- Horizon AI
- Posts
- Samsung Unveils Its Own ChatGPT Competitor ๐ฑ
Samsung Unveils Its Own ChatGPT Competitor ๐ฑ
+ Figma Enhances FigJam Whiteboard with New AI Capabilities
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/0e8c9b02-aece-407f-b87c-2d649a71fdd0/WhatsApp_Image_2023-08-18_at_11.51.49_PM.jpeg)
Welcome to another edition of Horizon AI,
Samsung has unveiled its own generative AI model. How will it fare against the competition? We can expect to see more ChatGPT alternatives as companies rush to establish their presence in the AI race
Let's jump into it!
Read Time: 3.5 min
Here's what's new today in the Horizon AI
Samsung Unveils Its Own ChatGPT Competitor ๐ฑ
Figma Enhances FigJam Whiteboard with New AI Capabilities
Free ChatGPT Guide ๐
AI Research: Google Researchers Demonstrate a New Model for Image Generation
AI Tutorial: Create brand logos using ChatGPT
AI Image of The Day๐จ: Graffiti Murals
AI News
SAMSUNG
Samsung Unveils Its Own ChatGPT Competitor
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/20585908-0e09-44c0-81b7-6b1a5fbf2b54/gauss.png)
Samsung has jumped into the generative AI race with the launch of Samsung Gauss, a generative AI model that consists of three tools: Samsung Gauss Language, Samsung Gauss Code, and Samsung Gauss Image.
Details:
Samsung Gauss Language - This large language model can understand natural language, answer questions, summarize documents, translate text, and more. Samsung plans to integrate it into devices like phones and laptops.
Samsung Gauss Code - Focused on AI-assisted coding, this tool helps developers quickly generate and test code.
Samsung Gauss Image - Image generation and editing feature that uses AI to, for example, upscale images and convert low-res to high-res.
Samsung unveiled Gauss at its AI Forum event just days after OpenAI's own AI announcement. The company said Gauss is currently only available internally but will open up to public users "in the near future."
FIGMA
Figma Enhances FigJam Whiteboard with New AI Capabilities
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5b190ff0-75ca-4755-a1d8-d0567d6129ad/ezgif-2-c6b39663d1.gif)
Source: Figma
Figma has announced the addition of three new AI-powered features to FigJam, its online collaborative whiteboard tool, to improve usability and organization.
Details:
Generative AI helps users create new FigJam boards more easily by generating templates from prompts describing required elements like calendars or timelines.
Automatic sorting organizes the sticky notes on a crowded board into logical groups based on subject matter, owner, or other attributes.
Summarization uses AI to scan all sticky notes on a board and produce a high-level summary, allowing users to quickly grasp key themes and ideas.
The features employ OpenAI models customized by Figma to understand FigJam objects and use cases. Figma states over two-thirds of FigJam's active users are non-designers, underscoring the need to optimize the tool for universal collaboration across teams.
AI RESEARCH
Google Researchers Demonstrate a New Model for Image Generation
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/f0a75008-18a7-4186-9511-a948fbc9fbd4/image.png)
Source: Google Research
Google and UC Berkeley researchers have presented a new AI method for generating images that could replace existing diffusion models. The proposed "Idempotent Generative Networks" (IGNs) aim to map any input to a target image in a single step.
IGN is a "self-antagonistic" model that combines generator and discriminator roles in one, unlike GANs which require separate models.
It attempts to project any input to a target image in one shot, unlike diffusion models that take incremental steps.
Early examples show converting sketches to photos, generating images from noise, and repairing damaged images.
While image quality does not yet match state-of-the-art models, the examples demonstrate the potential of IGNs. The model can handle diverse inputs like sketches or corrupted images.
FREE CHATGPT GUIDE
Get our latest ChatGPT Guide to learn how to build your brand on Linkedin ๐ผ
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/747edc5f-45b3-4ff2-8b77-1f2c97d53275/Final_v2.png)
Download it here.
Not subscribed yet? Subscribe here and get all our free resources!
AI Tutorial
Create brand logos using ChatGPT
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/f61119aa-4e07-4f1b-a9ac-f6c8b820dc5d/image.png)
You can generate stunning logos with the help of GPT-4 Vision and DALL-E 3.
Upload a logo for reference to GPT Vision and add this prompt: "I want to generate a logo that is similar to this one in DALL-E3. Craft a DALL-E 3 prompt that will make the logo for my [Company] similar to the one in the example."
Go to DALL-E3 and input the prompt you obtained before, you will get a logo for your company in a similar style as the one you gave.
You can add more details depending on what you want to obtain.
AI Image of The Day
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/7e34fd81-34f6-437a-b5f7-acf2e774e363/Dise%C3%B1o_sin_t%C3%ADtulo__95_.png)
Graffiti Murals
Send your AI-generated images for a chance to get published here: [email protected]
Source: @xv45h & @aussiejay on Leonardo AI Discord
Thatโs a wrap!