Sponsored by
Hello members.
This newsletter is the first to be sent to everyone in 2024.
For those who want to unsubscribe from Insiderly.ai AI updates, you can click the Unsubscribe link at the end of the email.
This year, the movement in the global AI industry is very busy.
In the past week, there have been 4 very interesting AI news:
1. The OpenAI drama is here again
When Elon Musk filed a court case, accusing Sam Altman of betraying OpenAI's goal of benefiting humanity.
Focus on business interests, especially with Microsoft
This goal was previously included in the Founders' Agreement.
In the picture, this is the email that Sam talked to Elon when he set up OpenAI in June 2015.
I don't know what the impact will be. Please follow
Details of the lawsuit filed with the court See here
2. Alibaba Research Team Develops EMO Model, Transforming Still Images to Move, Speak, and Express Emotions
The model is an Audio2Video Diffusion model called EMO or Emote Portrait Alive.
It helps to create realistic and lively portrait videos.
It is cooler than the previous image creation technology that gives a picture of a person speaking who looks stiff and emotionless.
What can EMO do?
- Change the still image to move your mouth to match your voice.
- Facial expressions are consistent with speech.
- Let it sing in a variety of styles.
- Create videos up to 1 and a half minutes long.
How does EMO work?
EMO creates videos using audio and human images, just upload photos and audio files, and you will get a smooth, realistic video that captures all the details and uniqueness of the person.
Compared to before, it was necessary to take 3D images or capture facial movements with a camera.
It is another wow development in the AI world today from Alibaba.
China's AI Developments It's not very exciting, but it's important.
3. Klarna: Success Case of using Chatbot to work in Customer Services
Klarna, a Swedish digital payment company. Available worldwide in 45 countries
It has over 150 million active users and over 500,000 merchants that support the payment system.
announced the success of bringing AI Assistant from collaboration with OpenAI to customer service simultaneously around the world.
detail
- Over 2.3 million customer conversations, or two-thirds of all customer conversations, are managed with AI Assistant.
- This AI Assistant works as much as hiring over 700 agents on a full-time basis.
- Klarna hires more than 3,000 full-time employees to do customer service + outsource to partners.
- In terms of customer satisfaction, AI Assistant scores the same Customer Satisfaction Score from customers as it does when using an agent as an answerer.
- Reduce repetitive questions by 25%
- Help solve the problem for the customer within 2 minutes, compared to 11 minutes when the agent is the one who does it.
- Available in 23 countries, 24/7 and available in over 35 languages.
- There is still an option to chat with the staff.
- Klarna is estimated to earn $40 million in revenue in 2024 from the use of AI Assistant in this customer service event.
You can see that Klarna is clearly measuring ROI. Metrics are used for improvement, such as waiting time, repetitive queries compared between humans and bots.
It will be interesting to see if the Customer Services case will be replaced by AI 100%.
From the Klarnar CEO's tweet: https://twitter.com/klarnaseb/status/1762508581679640814
Official Press Release :
4. Create video games in one prompt with Genie from Google
After letting Sora, a Text-to-Video Games model from OpenAI, dominate the AI world wow scene for several days in a row.
Last night, Google DeepMind released a new research that shook the world again.
With a new AI model that can create video games in a single prompt.
This model is called "Genie".
- It has been trained with over 200,000 hours of gaming content with over 11 billion parameters to help it understand visual scenes, interactions, and other elements of the game.
- You can create video games that can be played. Actually control the characters in the game. The image of the Mario game floated immediately.
- The games created from this model are action-controllable 2D worlds.
- The quality of video games that can be created right now is not high (1 FPS = 1 frame per second). Famicom Game Mood
- Create game scenes, create actions such as how to walk. Jump of the game character, including the level of the game.
- It supports prompt text-to-video games or converts images into 2D games or image-to-video games.
What is expected to be the consequences
- In the video game industry, of course, game creators should make it easier to create various components in the game or make demos of games.
- Gamers may be able to create simple video games to play by themselves, or they may also be able to sell them.
- Virtual Worlds and the Metaverse are likely to hit as well. This is the beginning of creating different virtual worlds.
- Because the model can be trained to be applied to various robotics.
The quality may not be able to compete with modern games, but it is a very interesting starting point.
There are many more use cases that can be used.
Sora can create a beautiful 1-minute short video.
But people can't interact with it. You can only see it.
But Genie people can interact with what they create.
It's very addictive.
Source: From X of one of Deep Mind's Team Leads, Tim Rocktäschel ( https://x.com/_rockt )
Next Newsletter It will be sent to you on Sunday. If you find it useful, update your knowledge and movements in the AI industry, please advise me. Tell your friends.
I am currently working on a small e-book to distribute the 2024 updated version of Generative AI Tools to people who are interested. This March, it will be distributed to all members for free.
Thank you to everyone who followed.
Insiderly.ai Team
last Allow me to introduce another foreign website that updates interesting AI content. The name is The Rundown AI, so it is recommended that you press Register. Updates on news and interesting AI tools
Learn how to make AI work for you.
How do you stay up-to-date with the insane pace of AI? Join The Rundown – the world’s fastest-growing AI newsletter with over 500,000+ readers learning how to become more productive using AI every morning.
1. Our team spends all day researching and talking with industry experts.
2. We send you updates on the latest AI news and how to apply it in 5 minutes a day.
3. You learn how to become 2x more productive by leveraging AI.