The First LLM Family in 2025 | Weekly Edition

PLUS HOT AI Tools & Tutorials

Jan 03, 2025

Hello and welcome to our weekly roundup!

How are your first few days of the new year? Hopefully good, because it's time to get back on track! Although 2025 has only just begun, many companies and developers already have something to share.

Let's discuss what happened in AI over the past week.

This Creators’ AI Edition:

Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁

Your advertisement could be featured here!

Sponsor a spot in our newsletter to connect with subscribers interested in AI, technology, and startups. For inquiries, click the button below.

Featured Material 🎟️

Rubik's AI Reveals Its First Family of LLMs

An unexpected but important release at the very beginning of the year. Startup Rubik's AI introduced Sonus-1, its first family of large language models. It includes four different options: Mini, Air, Pro, and Pro (w/ Reasoning). Images provided by the developer show that the flagship Pro (w/ Reasoning) model is on par with top solutions, including o1 by OpenAI and Llama 3.1 by Meta.

Here's how Rubik's AI describes its models:

Sonus-1 Mini: For when speed is of the essence, offering high-speed and cost-effective solutions
Sonus-1 Air: A versatile model with a balance of performance and resource usage.
Sonus-1 Pro: Top-tier model, optimized for complex tasks requiring the best performance.
Sonus-1 Pro (w/ Reasoning): Flagship model featuring chain of thought reasoning for even more capability.

Keep your mailbox updated with key knowledge & news from the AI industry

In short, I advise paying attention to Sonus-1 Pro (Reasoning). This AI excelled in several tests, especially reasoning and math problems, demonstrating a good ability to outperform other proprietary models. The developer claims that AI experts are already using its solutions.

However, we can’t confirm this information.

Good news: You can already try the Sonus-1 suite at chat.sonus.ai, including Pro Reasoning, for free. Right after the release, the company also integrated image generation with Flux into its platform.

Rubik's AI picked an interesting moment to pick at the market. While OpenAI, Google, and others are gaining momentum and preparing their following products, a small startup has decided to fill the space with its models.
That said, virtually nothing is known about the startup right now. So, I urge caution and treat it with a little bit of distrust. Whether Sonus-1 is worth attention against the background of all the famous “big guys” we will find out very soon.

News Of The Week 🌍

OpenAI Lays Out Its For-Profit Transition Plans

OpenAI has announced plans to transition from its non-profit structure to a for-profit Public Benefit Corporation (PBC) by 2025. This restructuring aims to enhance its ability to attract significant investments necessary for the development of AGI. The decision was communicated in a blog post, where OpenAI emphasized the need for a clearer financial framework to meet the capital demands of large-scale AI projects.

Under the new structure, OpenAI's for-profit arm will take control of business operations, while its non-profit division will focus on charitable initiatives.

OpenAI's board acknowledged that the current structure limits its ability to attract conventional equity investments. The company has raised $6.6B thus far but estimates it will need to secure hundreds of billions more to remain competitive.

Research: AI Tools May Soon Manipulate People's Decisions

Researchers from the University of Cambridge have raised alarms about the potential for AI tools to manipulate human decision-making in a new marketplace termed the "intention economy." This concept suggests that AI could soon predict and influence users' choices based on their behavioral and psychological data. The study indicates that AI systems, like ChatGPT, could analyze users' online activities and personal interactions to anticipate their intentions.

This capability would allow companies to not only forecast decisions—ranging from consumer purchases to political votes—but also sell this information to advertisers, creating a lucrative market for "digital signals of intent" before users even make their choices.

I don't want to be overly skeptical, but this is one of the most off-kilter research studies I've read in a while. And I don't want to defend big tech, but like... Cambridge researchers just described how digital marketing works.

Sharing is caring! Refer someone who recently started a learning journey in AI. Make them more productive and earn rewards!

Refer a friend

Samsung Invests $181М Into Rainbow Robotics

Robotics stocks soar over positive outlooks on future growth - The Korea Times

And another big news for the world of robotics. Samsung has announced an approximately $181M investment to increase its stake in South Korean startup Rainbow Robotics from 14.7% to 35%. This transaction makes Samsung the largest shareholder in the company. Rainbow Robotics specializes in humanoid robots and advanced robotic systems. The startup is known for developing South Korea's first two-legged humanoid robot, Hubo.

The investment will allow Samsung to integrate Rainbow Robotics as a subsidiary and establish a Future Robotics Office to enhance innovation in robotics. The deal is expected to close in February 2025.

Jetson Thor: Nvidia's Upcoming Release for Humanoid Robotics

According to the Financial Times, Nvidia believes 2025 is "The ChatGPT Moment" for physical AI and robots. To that end, the company is set to launch Jetson Thor in the first half of 2025. This new compact computer will feature the Blackwell architecture, offering an 8X increase in GPU performance and 2.6X improvement in CPU capabilities compared to Jetson AGX Orin, and support for up to 128 GB of RAM. The Jetson Thor aims to enhance AI operations like real-time object detection and natural language understanding.

The global robotics industry is expected to grow to $165B by 2029.

Gemini is Google's 'Main Focus' for 2025

Google has set its strategy for 2025. CEO Sundar Pichai said the company will focus on developing its Gemini model as its flagship product over the next 12 months. Google's goal is to increase the audience of regular users for its chatbot to 500 million people. Pichai noted that while there has been strong momentum with the Gemini app, significant work remains to close the gap with competitors like OpenAI.

He called 2025 a pivotal year for Google. Now, Gemini must solve real user problems and strengthen the company's position in the AI market.

Most of 2024 didn't go well for Google. AI Overview was an incredibly raw product, and the company faced much criticism over its image generator. Let’s hope things are better this year.

Useful Tools ⚒️

Erayaha.AI – Smarter contract reviews & insights in MS Word / Google Docs

Automaticall – Warm up your leads with AI phone calling

Monkt – Transform files and web pages into AI-ready Markdown or JSON

DataMorf – Build automated workflows for complex tasks

FirstVersion – Rebuilding the first versions of great products with AI

Instant3d.ai – Copilot for generating 3d assets for Blender/Unity

Instant3D.ai is a platform that allows you to generate 3D assets from text prompts in a short time, typically under 20 seconds. It employs a feedforward neural network to create a 3D triplane based on the input text, which is more efficient than traditional modeling methods that can take hours. The technology uses techniques such as score distillation sampling and CLIP loss to ensure the generated models align closely with the provided descriptions.