In partnership with
Happy Friday, AI & Data Enthusiasts! Big moves in AI today! OpenAI has unveiled GPT-4.5 ‘Orion,’ its most advanced model yet, while also dealing with GPU shortages delaying its rollout. Meta’s new Aria Gen 2 glasses are pushing AI research forward, and Microsoft is bringing Copilot to macOS for a dedicated AI assistant experience. Let’s dive in!
In today’s edition:
🚀 OpenAI Unveils GPT-4.5 'Orion'
🕶️ Meta Launches Aria Gen 2
🧠 GPU Shortage Delays GPT-4.5
💻 Copilot Now Available on Mac
- Naseema Perveen
WHAT CAUGHT OUR ATTENTION MOST
🚀 OpenAI Unveils GPT-4.5 'Orion' – its Largest AI Model Yet
OpenAI has officially launched GPT-4.5, code-named Orion, its largest AI model yet. Trained on more computing power and data than any previous model, GPT-4.5 brings improvements in reasoning, emotional intelligence, and creative capabilities.
Key Features and Availability
Bigger, but not a ‘Frontier’ Model – OpenAI initially stated that GPT-4.5 isn’t a frontier AI model, then quietly removed that claim from its white paper.
Limited Access – GPT-4.5 is available to ChatGPT Pro users ($200/month) and select API developers now, with broader access coming next week.
Performance Gains – The model excels at factual accuracy, coding, and creative tasks like design and writing. However, it still lags behind specialized reasoning models in complex problem-solving.
The Cost of Scaling Up
GPT-4.5’s advanced capabilities come at a steep price. OpenAI admits the model is extremely costly to run, charging developers $75 per million input tokens - far pricier than GPT-4o. This raises questions about the long-term viability of such massive AI models.
A Sign of AI’s Future?
Despite its mixed performance, GPT-4.5 represents a step toward OpenAI’s next big leap: combining its GPT series with “o” reasoning models in GPT-5. Whether this hybrid approach will redefine AI’s future remains to be seen, but one thing is clear - scaling alone is no longer enough to push AI forward.
IN PARTNERSHIP WITH THE AI REPORT
There’s a reason 400,000 professionals read this daily.
Join The AI Report, trusted by 400,000+ professionals at Google, Microsoft, and OpenAI. Get daily insights, tools, and strategies to master practical AI skills that drive results.
WORTH WATCHING
The impact of artificial intelligence in 2025
Join experts from Duke, Google, and IBM as they discuss the latest advancements, challenges, and future implications of AI. From ethical concerns to groundbreaking innovations, this panel will dive into the evolving role of AI across sectors.
Don’t miss this insightful conversation!
KEEP YOUR EYE ON IT
🕶️ Meta Unveils Aria Gen 2 for Advanced AI Research
Meta has introduced Aria Gen 2, the latest version of its research-focused AR glasses, designed to push the boundaries of machine perception, contextual AI, and robotics. This next-generation hardware enhances research capabilities with upgraded sensors, improved AI processing, and extended usability.
🔬 Key Enhancements in Aria Gen 2
Advanced Sensor Suite: Includes an RGB camera, SLAM cameras, eye-tracking sensors, spatial microphones, and new heart rate monitoring and voice-isolating microphones.
On-Device AI Processing: Powered by Meta’s custom silicon, supporting SLAM, eye tracking, hand tracking, and speech recognition.
Extended Usability: Lightweight (75g), foldable design with up to eight hours of battery life.
Audio Interaction: Features open-ear force-canceling speakers for real-time user feedback.
🎯 Expanding Research Applications
Aria Gen 2 builds on the success of Project Aria, which has fueled advancements in computer vision, robotics, and accessibility. Early adopters, including Envision, are exploring its potential for assistive technologies, such as AI-driven indoor navigation for blind and low-vision individuals.
Meta plans to roll out Aria Gen 2 to academic and commercial research labs, with further details on availability coming soon. Researchers can sign up for updates to gain early access.
🧠 OpenAI’s GPU Shortage Delays GPT-4.5 Rollout
OpenAI CEO Sam Altman has revealed that the company is “out of GPUs”, forcing a staggered rollout of its latest model, GPT-4.5. While early access is starting for ChatGPT Pro users, others will have to wait—raising questions about AI infrastructure bottlenecks and costs.
🔧 GPU Shortage Halts OpenAI’s Expansion
Altman acknowledged that OpenAI’s rapid growth has outpaced its GPU supply, delaying broader access to GPT-4.5. The company plans to add tens of thousands of GPUs next week to support wider availability.
💰 GPT-4.5 is Expensive
The model isn’t just big—it’s expensive. OpenAI is charging $75 per million input tokens and $150 per million output tokens, 30x and 15x the cost of GPT-4o, respectively. This pricing has sparked debate about whether larger AI models are becoming unsustainable.
🚀 What’s Next? OpenAI’s Long-Term Plans
To avoid future GPU shortages, OpenAI is looking to develop its own AI chips and expand its data center infrastructure. These moves aim to secure its dominance in AI while reducing dependency on third-party chipmakers.
🔮 The Big Picture
The GPU crunch highlights the growing pains of AI’s rapid evolution. OpenAI’s struggle to meet demand underscores a larger industry-wide challenge—scaling AI infrastructure fast enough to keep up with advancements. While developing in-house AI chips and expanding data centers could offer long-term solutions, these efforts will take years to materialize.
In the meantime, high costs and limited computing power may slow AI accessibility, favoring enterprise customers over individual users. As AI models grow larger and more expensive, the industry will need to balance innovation with affordability to avoid bottlenecks that hinder widespread adoption.
💻 Microsoft Copilot Launches macOS App
Microsoft has officially released a dedicated Copilot app for macOS, bringing its AI-powered assistant to Apple users for the first time. Previously available only via browsers and mobile devices, this standalone app offers a seamless AI experience on Mac.
🚀 What’s New?
AI-Powered Assistance: Copilot can draft emails, summarize documents, generate images (via DALL-E 3), and more.
Mac Compatibility: Requires macOS 14.0+ and an M1 chip or later.
Expanded Availability: Copilot is now accessible across Windows, Mac, iOS, Android, and Microsoft apps like Teams and Outlook.
🆓 Extra Features for All Users
Microsoft also announced that Voice (for language practice) and Think Deeper (for complex reasoning tasks powered by OpenAI’s o1 model) are now free and unlimited for all Copilot users.
Mac users can now download Copilot and explore its full capabilities directly from their desktop.
ICYMI
$$$ MONEY MATTERS
Perfect taps $23M to fix the flaws in recruitment with AI.
Apple commits $500B to US manufacturing.
AI-coding startup Codeium in talks to raise at an almost $3B valuation.
Karman+ digs up $20M to build an asteroid-mining autonomous spacecraft.
Anthropic reportedly ups its next funding round to $3.5B.
That’s all for now. And, thanks for staying with us. If you have specific feedback, please let us know by leaving a comment or emailing us. We are here to serve you!
Join 130k+ AI and Data enthusiasts by subscribing to our LinkedIn page.
Become a sponsor of our next newsletter and connect with industry leaders and innovators.
