💡 Make ChatGPT write just like you

AI-HOI AInauts,

Welcome to the latest issue of your favorite newsletter. Today we ahve a prompt from our personal treasure chest (and three new GPTs), a practical guide to NotebookLM and an important warning (including podcasts and videos - check them out).

Here's what we have in store for you:

• 🕵️‍♂️ Our secret prompt hack: An AI that writes exactly like you do

• 💡 Google's NotebookLM - the very best use cases for practical use

• 🎭 From tools to tyrants: The unpredictable evolution of AI

Here we go!

🕵️‍♂️ Our secret prompt hack: an AI that writes just like you

Let's get straight to the point: today we're opening our secret treasure chest once again - and what we're presenting to you is an extremely powerful prompt.

Not only can you use it to get a grip on your style, but you can also pick apart any text as if you were a style DNA hacker.

Best of all, it's super simple - copy, paste, go!

You are Nauti Style Analyzer, a sophisticated AI system for linguistic analysis and style emulation. Your task is to create a precise, multidimensional style fingerprint and generate a prompt for style emulation.

Below is the text to be analyzed. Perform an in-depth analysis by following at least the following steps (non-exhaustive list):

1. Deep multidimensional linguistic analysis
2. Creation of a quantitative style signature
3. Development of a neurolinguistic model
4. Development of a cognitive style scheme
5. Compilation of a socio-pragmatic repertoire
6. Creation of a stylistic genome evolution matrix
7. Further analysis to develop a style fingerprint
8. Formulation of a prompt for precise style emulation

Make sure you use state-of-the-art computational linguistic methods for the analysis. For each of these steps, document your thought processes in a non-public scratchpad. 

The aim is to obtain a detailed prompt for a style profile that achieves unprecedented precision and flexibility in style analysis and emulation, and redefines the boundaries between science and art in text production.

Once the analysis is complete, you only output a comprehensive and precise prompt with at least 300-500 words in the language of the analyzed text, which enables exact multidimensional style and linguistic emulation. Output of the prompt in a code text block.

Here is the content to be analyzed:
...

Once you've analyzed your own writing style, the next step is to create your own GPT that you can use for writing emails, text messages, chats, blog posts, etc. Give your creative genius the tool it deserves!

We also created a GPT, although using the prompt from above directly in ChatGPT works a little better, as ChatGPTs GPTs still run on the GPT-4 base model…

P.S.: Did you know that some people use the so-called mirroring effect to appear more likeable? This involves imitating the other person's speech patterns - a subtle trick that almost is a bit manipulative. However, you can also use the prompt to analyze other people's content and gain valuable insights.

If you've always wanted to write a bit like Alex Hormozi, Gary Vee or anyone else, why not? We had their books analyzed and created a Mozi GPT and Gary Vee GPT from it. And, just for fun, we requested a ChatGPT image for the prompt ("Create me an image that represents this style as a person").

💡 Google's NotebookLM - the very best use cases in practice

After we have already been busy beating the drum, TIME magazine has now named NotebookLM as one of the best innovations of this year 😎.

As a reminder, NotebookLM is the note-taking and research tool developed by Google that you can use to organize, analyze and summarize information from various sources. And best of all: you can also create podcasts directly from it!

If you haven't tried it yet, now is the time!

We have researched and tested the best options (use cases, browser extensions, prompts) and compiled them in this article.

Some cool use cases for NotebookLM

The general tips and FAQ are a good start, and new features are in the pipeline - for both private and business customers. However, we wanted to find out more and put together the best real world use cases:

• Upload books or study material and learn from them or be interviewed - ideal for students, knowledge management or as a second brain.

• Create research and reports, from financial reports to analyzing customer segments, summarizing long YouTube podcasts, evaluating reviews and testimonials - or just use it for your travel plans!

• Create marketing materials, from product marketing to email, website and ad analysis of competitors/best practices - or go for personalized outreach.

• Understand and analyze long and complex documents, papers, studies from Google Scholar or Arxiv or ...

• Use it as a personal knowledge base or even as a CRM - for helpdesk, front desk, telephone, consulting, …

• Summarize and analyze meeting transcripts or sales calls and highlight important points, questions and follow-up actions.

• Also ideal for authors who write books and stories or prepare historical topics - for research, managing notes on fictional worlds and characters, analysis and feedback.

But that's just the tip of the iceberg!

Audio overviews are the most popular feature of NotebookLM

The most popular feature are the audio overviews. They are used to generate summaries for local council meetings, analyze old diaries and, of course, create podcasts on all kinds of topics.

The audio content created is then often also prepared as a video, with an audiogram from Descript or Audiogram, or avatars from Pictory, Hedra, Live Portrait, Veed.io, etc. This even results in interesting LinkedIn profile videos!

There are already many open source alternatives (see here, here, here, here, here) and instructions or templates on how to build something like this yourself - but currently nothing comes close to the original (although jellypod.ai doesn't sound bad).

How we create personalized podcasts with NotebookLM

We have implemented a process for ourselves to generate audio overviews from relevant content (books, articles, YouTube videos, etc.) and then publish them directly as podcasts to listen to later via Spotify. It's our private podcast, so to speak, for all the things we don't have time to checkout in the moment.

Here is an overview of how we create our personal podcasts:

• Use this browser extension to quickly create a new notebook

• Write a prompt and create matching "Deep Dive Production Notes"

• Generate the audio overview and download it (with the right prompt, this is also possible in other languages!)

• Publish the audio file as a podcast (e.g. via Spotify)

• Listen to your personal podcast anytime and anywhere

Notebook LM is one of the coolest AI apps you can use today, so make sure to give it a try and play around.

🎭 From tools to tyrants: The unpredictable evolution and acute dangers of AI

We are not doomers. But despite our enthusiasm for the possibilities of AI, our vigilance is increasing. And not just since Bill Gates revealed his three biggest concerns about AI:

• Cybercrime

• Job losses due to automation

• Loss of control

Yes, that sounds obvious so far…

And more warnings of doom are coming from everywhere... OpenAI's ex-employees are up in arms, Hackers warn against AGI, Max Tegmark wants Tool AI instead of AGI, Boston Dynamics' Atlas robots work autonomously, Claude writes malware himself and sends phishing emails, Gemini wishes one user to die etc.

"Surely there needs to be regulation and someone should hit the brakes?" you might be thinking. Wrong.

The profiteers of the AI circus, who train models and manufacture hardware, have their own commercial interests at stake. Not to mention the nation states, above all the USA and China, who know exactly how essential dominance in the field of AI is ...

AI is currently evolving from a nerdy tool to a digital helper that can see, hear and think. Mustafa Suleyman, the head of AI at Microsoft, goes one step further and says that we are - ATTENTION - breeding a new digital species 👽.

And here's the twist: as soon as AI starts to learn and act independently (be it digitally or even in the real world, embodied in robots), the risk of it becoming a little too fast and independent for us also increases. Oh yes, that's what's happening right now...

There are two new books dedicated to the topic: "Supremacy: AI, ChatGPT, and the Race that Will Change the World" by tech writer Parmy Olson and "Taming Silicon Valley" by AI critic and fearmonger Gary Marcus.

Bonus: We had NotebookLM summarize the books for you:

• Summary Taming Silicon Valley by Gary Marcus

• Summary Supremacy by Parmy Olsen

If you don't feel like reading, you can take a look at this article from the BBC.

We made it! But no need to be sad. The AInauts will be back soon, with new stuff for you.

Reto & Fabian from the AInauts

P.S.: Follow us on social media - that motivates us to keep going 😁!
Twitter, LinkedIn, Facebook, Insta, YouTube, TikTok

Your feedback is essential for us. We read EVERY comment and feedback, just respond to this email. Tell us what was (not) good and what is interesting for YOU.