👨‍🚀 The crazy story about the dirty AI work

Hello AInauts,

Welcome to the latest issue of your favorite newsletter! Three related topics today.

We start with inspiration on how much money people are currently earning with AI. Then we'll learn about the "dirty work" for AI, which can also be lucrative. And we conclude with a concrete benefit for you!

This is what we have in store for you today:

• 🤑 AI researchers: Are now more expensive than soccer pros

• 🔥 Data labeling: The crazy story about the dirty work for AI

• 🤔 Is this the next top job in AI for all non-techies?

Here we go!

🤑 AI researchers are more expensive than soccer pros

Let's start with a little gossip.

It's been making the rounds for several days that Meta (the company behind Facebook, Instagram, WhatsApp) wants to put together a new AI team and is aggressively poaching talent from competitors (see also our report on the "takeover" of Scale AI here).

Specifically, the team is called "Meta Superintelligence Labs" and has no lesser goal than to take the lead in the AI sector.

While it initially looked as if Meta was not particularly successful in recruiting, it has now been shown that money can solve a lot of problems after all …

Meta is said to have offered individual AI researchers up to 100 million dollars as a signing bonus... As it turns out: with success!

Mark Zuckerberg has now introduced the team in an internal memo. It includes leading names who have made important breakthroughs in OpenAI, DeepMind and the like.

Why is this relevant?

Meta's massive investment shows that their strategy has changed. It is no longer just about developing "good" open source language models, but rather about being a leader on the path to AGI.

The big advantage that Meta already has is the billions of users of its platforms, the existing data center infrastructure and, of course, the vast amounts of data about all of us.

There is also a very important edge in terms of user experience: the Meta Smart Glasses

If you've been reading us for a while, you'll know that we're big fans of Meta's smart glasses.

Maybe OpenAI and Jony Ive are building a completely new AI device that we can't yet imagine.

But for us, smart glasses are the next logical step in how people use AI in our everyday lives.

• AI that sees what you see,

• that you can talk to comfortably and discreetly and

• get answers through the glasses instead of headphones

And if display projection in the lenses works, it will be the next big thing for us in the field of AI devices.

With the new team and the existing resources, products and reach, we personally wouldn't go short on Meta for the next few years*.

We are very curious to see how the competition reacts and how the whole thing develops.

(* This is definitely not an investment recommendation. We have no clue. 😉)

🔥 Data labeling: The crazy story about the dirty AI work

Because it fits the topic so well, we would like to take a look at the hidden champions of AI. These are companies that hardly anyone knows - but which are absolutely essential for AI! And there are also two crazy success stories that we find very exciting.

Meta buys 49% of Scale AI for over $14 billion

We have already written about how much Meta is currently spending on AI talent.

A few weeks ago, it was also announced that Meta had secured around 49% of Scale AI for over 14 billion US dollars.

The 28-year-old 🤯 CEO of Scale AI is now in charge of the above-mentioned Meta Superintelligence Lab.

But what does Scale AI actually do, and why is it worth so much money?

Why Scale AI & Co. are so important

Companies like Scale AI (or even competitors like Surge AI, more on that below) are the invisible backbone of the whole AI revolution!

While everyone is talking about ChatGPT, these companies are doing the real “dirty” work. They label and curate the data that is used to train all the big AI models.

What they do:

• Data labeling: Humans mark images, rate texts, correct AI outputs

• RLHF (reinforcement learning from human feedback): They teach AI models what is "good" and what is "bad"

• Quality control: They ensure that AI training data is high quality and not biased

Why this is mega important: A relatively simple task that only humans can do is very important. Without these companies, there would be no ChatGPT, no Claude, no autonomous cars - nothing! 🤯

These companies work with thousands of contractors, often freelancers, who categorize the data on their behalf.

One of Scale AI's competitors, for example, is Surge AI. The story behind it is also pretty crazy…

How can you tell that companies are printing money? Exactly, when they have websites like this ↑. Without call-to-action buttons. Without pop-ups. Without lead forms.

The founder started the company completely without external funding. And it is profitable since day 1. Over 1 billion in sales in 2024!

Let's just assume that approximately 30% will be paid out to the data labler. Then some overhead. The bottom line should be a few hundred million in profit.

And he did this simply by approaching young college people and showing them how to categorize data - and then selling it on to OpenAI & Co. for twice as much.

A relatively simple job that (still) requires people is the backbone of our AI!

P.S. For those interested in the topic, up to six-figure salaries are possible through data labeling. Just take a look at the job pages of the two players and search for Data Annotator or Generative AI Generalist.

🤔 Is this the next top job in AI for all non-techies?

So, what do we do with this information? Or in other words: how can we and you benefit or take advantage of the incredible opportunities?

Let's end the newsletter with a topic that can be useful for you.

Because, if we're honest, we also know that very few of us have the skills to become an AI researcher that Meta then hires for expensive money. But data categorization is also rather boring and unspectacular.

However, a skill, or rather a new job description, is currently sweeping the world that we can all learn. Even as non-techies.

May we introduce: The AI Automation Engineer

Here, for example, the post by Wade Foster, founder and CEO of the popular no-code platform Zapier:

The statement is strong: "If you're an AI Automation Engineer, we'll hire you. In every position we have open."

What is an AI automation engineer?

Before we get to how you can become an AI automation engineer with a manageable amount of effort, here is a brief explanation of what it is:

An AI Automation Engineer is someone who uses AI concretely and directly to make everyday problems and workflows easier and more productive. Not a theoretical researcher, but someone in the thick of the action.

Imagine you see any little annoying task in sales, customer support or HR. Instead of solving them manually over and over again, you build fast, practical automations.

You turn simple ideas directly into executable solutions, using tools such as Zapier, n8n, Make, Airtable, etc.

Your job is to identify the crucial points where automation really makes sense.

You build prototypes within a few days, accompany teams on site, understand their workflows and optimize them with your AI automations.

Why is this role so important?

Because many teams know that AI could make a difference, but don't know exactly how to get started.

The AI Automation Engineer closes precisely this gap between theory and practical implementation. He brings directly visible results, saves working hours and improves the quality of processes in the company.

And who has the best prerequisites for this?

• No-code enthusiasts who work intensively with Zapier or Airtable, for example.

• Generalists from start-ups who have a flair for automation.

• Prompt and context engineers or LLM product hackers who like to experiment and implement.

• People interested in AI/automation who are willing to learn and have stamina 😄

The great thing about it is that you don't need years of technical training. Above all, you need curiosity, practical thinking and the ability to cleverly combine tools.

In short: you become the interface between humans and AI.

How can you learn it?

If you've been reading us for a while, you'll know that we wrote about this very job description over a year ago.

We are delighted that large companies such as Zapier are now specifically looking for it, and this also confirms that we were right to focus on this important topic.

For you, this means that with a little training, you can follow this path and develop into an AI automation engineer.

It is a skill that is already in demand today and whose importance will increase massively in the future.

The world needs exactly this practical combination of AI and automation, and YOU can do this job. YouTube has everything you need, and if you join a community of likeminded folks, you are on the right track.

If you are passionate about this topic and would like to learn more about it, please let us know!

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 - it motivates us to keep going 😁!
X, 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.