This issue is brought to you by:
AI-HOI AInauts,
Welcome to the new issue of your favorite newsletter. The feedback about our issue on how to train your own image AI was clear: you want more of this!
That's why today we're taking a look at how you can create any type of app with the help of AI. Whether it's a web app, WordPress plugin, browser extension, website, ...
Implement your own ideas - or build a side business with them. Even if you've never written a line of code before.
That's what we've got for you today:
⭐ 8-year-old builds app with new tool - what can you do with it?
📰 AI news quickie: The HAIlights from the industry
Here we go!
⭐ 8-year-old builds app with new tool - what can you do with it?
Imagine you could develop complex apps without writing a single line of code. Sounds too good to be true? We assure you: it's real! And: you can do it too.
From idea to app in no time
You can now go from an initial idea to a finished app in record time. Whether it's a task manager, chatbot, website, browser extension or a complex web app - with Cursor, you can turn your vague idea into a concrete project in no time.
The tool has won the hearts of early adopters worldwide. Simply say what you want using natural language - and the AI creates the app for you. THIS is the future present of coding (sorry, devs). Solopreneurs love it!
Just watch this video. If an 8-year-old can create a Harry Potter app with a few tips from her dad, what can you build with it?
Let's get started - your first app in 15 minutes
You know, we're not coders either - but when we saw this video, our ambition got the better of us. And we picked out a simple use case that we can implement together, step by step.
This will turn you into a junior developer in fifteen minutes 😎 And don't worry, it's not complicated - we promise!
Let’s create a simple browser extension together.
Step-by-step instructions
Download and install Cursor.com
Create a new folder for the app on your computer and open/drag the folder into Cursor
Now use the 'secret' key combination: Cmd + I (or Ctrl + I on Windows). This opens the small but powerful "Composer" window - this is the wizard that does everything for you. (There is also the AI pane at the top right - but the Composer is the better choice and can do more).
There, you simply enter the prompt for the desired extension ("Create a browser extension that ...") and click Enter.
The app now creates the necessary files directly, and you can transfer the generated code with "Accept All".
Now save all files with Cmd + Option + S (or Menu File → Save All).
Switch to the Extensions in the browser and activate Developer mode. Then drag the extension in there and activate it.
Test the extension in the browser. The code might not work straight away. But no problem, just copy the error messages to Cursor and keep optimizing the app.
We love it! If you want to go deeper, YouTube and X have lots of tips.
Our take: Cursor democratizes app development
No question: with Cursor, anyone can develop an app, plugin, etc., using only natural language. And anyone who can code gets a powerful assistant. This marks the beginning of a new era of coding and will be an important building block for AI engineers.
While other AI coding tools such as GitHub Copilot, Codeium or Poolside often come as add-ons, Cursor's AI is deeply integrated into the editor. And if you like, you can even operate it via voice. Models such as GPT-4 and Claude 3.5 are running in the background.
Another super practical feature is that you can simply drag a screenshot of a design into the chat and the AI can generate corresponding code. This allows you to turn your creative visions into functional prototypes or rebuild existing apps at lightning speed.
And thanks to Replit, this can also be deployed immediately and made accessible to others. This is not only super efficient, but also damn cool ... and the $60 million dollars in funding they’ve just received will certainly lead to even more upgrades from Cursor soon!
📰 AI news quickie: HAIlights from the industry
Now, that we've knocked the sand out of our sandals, it's time to summarize the most important news and updates for you:
OpenAI
Lots of hype about 🍓... OpenAI's "Project Strawberry" promises AI with real thinking skills soon: instead of just generating text, the new technology should be able to solve problems and plan independently.
The number of ChatGPT users has doubled in the last year - 200 million users per week! And API usage has even doubled in just the last few weeks. Wow!
The rapid adoption of ChatGPT is progressing in many industries.
But OpenAI does not want to rest on its laurels and is in talks about a new financing round with a valuation of 100 billion US dollars. The main investors could be Apple (!) and NVIDIA.
But before the next models from OpenAI and Anthropic are released, the government will keep an eye on it ...
Anthropic
Claude Artifacts has now been rolled out on mobile - and it just keeps getting better!
Anthropic also released the "system prompts" of the Claude models - you can now take a look behind the scenes.
Amazon revolutionizes Alexa with Claude. Starting in October, the smart assistant can be upgraded for a fee.
Google's GameNGen (video, PDF) simulates the game Doom in real time, without a game engine. In other words: the AI generates interactive gameplay and simulates a world. Holy moly!
The Gemini Live Assistant on mobile has been expanded to include over 40 languages - get an Advanced subscription, set up English as the main language and your language of choice as the second language, and you can talk with the bot in real time.
Google Meet now also takes voice notes (in English, for now).
Meta
Meta secretly sneaks in a new web crawler to collect a bunch of data for its AI models.
The latest version of the Meta AR glasses will be unveiled soon!
Video & Image
With Runway, you can now extend Gen 3 Alpha videos up to 40 seconds.
Are you a video enthusiast? Try the new tools Hotshot, MiniMax, Open Source CogVideoX, and LTX Studio.
Image generator Freepik Mystic impresses with striking details and prompt accuracy.
Midjourney seems to want to get into the hardware business.
Surveys
According to Gartner's Hype Cycle 2024, Generative AI is already past the peak of exaggerated expectations. Next comes the valley of despair...
But the most important AI companies are all doing well financially.
A survey shows that three quarters of Americans have concerns about AI in areas such as data protection, job loss and fake news.
What is the best way for a company to develop an AI strategy? This playbook provides some initial pointers.
According to the KPMG survey, 78% of CEOs expect rapid returns from GenAI in the next three years.
At the same time, 56% of Fortune 500 companies see AI as a "risk factor".
MIT has published a comprehensive database (PDF, table) with over 700 AI risks. You can now look up what could threaten you in the AI revolution.
Industry
A new Californian law wants to make AI companies liable (details here) ... and the FTC bans AI-generated fake reviews and testimonials.
These are the most important AI companies in the field of medicine and research.
Magic.dev has introduced an AI with a 100 million token context window - a new record! And raises $320 million in investments.
Cerebras has developed a new solution that is almost twice as the prior leader Groq.
NVIDIA continues to report strong growth, but the competition is increasing!
That's it! Thank you for joining us.
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.
