📫 Subscribe | 🗣️ Hire us to speak | 🎓 Online Courses | 📅 Events

On Thursday OpenAI released GPT-5.5, which the company referred to as the “smartest and most intuitive to use model” it has released to date. According to OpenAI president Greg Brockman, GPT-5.5 is “a faster, sharper thinker for fewer tokens compared to something like 5.4.” GPT-5.5 reportedly outperforms its predecessors in agentic coding and knowledge work and has demonstrated facility in mathematics and scientific research. The model is only currently available to paid ChatGPT users.

Taylor Swift’s TAS Rights Management filed three trademark applications to protect her voice and likeness from AI misuse. The filings cover specific audio clips and a detailed visual of her onstage persona. By pursuing federal trademarks, Swift aims to create a stronger legal perimeter against unauthorized deepfakes and AI clones, moving beyond limited state-level publicity rights. While the approach remains untested in court, it marks a significant escalation in how superstars defend their digital identities against generative AI.

It was the best of times, it was the worst of times, at least in this tale of two AI partnerships.

Google has announced an expanded investment in Anthropic, the company behind the Claude family of models, committing an initial $10 billion with the option to invest up to $40 billion more. The deal also gives Anthropic increased access to Google Cloud infrastructure.

Meanwhile, OpenAI and Microsoft have updated their partnership on notably different terms. OpenAI gained the freedom to serve customers on any cloud provider rather than being tied to Microsoft Azure, along with the ability to restrict Microsoft's access to future OpenAI technology. In exchange, Microsoft will no longer receive revenue share payments from OpenAI, though OpenAI will continue those payments through 2030, subject to certain conditions.

These shifting alliances add further intrigue to the evolving LLM landscape.

Lean TECHniques works alongside Midwest companies to leverage technology for business impact. Through advisory, coaching, and embedded delivery, Lean TECHniques helps organizations develop high-performing teams, create environments that fuel innovation, and deliver best-in-class customer experiences. Core areas of expertise include software, cloud, artificial intelligence, product management, and technology leadership.

OpenAI shipped GPT-Image-2 this week, and the real headline isn't better pictures. It's that the model actually thinks before it draws. It scored 1,512 on Image Arena, 242 points ahead of the next closest competitor (Nano Banana). Text inside images finally works, including non-Latin scripts like Japanese, Korean, and Bengali. A new Thinking mode reasons through the layout, searches the web for references, and can produce up to eight coherent images from a single prompt. OpenAI's framing that "images are a language, not decoration" is the real shift, because image generation just joined the same advanced level text models entered a year ago.

The shape of the state is definitely off, but I would say ChatGPT Image 2 showed itself off pretty well here.

Odyssey released Odyssey-2 Max, the company’s most powerful world model to date. It simulates physical environments in real time, which some researchers say is the next frontier after language models. See it in action: