AI revolution: from Duolingo replacement to neural networks for presentations, videos, and fonts – new and updated in 2024

 

1. AI replacement for Duolingo. A neural network has been found that will help with language learning – it can be used to memorize new words through mini-games. Babadum allows you to adjust to your language level and help you improve your vocabulary faster.
2. Check the accuracy of the information in seconds. This neuron searches only for reliable sources, weeding out fakes and myths. Simply enter a query that raises doubts and get a clear answer without unnecessary information.
3. Suno has received an impressive update. Now you can generate songs from photos or even videos. Under the hood is a model that analyzes events in the file and adjusts the mood of the song and lyrics to a given topic.
4. ChatGPT is now available on Windows. OpenAI has finally released a desktop version of the chatbot for Windows – you can download it from here, it works only for paid subscription holders.
5. Now you can generate videos with believable gestures. TANGO neural network has been released – it takes a video reference and your audio and generates not only lipsync but also realistic gestures that match the tempo of the audio. Under the hood, there is a full-fledged combine: during the generation process, a 3D model of a person is created and animated in Blender. There is already a demo, but the developers warn that it is not very good.
6. A new AI assistant for Excel. The WalloAI neural network analyzes your projects and helps to select formulas and format your entire Excel spreadsheet.
7. Google’s GenType will create any font to your liking. If you suddenly run out of free fonts on the Internet and want something extraordinary, this service is for you. You can generate a font from food, from keyboards, or from spaceships – just type in the font and the neuron will draw anything.
8. The Unwatermark neural network effectively removes any watermarks from images, even the most subtle ones. The tool works right in the browser and is completely free.
9. Hassaan Raza has created the first AI bot of its kind that communicates with users in a video call format. Unlike OpenAI-o1, this neuron reacts to the environment and is able to maintain a dialog on its own. The AI interlocutor can even get offended and leave the user.
10. Ideogram releases its Photoshop killer. Ideogram released Canvas, a neural network that allows not only generating images but also changing them. There is only one drawback: they ask for money to access the features.
11. ElevenLabs has received an update. Now it is possible to generate voices with text prompts. Neuron has completely redesigned Voice Design – you just need to describe the voice you want and provide test audio. There are no other restrictions except for the number of voices.
12. AI that can control your PC. Under the hood of this tool is ChatGPT’s main competitor, Claude, which can do almost anything: order your food, choose train tickets, or even find a job. Installation manual.
13. Runway has released the Act-One feature. In fact, it is a video-to-video generation where AI turns your facial expressions and gestures into a generated character. Take a closer look at the first video – both characters are played by the same actor.
14. AI from Stanford generates vast research on any topic. The STORM neuron is your personal academic that writes essays, articles, and PhD-level coursework in one click.
    – A detailed introduction and a short brief.
    – Scours the entire Internet for information.
    – Sews references and quotes into the text to support the thesis.
    – You can download the fully finished result in PDF.
    – It is absolutely free.
15. AI generates ADHD videos in the background of Minecraft parkour. The service works very simply: you specify a PDF document on the desired topic, and the models will turn it into a typical buzzer video. Subway Surfers is also available.
16. A full-fledged desktop version of ComfyUI is coming out. For those who don’t know, ComfyUI is a popular and probably the best way to run local models to create images. In addition to the embedded Python, the developers have also redesigned the interface to make it more comfortable and understandable. It works on all current platforms: Windows, MacOS, and Linux.
17. AI sound for videos for free. The service allows you to customize the voice, choose different speakers, and support a lot of languages. It works without registration and allows you to make voiceover in Ukrainian.
18. A new neural network for presentations. SlidesAI creates presentations on demand and allows you to easily edit them, so you no longer need to work with slides manually.
19. Entering all text on a computer by voice. Wispr Flow integrates voice input into a whole bunch of Windows applications, from instant messengers to text editors. The neural engine can even understand pauses and accents in your speech. After recording, the text is carefully formatted.
20. New AI for designers. A very useful neural network for designers has been released – Recraft V3, which includes two models:
    – The first one is responsible for generating very realistic images comparable in quality to Midjourney or DALL-E.
    – The second one generates vector graphics – it’s not very good with three-dimensional images yet, but logos are fine.
21. There is also an editor where you can combine your text and a picture in a given style, which is an extremely unusual feature.
22. Google’s new neuron will teach you anything. Learn About will help you learn any topic quickly, easily, and clearly. It’s simple: enter your question or specify a topic and get an interactive course with pictures! The tutorial is built on the go – you go deeper into the topic or ask questions, and AI continues to generate new pages. And all this is free. To access it, you need to enable a VPN in the United States.
23. Hedra update was released. Now the neural network can animate your photo and make a lip sync with your audio. Upload your photo, record your voice, and get a ready result where your photo comes to life and sounds like audio.
24. PixVerse update. The new V3 version of PixVerse has several new features, so the video generator has become more powerful and faster:
    – There are many video effects that resemble those of Pika – there is a “crush” effect and a “Lego explosion”.
    – AI can also do lip-syncing: video characters can speak much more “lively” and in sync with the audio.
    – Other innovations include support for different aspect ratios and a choice of video style and duration.