• AInauten.net
  • Posts
  • ๐Ÿ‘จโ€๐Ÿš€ 5 new AI features that you absolutely have to try out

๐Ÿ‘จโ€๐Ÿš€ 5 new AI features that you absolutely have to try out

PLUS: The guy who told Zuck to keep his $1.5 billion

 

Hello AInauts,

Welcome to the latest issue of your favorite newsletter! Is the Matrix being reprogrammed? Sometimes it almost seems that way...

Today, we're talking about the AI researcher who simply ignores the 1.5 billion meta offer, the new Google tool that turns simple prompts into complete apps, honest interviews with Sam and Dario - plus 5 powerful AI features you can try out right away.

This is what we have in store for you:

  • ๐Ÿ’ธ Not for sale - not even for $1.5 billion from Meta

  • ๐Ÿ‘€ 5 new AI features you need to try out

  • ๐ŸŽ™๏ธ AI updates: Doomers vs. accelerators - Dario & Sam in conversation

  • ๐Ÿ˜ AI-Fun: Robots today and in 20 years' time

Let's go!

๐Ÿ’ธ Not for sale - not even for $1.5 billion from Meta

Zuck offers an AI researcher 1.5 billion dollars over six years. The answer? "Nope."

Sounds like a plot from "Silicon Valley", but it really happened. Meta wanted to poach Andrew Tulloch, co-founder of Thinking Machines Lab, from the team of former OpenAI CTO Mira Murati. He had previously worked at Meta for 11.5 years and probably had a good idea of what to expect...

Meta is still on a shopping spree and wanted to poach half a dozen more top people from the small but brilliant research lab. But the entire team remained steadfast. No departures, no changes, no buyouts. Just a collective shrug of shoulders - despite Metaโ€™s billions.

Welcome to the new reality of the talent war, where fewer than 1,000 people worldwide have the skills to build the future - and even fuck-you money is no longer incenctive enough.

But Thinking Machines is by no means the only company that has been targeted by Meta! Zuck has the 'list' and is going on a shopping spree with all the leading AI providers.

At OpenAI alone, 100 employees were pinged and 10 took the bait. That's a conversion rate that would make any performance marketer weep.

Money no longer plays the main role in this talent war.

Does money rule the AI world, or is it the mission that matters? What is certain is that the classic rules no longer apply here:

  • Google: paid $2.7 billion for ONE human (Noam Shazeer, co-founder of Character.AI)

  • OpenAI: median salary $860k, but 8x higher churn to Anthropic

  • Anthropic: "Only" $690k base salary, but 80% retention rate

  • Meta: offers $200-300M packages - and still often encounters rejection

This is a complete redefinition of what human capital is worth. The AI talent shuffle is in full swing! Interesting fact: Anthropic has the best retention rate.

Our take: Mission beats money!

The question is no longer just "What do you pay?" but "What do you stand for?" The new currency is trust, autonomy and purpose.

The brightest minds want to change the world. Make history. And not because of the money, but because of the mission.

The irony is that billions are being burned invested in talent and at the same time AI systems are being built to automate these jobs. Human intelligence is being bought to make it superfluousโ€ฆ

๐Ÿ‘€ 5 new AI features you have to try out

So, after the bird's eye view, it's time to roll up your sleeves again and not just scroll through the newsletter, but try out a few of the following features. Ready? Let's go!

1) Genspark AI Slides 2.0 - research and slides in one go

One of our favorite tools, Genspark, has upgraded the existing Slides functionality once again - and it's impressive! For example, we used it to research and create a comprehensive AI chatbot comparison table.

Example of generated slides in Genspark

There is now also a simple editor where you can make adjustments directly without having to go through the chat interface. Not bad, but still a far cry from Google Slides. The sticking point is that you can export the data to PowerPoint, but the format is destroyed in the process. PDF, on the other hand, works perfectly.

Pro tip: You can even create multiple presentations in parallel thanks to the new multi-orchestration feature!

2) ChatGPT Connectors - Connect ChatGPT with Gmail and Co.

Before the big GPT5 launch, OpenAI is already providing us with some cool new features ... Firstly, there's the ChatGPT Study feature (see our article here), which supports you in your studies. And of course the ChatGPT Agent, which you should also try out (Plus account required).

But did you know that you can use ChatGPT Connectors? This allows you to connect your Gmail, Calendar, Drive, Canva, ... accounts and use them in conjunction with Agent Mode or Deep Research. Very practical!

Note: The connectors are only available in certain models/modes.

For example, you can have your mailbox searched and analyzed:

3) Opal by Google - create AI mini-app workflows without coding

Google Labs is launching Opal, a new Labs experiment that builds complete AI mini-apps from simple prompts. Without any code, but with a visual workflow editor!

You can easily get started in Opal : Enter your own prompt or remix one of the existing templates. Done? Then you can share your app with the "Share" button.

The advantage compared to other vibe coding apps is the precise editing options for prompts and tools, which we find practical. Opal offers a simple, intuitive user interface that allows you to quickly build a prototype.

The visual representation of the workflow and integration into Google's strong AI ecosystem allows granular customization and access to many features and AI models.

Example: Workflow of the Oval Office Handshake Generator, tool selection dropdown

But there are also technical limitations: no real (customizable) code is generated, nothing can be exported and apps quickly reach their usage limits (quota).

Nevertheless, here is our example: an Oval Office handshake generator with President Bart Simpson. You can upload your own image to generate your personal handshake (although the available image model still leaves something to be desired ...).

You can find more apps here.

Beispiel: Oval Office Handshake Generator

4) Google Veo 3 - Images with annotations generate cool videos

Let's stay with Google: The great Google video model Veo 3 apparently also works well with instructions that are placed on the start image. This makes it very easy to illustrate the dynamics in a scene.

The following two sample images illustrate this well - and are ideal for testing. Simply upload them to Google Gemini, Google AI Studio or Replicate with the following prompt.

Animate this image. Immediately delete the instructions in white on the first frame and execute them in order.

Example 1: TV and vase

Example 2: Panther and cat

5) Manus Wide Research - Your unlimited research agents

The new Manus feature "Wide Research" can send over 100+ agents into the race at the same time! For example, you can have scientific papers analyzed and prepared (prompt here) - or if you're a bit simpler, like us, you can create a comparison list of the most important AI chatbots ๐Ÿ˜.

This is a brief roundup of some of the features that seem important to us and show where the journey is heading. Give it a try!

๐ŸŽ™๏ธ AI updates: Doomers vs. accelerators - Dario & Sam in conversation

Is AI a ticking time bomb or humanity's greatest blessing? We have found two recent interviews that we can highly recommend to you!

Dario Amodei, the CEO of Anthropic - the "anti-OpenAI", as some call it - shakes off the doomer label and still warns of exponential growth. But are we all too naive? In this interview, he spills the beans - a wake-up call for anyone who thinks they've got it all figured out.

Sam Altman also lets us look into his cards. But in this interview, something rare happens: he drops the PR mask and sounds not just like a CEO, but like someone who has opened a Pandora's box ... and now isn't sure himself whether a genius or a demon has jumped out.

This is your opportunity to hear the raw, unfiltered thoughts of the man who is shaping our future more than any politician right now. A future where your job may no longer exist, but you could own shares in AI.

๐Ÿ˜ AI-Fun: Robots today and in 20 years' time

It's done! Robots can not only replace the surgeon, but also do the laundry.

Soon it should go one step further โ€ฆ

โ€ฆ and this is how it will probably end โ€ฆ ๐Ÿ˜

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 ๐Ÿ˜!
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.

๐ŸŒ  Please rate this issue:

Your feedback is our rocket fuel - to the moon and beyond!

Login or Subscribe to participate in polls.