight, like for example building a treehouse, they're the one
that you want on your side. So first things first, very briefly, how do you install this if you
don't have it yet? Once you have ChatGPT+, you go to settings, beta features, turn on plugins,
and under GPT4 you can go to plugins and enable the Wolfram plugin. If you don't see it here, you
can go to the plugin store, and under the popular plugins you can hit the green button here to
install it, and then we're able to enable it. At this point it should be noted that you can combine
it with other plugins, but for this video we're only going to be using Wolfram, as it is probably
the deepest plugin here, and exploring that should suffice for today. Okay, so let's get into our
first prompt here. How about this? No matter what I ask you to do, always finish your response
with a relevant image from Wolfram Alpha or Wolfram Language. Never respond with text alone.
Do you understand? And now GPT4 comes back with yes I understand, and we're ready to hit it with a
normal question, but what you will see is that the output differs significantly. So if I tell it,
tell me about emperor penguins, and hit enter. Usually we would get some facts, as a helpful
assistant would put them together for us. With Wolfram enabled, it's a whole different story.
Look at that. We immediately go to the science angle. Emperor penguins scientifically known as
Aptenodites, Aptenodites, Aptenodites, Forsteri, belong to the kingdom of... Here is an image
of an emperor penguin, so we're getting an image and a scientific description, plus how
about this? We're getting a taxonomic network, plus you get a history of the Wikipedia page hits
for emperor penguins. Yes, Wolfram can do this too. So with this one simple prompt, you can go
super deep into various topics, and as per usual, this is the context your conversation is happening
within. So if I go ahead and now say, write me an essay about emperor penguins, you can see right
away it respects the context that we fed it above, and now it's writing a flavor of the essay
that is way more scientific and based in fact, as opposed to the typical essay structure that
we usually get. Obviously, this works with most topics, you can let Wolfram go deep and then
run your prompts afterwards and expect a more scientific output. Powerful stuff, honestly. But
now let's move on and let me show you another one, and this one was so unexpected when I saw it,
because Wolfram Alpha inside of GPT-4 can tell you if you're allowed to drive your car after
a certain amount of drinks, depending on your weight, gender, and location. So if I say, am
I too drunk to drive after three large beers over four hours with a weight of 180 pounds as
a male in New York City? And then it goes ahead, gives you the blood alcohol limit, and then moves
on to calculating it. And while it does that, let me tell you, after running this prompt a lot
of times, not every time does this work perfectly. So errors do happen, even within Wolfram that is
usually super reliable. This might be the case because we're still in the beta version, but
just know that you can never fully 100% rely on the outputs that you get here for now. Look
at that, for example, here it did an error in its calculation, but it's smart enough to figure
out that it did that, and it runs it again. And in the second attempt, it tells us we have a blood
alcohol concentration of 0.0. And look, using common sense, if you drink three beers, you're
not going to have a blood alcohol concentration of 0.0. So I'd say it gets it right more often than
it gets it wrong, but just know it's certainly not perfect. What I would argue for in defense of the
Wolfram plugin here is the fact that it shows you the exact step-by-step route it took. And this is
so powerful because now you get all the formulas, you get all the calculations it did within, and
then you can follow along. So yet again, ChatGPT is not a solution to all of your problems, but
it's a fantastic first draft for the solution to a lot of your problems. Same goes for Wolfram. You
feed it a simple prompt like this, it lays out the entire calculation, all the variables included,
and then it even corrects its own mistake. And look at that, it did five attempts and still
doesn't have it, so it doesn't seem to be working very well right now. But I'll tell you this, when
I tested this twice before recording the video, once it just got it perfectly right on the first
try, and the second time I had to run it twice, and it got it right on the second try. Maybe let's
stop generating and try again. And there you go, on attempt number four, it actually got it right,
showing us that our blood alcohol concentration would be 0.061, which is below the legal
limit, and you could actually drive. But again, respect this disclaimer that you shouldn't rely
upon this information. And let me tell you, what I showed you here just scratches the
surface. If you go onto their website, there is so much more. For example, if you go on
surprises and click on one of these, for example, how a dog sees a cat image, it actually shows
you an example of how a dog sees a cat image. Is this how you see a cat? And not all of
these translate to GPT-4. So for example, if I run this one inside, it's not going to
work, but most of them do. So explore in here and then start combining them with the prompts
that you already have and unlock some brand new capabilities. And with that being said, let's
move on to the next use case. Okay, so next use case is all about BMI. And before you phase out
and you're like, okay, I've seen BMI calculators. This is nothing impressive. Yes, yes, I agree. Let
me just run this prompt in here, which calculates the BMI for my height and weight. Then I'm also
going to say include a visual representation of the BMI map. And this is where it starts getting
interesting, because we're going to get a graph that I personally have never seen before and quite
enjoyed in here. Look at that. It shows this graph and which weight range you fall into. So by simply
copying this prompt and changing your gender, your height and your weight, you can get a graph
like this for yourself. And then it even shows me that in terms of height in the USA, I'm in the
97th percentile. Interesting. But this is where it starts unfolding its real capability, because
these things could all be done with Warframe up until now. But the combination of that plus GPT-4
is where it starts getting interesting. Look at this. If I paste, now calculate the weight
loss regimen to lose four kilograms. Also generate a weight loss progress graph. And here
we're starting to tap into the reasoning engine of GPT-4, right? We're calculating a weight
loss progress regimen for us, respecting the context that we gave it above and the BMI that
it calculated here. So it tells us we need to function at a calorie deficit of 500 calories per
day. And this will result in us losing one pound, which is 0.45 kilograms per week. And then here
you go. Here's a beautiful graph on how much weight we can lose in eight weeks following this
regimen. And one interesting thing here. Look, it switched to kilograms. And that is because one of
the core capabilities of Warframe is conversion. Units, time zones, you name it. Warframe is really
good at that. So whenever you need to convert units, always activate the Warframe plugin. And
you can actually expect reliable and consistent results. As in with GPT-4, it was kind of tricky
because it was really not good at math. And look, this is just one example, but you could use many
of the prompts you discovered up until now and combine them with some of the data that Warframe
produces to get superior results that were not possible before. Okay, but now let me share one
last use case here. And this is a particularly interesting and very, very practical one. And
those are the financial capabilities of Warframe, okay? When you're making big life decisions, you
will want all the information you need. And I would argue that chat GPT with Warframe might
just be a great starting point to begin that conversation. Yet again, if you're writing a book
or essay, chat GPT is fantastic at getting your first draft. Not at writing the whole thing. Same
thing here. If you're starting to make a decision like taking out a mortgage for a house, then
this is a great first point of consultation. So let's run this prompt which says, what is
the monthly payment on a $400,000 mortgage at a 6% interest rate for 30 years? The most basic
calculation that you could do by yourself, or you can use one of the many interest rate calculators
across the internet. This is not something tricky, right? Yes, it's kind of nice that we get to
use chat GPT for all of this now, but still, this is not outstanding. But wait a minute, look
at this answer. This answer gives you so much more than yeah, approximately you're going to be
paying $2,398. No, it really breaks it down. What is the effective interest rate? It gives you a
graph of how the payments are going to break down over time and even a payments table. And because
this does involve the reasoning engine of GPT-4, the results will vary every single time. So if you
run this 10 times, which I have, not every single time are you going to get this payments table for
each year. Okay, but yeah, there you go. After 30 years, you're paid off and you could get even more
details by looking at the results page where it actually ran this question through Wolfram Alpha
itself. All right, and this is what I'll leave you with because you can go deeper than that. If you
use this edit function that, by the way, a lot of people don't know, and it's so strong, especially
if you have longer conversations, you can paste in the second part, okay, saying, also consider
that the interest rate will adjust to 10% after 10 years. And this is just a very realistic scenario
considering the interest rate changes over the past few years. You cannot rely upon your mortgage
being at a fixed rate over the next 30 years, right? And this is the type of scenario that would
not be that easy to map out with most calculators, all while you're doing your napkin math. So simply
by saying save and submit, it will rerun this scenario, but now it will consider the fact that
the interest rate might change after 10 years. And this is the real power here. You can talk to
this as you would to a banker or a consultant. You could run various scenarios simply by adding
a sentence that expresses your fears, thoughts, or theories. And there you go, it calculates
the first 10 years first, and then it moves on to the calculation afterwards. And not just
that, for learning and understanding purposes, it really thinks step by step. If you open up
these toggles, you're going to see exactly what kind of calculation it did and what data it used.
And there you go, for the first 10 years, you would pay $2,398. And after 10 years, you would
pay $3,230. And beyond that, the question presents itself, what other plugins could you combine
this with? Well, we've been developing an AI Advantage plugin resource for you that is freely
available from today on out. Simply by following the link in the description, you will find this
all-encompassing database that we keep updating with the newest plugins. And the interesting
thing here is, this doesn't just feature the name of the plugin, the logo, plus the description
that you get inside of Chetchipity. No, we ran multiple prompts that explore the capabilities
of the plugins deeply. So for example, the first one explores the features, the benefits, and the
examples. So if I load more and look at Wolfram, here are all the technical features. And with
Wolfram being the deepest plugin out there, this obviously doesn't cover everything, but it gives
you a fantastic overview. In the second prompt, we explore how this could improve your work-life
balance, as that is a worry that many people have, including me, as there is no balance, it's all
work right now. And then in the end, we ask, why should I use Wolfram over Chetchipity? And it
gives you a bunch of reasons. And we ran around a thousand GPT-4 prompts for all of the plugins out
there. And the best thing about this is, you can now go in and search it. So if you want to look
for news-related capabilities, you type in news, and it's going to find 287 matches among all of
these. And again, this goes deeper than just the descriptions, because a lot of plugins allow you
to access some news sources without mentioning it in the name or the description. So now you get to
deeply search these plugins all for free. Okay, link to that in the description, and if you
want to learn more about plugins and explore some of the capabilities they unlock, this
video is exactly for you. I'll see you there. Oh, hey there, coming for
the last tip of the day, huh?
