😱 OpenAI's $200 ChatGPT subscription

Also: Copilot Vision now available as a preview for some U.S. users

In partnership with

Welcome, AI enthusiasts

12 Days of OpenAI have begun. They're hosting 12 livestreams in a row, announcing a mix of big and small updates. First up, OpenAI's o1 model is now fully released, along with a supercharged Pro subscription offering advanced capabilities. Meanwhile, Copilot Vision has introduced a browsing companion that makes web exploration smoother. And if that wasn't enough, DeepMind's Genie 2 has jumped from flat 2D to immersive 3D, generating stunning, dynamic environments. Let’s dive in! 

In today’s insights:

  • OpenAI rolls out o1 and Pro tier for power users

  • Copilot Vision previewed for smarter browsing

  • Google’s Genie 2 creates 3D interactive worlds

Read time: 4 minutes

🗞️ LATEST DEVELOPMENTS

Evolving AI: OpenAI has launched the full o1 model and a Pro version for people who need more power.

Key Points:

  • OpenAI launched the o1 model with two pricing options —$20 for Plus and $200 for Pro.

  • ChatGPT Pro has more power for hard tasks like coding and legal research.

  • OpenAI also started a grant program to give ten Pro subscriptions to medical researchers.

Details:

OpenAI has introduced a new subscription called ChatGPT Pro, which costs $200 a month. This version gives unlimited access to the o1 Pro model, aimed at professionals needing more power. Data scientists and engineers are pushing ChatGPT to its limits. The Pro model helps solve difficult problems with greater accuracy and speed. The regular o1 model is also available for Plus subscribers at $20 a month, now with vision capabilities, allowing it to analyze images. Tests show that the o1 model performs better on benchmarks compared to the older GPT-4o. OpenAI is also giving grants to support researchers, starting with experts in the medical field.

Source: OpenAI YouTube livestream

Why It Matters:

OpenAI's release of the full o1 model marks a major milestone, giving 300 million users access to the most advanced AI model capable of reasoning. Compared to o1-preview, users can expect faster, more accurate responses, especially for coding, math, and even interpreting images. Plus, it’s trained to be more concise, resulting in quicker replies.

The future of presentations, powered by AI

Gamma is a modern alternative to slides, powered by AI. Create beautiful and engaging presentations in minutes. Try it free today.

Evolving AI: Copilot Vision introduces a companion that reads and helps you browse more efficiently.

Key Points:

  • Copilot Vision is a new feature for Pro users that scans, analyzes, and assists as you browse.

  • Privacy is prioritized: Vision is opt-in and deletes session data after use.

  • Early access is limited, rolling out first for select websites and Pro users in the U.S.

Details:

Copilot Vision, now available as a preview for some Pro users in the U.S., adds a new layer of assistance to your browsing. Embedded in Microsoft Edge, Vision provides contextual insights on web pages, helping you make decisions and gather information without navigating alone. It’s a fully opt-in feature, designed with privacy in mind, and only works on certain websites for now. Feedback will shape its broader release and future capabilities.

Why It Matters:

Copilot Vision aims to turn browsing into a more collaborative activity. Whether planning outings or comparing products, Vision helps refine your decisions by directly analyzing what’s on your screen, potentially saving you time and effort. For digital consumers, this could make sifting through countless tabs a thing of the past — are we seeing the start of a new way to browse?

Evolving AI: DeepMind's Genie 2 moves beyond 2D, now generating dynamic 3D environments.

Key Points:

  • Genie 2 can generate 3D worlds and sustain them longer than previous models.

  • Uses a diffusion model for scene rendering, supporting various viewpoints.

  • Capable of maintaining scene consistency for about 60 seconds.

Details:

DeepMind's new Genie 2 model advances significantly over its predecessor, generating interactive 3D environments in real-time. It uses a diffusion model that renders scenes as the player moves, supporting multiple perspectives and reconstructing scenes when they re-enter view. The model can maintain consistency for up to 60 seconds, though some visual artifacts still appear over time, highlighting current limitations. The key here is: there's no game engine behind this. Everything you see is generated for the first time, in real-time, while being playable.

Why It Matters:

Genie 2 is a game changer. It offers a powerful tool for training AI agents, enabling them to interact with dynamic environments that mimic real-world conditions. It also opens new possibilities for artists and designers needing rapid prototypes of 3D spaces. Eventually, it could lead to more immersive virtual experiences. The creative potential for AI-driven game environments or virtual training grounds is immense — we're in for an exciting future.

 👀 Click on the image you think is real

🎯SNAPSHOTS

Direct links to relevant AI articles.

🏛️ David Sacks is Trump’s AI and crypto ‘czar’.

💰 xAI has raised $6 billion.

🤝 OpenAI is partnering with Anduril, a military contractor.

👩‍💻 Researchers can now detect phishing emails with open-source AI.

📈 Trending AI Tools

  • 💸 Expense Sorted - AI-powered tool that automatically categorizes your expenses (link)

  • 💭 Voicenotes - A place to dump your thoughts (link)

  • 👗 Petals-AI Wear - Unique custom clothing designs with AI (link)

  • 🎬 Zebracat - A tool to create videos from prompts (link)

  • 🎞️ HitPaw - AI upscale video software (link)

Reply

or to participate.