OpenAI’s Sora 2 takes social media by storm, while Thinking Machines Lab announces its first-ever product, and Anthropic updates its core technical team.
Here are our key takeaways in more detail:
OpenAI’s Sora 2 makes its grand reveal together with a new social media app for iOS
Mira Murati’s Thinking Machines Lab announces Tinker, a powerful tool for fine-tuning AI language models
Anthropic appoints former Stripe CTO Rahul Patil as its new chief technical officer
Join us at AI Tangle as we untangle this week's happenings in AI!
AI LIGHTNING ROUND
💽 Embracing AI in Music - Major record labels Universal Music and Warner Music are reportedly closing in on landmark AI licensing deals with Suno, Udio, and more.
☄️ A Comet For Everyone - Perplexity is making its AI-powered Comet browser free for everyone, while paid users get a new "background assistant."
💸 Financing Competition - Rising upstart Cerebras has recently locked in $1.1 billion in funding as it looks to go head-to-head with Nvidia in AI hardware.
💪 AI-Powered Fitness - Peloton has announced a sweeping revamp across all its exercise machines with added AI ahead of the holiday season.
5 QUICK HITS
Following up on last year's success, OpenAI has released Sora 2, its most capable video and audio generation model, announced in a recent debut stream. The model comes with improved realism and, importantly, better adherence to physics and scene consistency, with OpenAI demoing a few noteworthy examples. Alongside the model came the Sora app for iOS, a TikTok-style clips sharing social media app with a feature to "cameo" in friends' videos. Despite being primarily invite-only (with the exception of ChatGPT Pro users), the Sora app hit #3 on the App Store just a few days after launch.
Thinking Machines Lab (TML), a stealthy AI startup co-founded by former OpenAI executives Mira Murati and others, has launched its first-ever product: Tinker, a tool that automates fine-tuning of frontier AI models. Backed by $2 billion in seed funding, TML wants to make model customization more accessible by removing technical barriers to reinforcement learning and supervised training. Although Tinker only supports Meta's Llama and Alibaba's Qwen models for now, beta testers have been very positive about TML's debut tool, notably its ease of use and power.
Former Stripe CTO Rahul Patil has recently joined Anthropic as its new CTO earlier this week, as part of a recent update to the structure of its core technical team. Patil takes over co-founder Sam McCandlish's position, who has now become Anthropi's chief architect. As CTO, Patil will report directly to Anthropic president Daniela Amodei and oversee the company's compute, infrastructure, inference, and a range of engineering tasks, bringing with him 20 years of engineering experience throughout his career.
Not long after its $100 billion compute deal with Nvidia, OpenAI has announced a partnership with South Korea's Samsung and SK Hynix to sate Project Stargate's demand for chips. The pair said they would scale up production to meet OpenAI's needs, with the AI startup expecting roughly 900,000 semiconductor wafers each month from the both of them. At the same time, OpenAI is also reportedly looking into opening two new AI data centers in South Korea, with Samsung exploring the possibility of floating data centers.
Uber has recently acquired Segments.ai, a multi-sensor data labeling startup specializing in lidar annotation for robotics and autonomous vehicles, from drones to self-driving cars. As part of the deal, Segments.ai's team, including CEO Otto Debals and co-founder Bert De Brabandere, will join the Uber AI Solutions team to "build the future together for lidar data annotation." The strategic acquisition looks to improve Uber's worldwide services while strengthening its decade-long work in AI, safety, and autonomy.
3 AI TOOLS
Huxe - Struggling to keep up with everything? Huxe is your personal superintelligence that gets you up to speed in the mornings, from five-minute morning briefs to podcasts that track your world.
AI EXTRA READ
Prompt Engineering VS Context Engineering (5-min read)
As prompt engineering continued to evolve in applied AI, a new term has begun to take the spotlight. In this research post by Anthropic, discover the emerging art of context engineering, the difference between it and prompt engineering, and how to make the most of it.
MORE FROM THE AIE NETWORK
📻 AI Confidential Podcast - Are LLMs Dead?
🔮 The Artificially Intelligent Enterprise - AI as Your Career Accelerator
💡 AI CIO - The Paradox of AI Reports
🎯 The AI Marketing Advantage - New AI Features Power Google Ads This Season
What did you think of this newsletter? Let us know!
Your AI Sherpa,
Mark R. Hinkle
Publisher, The Artificially Intelligent
Enterprise (TheAIE) Network
Connect with me on LinkedIn
Follow me on X
