Introduction

What we are about to do together is to put together a presentation from scratch. In 15, 20 minutes, we will have a presentation of five to 10 slides. Of course, we need a topic.

Registration.

QR code registration. Okay. Let's hold that.

I like it. No, no.

Maybe something a bit more, something that you might have to do at work. Any ideas?

Selecting a Topic for the Presentation

How to build a community? How to build a community. To build a community.

What kind of committee? How to build a community, OK.

Suggestions and Refining Ideas

Let's think about something a bit more researchy, where we also need to do a piece of research. Any ideas?

Some market research? Some? Market report? A market report?

OK, on what? What market? What sector do we want to cover?

Food industry.

I don't know. It's just an idea, like in the food industry.

Food industry.

Who is the audience? Investors. Investors, OK.

Food industry, food is really broad. Let's try to narrow it down a little bit.

Plant-based. Plant-based. Like where this is going, yes.

Segment.

Finalizing the Topic

OK. Plant-based.

What do we want to know about the plant-based segment? Size, growth.

Trends. Demand trends.

How about this? Then probably competitors or the supply landscape, right?

Anything else? Impact of AI chefs.

AI chefs. Wow, that's specific. Chad GPT is going to have a lot of fun with that.

Impact of AI chefs. Even if it's too hot.

OK.

Tools for Building the Presentation

Now let's see if this works. So I'm using Napkin AI. Many of you have heard of Napkin AI, yes?

Let's see.

I would like to put together a mind map of the plant-based market looking at these key elements.

Let's see if Napkin does the magic for us.

Using Napkin AI

So what I like about Napkin is I find it very useful at the brainstorming stage because it gives you a range of possible infographics that you can use to structure things. And it really picked up the plant-based theme.

See how it's giving us some little. And to me, this is more useful for brainstorming or to understand how you might put together a slide.

So we've got our list. Now, what do we do with this?

We go and charge GPT, and we create a project. Plant-based market research.

I've got my list. And I'm going to run a deep research to get a report on this.

So these models are getting really powerful. So a good deep research these days would probably take 15 minutes.

But we are going to do the short version of it. How many of you have heard about deep research?

So what deep research does Effectively, ChargeGPT would function as more of an agent looking autonomously at quite a bit of information, more information than it would otherwise do, and it would produce for you a report that could be anywhere in between five to 20, 25 pages.

And when you combine it with some of the more powerful models, how many of you have been playing with the different models that ChargeGPT has? But when you combine it with some of the more powerful models like O3, then it becomes independent and gives you quite a bit of information.

But we are going to stick to basics today to ensure that we can finish the research piece for the presentation. Imagine that you are an experienced market researchers.

Leveraging ChargeGPT for Research

I want a fast-tracked three-page overview of the plant-based market focusing on these key elements.

And you see that the typing or how things are arranged will not really matter. More questions.

We want a global overview or a region? Any takers? Europe.

What is the intended user audience of this report? We said investors.

Would you like recent data? Yes, I guess.

Impact of AI chefs. What do we mean? I don't know what we mean.

Who said AI chefs? Yeah, AI pizza. You know, like the pizza maker robotics.

AI chefs probably is not a good term. Maybe robotics. Impact of robotics.

So now that this is baking. we might start showing you how Gamma works a little bit.

Exploring Gamma for Slide Creation

How many of you have heard of Gamma before or have used Gamma? Gamma, like the Google ? No. Gamma is, you will see, a presentation tool.

1Actually, I have it open here. Let me reopen in one moment. This is gamma.

I think what you get with that, that, that, And so it's really an AI-first slide making tool, presentation tool, that starting from, and you have different options here, but you can effectively just put together some text. And then it would create a presentation for you, which is what we are going to do today.

Now, the advantage with respect to other tools is that now I will open a presentation that I delivered in Athens so that we can take a look. And it's really It comes with quite a few pre-built templates that you can just use.

I'll give you an example here. So here, it just gives you some pre-built layouts that you can use to change the slides. This is the first strength that I find in Gamma.

Then there is an AI generation feature, and you can get quite advanced both in the style that you decide for the images, the model selection, and the description of the images themselves. And then there is an actual AI interface. So for example, say, can you move the picture to the right?

It would just pick up your command and execute. So it really becomes this sort of AI first tool that was conceived to be used with generative AI that I feel tends to be faster if you put in a few hours at the beginning to learn how to use it.

So we've been through gamma. Let's see how this is going. This is still taking some time.

Image Style Options

So in the meantime, perhaps, things that we can start doing we can start thinking about how we want the images to look like. Because as you have seen in gamma, we can choose the style for the images.

So we can come up with a style that matches the report, maybe something like So we go back on our research. Imagine you are an experienced marketer that is putting together a presentation on the plant-based market.

Can you come up with a description for the style of images in a gamma presentation that is maximized for DALL-E.

So the Super Whisper app that I'm using clearly likes the painter a lot better than the AI model. What do we think about this?

Yes, you cannot sit. Bright modern editorial style illustrations with clean lines and soft natural lighting, featuring diverse people enjoying plant-based foods in everyday settings.

All right. Let's try something else. Can you ensure that it really appeals to an audience of investors?

All right. So it becomes more boardrooms. Fine. Let's see how that works.

Now, where are we with our deep research?

OK, this is still going. It will take a bit longer.

So what else can we do in the meantime?

Enhancing the Presentation Experience

Have you guys heard of Suno? What's used Suno before?

Yes.

How many? Can I guys show hands?

Incorporating Music

All right, so another thing we can do as we wait for the DRIP research to go is to put together a song that we can play to investors. Does anyone have very strong preferences on music and music genre?

Country. You want country. Country, plant-based.

All right. Country, sorry? I'm preparing a presentation for investors on the plant-based market.

Imagine that you are an experienced songwriter and you put together country song about the plant-based market that I can play before the start of the presentation. Of course, I paste it in the chat history.

Great. So the only reason I'm switching to Claude's. OK, I need to do this again.

Imagine that you are an experienced songwriter. I would like the lyrics for a country song on the plant-based market that I can play to investors at the beginning of the meeting.

The only reason I switched to Claude is that I want to leave GPD in peace while it's doing the deep research. Yes. What do you prefer, to use ChatGPT or DeepResearch?

DeepResearch, ChatGPT is still a bit better, I think. Gemini would be the best, probably for the connection with the Google ecosystem, but it also takes a long time.

Amrit, you are the judge. What do you think? You like it?

You cannot see it very well. I'm not going to sing it because I'm a terrible singer. So let's... Now, where are we?

Lyrics. What do we need to do? We need to do country, we said, right, Andre?

I think we should erase the first one. What did I do? Maybe you have to take off the first one, two points, because it suddenly just detects it.

Like the verse here? Yeah, maybe ask Claude to remove that. No, don't worry.

It's going to work. Yeah, it should. Maybe quicker in doing it that way. That's a good shout.

It's minimize opportunity for errors. Let's see how our deep research is going. Okay, still some time to go. I think, doesn't seem to be bad.

Alternative Ways to Engage the Audience

Anyway, this shows how you can use a tool, this sort of tool, I think, for some presentations where you're trying to show some marketing effort, I think it could be interesting.

beside for anything. You might want to have your own song happening.

So I think the internet is being very slow here. So, I'm wondering what we should do.

Perhaps just for the sake of showing you how Gamma works. And we go back to this later.

I think I run a parallel, the same search on clouds list we got going. So they should be quicker. It's not something I recommend, but it gets us through the demo.

Fortunately, the time that it takes ChartGPT to run these sort of searches is not entirely predictable, in spite of trying to control it for length. It didn't work out.

Okay, let's see how Claude does this. Okay, can you guys see? Now you can.

It's already pretty good. It would be interesting to compare to chat GPT later. I'm actually very impressed.

Maybe I should have used Claude. Who said I should have used Claude in the first place? Sorry? It depends for what.

We can chat over pizza and drinks. It's really good. Yeah, but for the presentation, it might be enough. Yeah, I mean, for the presentation, it's definitely enough.

I just wanted to have the deep research. All right, Claude. Now, imagine you are an experienced presentation maker. Can you give me an outline for

10 slides that summarizes this report and that is maximized for Gamma. How is it starting? Okay, title and hook. Snapshot.

Okay. Can you just leave the content and eliminate the gamma features? Okay.

Yep, that works. Okay, let's see if our, is it still going? Now, we are getting there.

What is the plugin you use for voice control? For the voice control? Yeah.

It's called Super Whisper. It is, so you see I'm not a native speaker. My accent is far from impeccable, but it mostly picks what I'm trying to say.

Okay, so we're almost done here. Now that's where the magic happens. So we go in gamma.

We paste in the text. It's a presentation generated from notes or an outline. Go on with slides.

Now, you remember our description of the slides or of the themes that we wanted for the slides. You see, writing for investors. Tony's analytical, informative, and optimistic. This is what we want.

for the style. Now, something green probably. You can also import your own custom themes. And we go and generate.

And so it's basically putting together a presentation that is not perfect, but really gives us a head start in terms of what could be the final version of the block. What happened here? I'm really impressed by the images.

I think they're quite consistent with the theme that we picked. And whether we like the theme or not is another question. So I think in the space of...

Conclusion and Reflections

10, 15 minutes, and in spite of not really having access to the deep research by CharGPT, we were able to put together from scratch a presentation and also a song that we can present to the board meeting. Now, whether the song is gonna have a good impact, well, that's for the investors to say. Now, are we done here?

Comparing Results and Effectiveness

I at least want to see, okay. Let's see how it compares to Claude quickly, if it was worth it. So this, I think you now see the difference with respect to what Claude has produced.

To me, this is a lot more detailed and better resourced. So you've got all the, you can double check every single point. So it would have been worth the wait, totally worth the wait for something important.

And I'm pretty sure if we now put together a presentation on the, an outline out of this research, the difference would be dramatic, but let's do something. Just so that. Imagine you are an experienced marketer researcher again.

I want you to evaluate this alternative research that has been put together on the basis of the report that you have just produced. And now I've got, let's do it with a presentation. Then we call it a day.

This is probably too long. Let's get into the first five. Let's try again.

It didn't finish. Ah, it didn't finish the previous one. No, I think it did.

Ah, OK. It was actually, I had not turned off deep research.

Duck. So I really like this flow when you're using one LLM to check the other LLM. I think it can be very powerful.

Challenges and Live Demo Insights

But you see that sort of highlights what I was saying about the deep research being worth it, because there are a lot of comments about it's vague, there needs to be more detail, so on and so forth. Sorry if it took longer than I expected.

It didn't go as expected, but that's the beauty of a live demo. We did figure out an alternative together.

Now, I don't want to take too much time from Amrit and Diana. So maybe if you have any questions for me, we can cover them during the pizzas.

But thank you so much.