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

OpenAI has pledged a $5 million donation to the Simons Foundation following heavy criticism from mathematicians over its AI-generated research. The backlash erupted after OpenAI announced the solution to hundreds of open math problems using its own models, drawing accusations of a disregard for scientific research norms from the Association for Human Mathematics (Porter wrote about this in Friday’s newsletter). The funding is part of a $15 million combined initiative with Anthropic and the Simons Foundation. The resources will support workshops, conferences, and provide researchers with access to AI models and computing power to help mathematicians examine, understand, and build upon AI-produced results.

Major institutions including Charles Schwab, JPMorgan Chase, Amazon, and Delta are actively blocking third-party AI agents, such as Meta's Muse, from logging into customer accounts or executing transactions. Financial leaders cite severe liability risks regarding automated tools moving funds and handling credentials, though Schwab plans to allow outside agents limited access via secure MCP servers

Google publicly launched Gemini 4 Argon, the first model in its new Gemini 4 generation, making it available to Google AI Pro and Ultra subscribers in the Gemini app. Built for complex tasks in software engineering, enterprise work, and cybersecurity defense, the frontier model features improved reasoning, reduced clichés, and lower introductory API pricing than key competitors. Google also integrated Antigravity's agentic coding interface on macOS, added a tasks mode for background workflows, and expanded app connectivity to over 40 services.

Microsoft AI CEO Mustafa Suleyman announced that the company has developed a new in-house AI model, MAI-Base-2, which outperforms a comparable OpenAI model while using just a quarter of the computing power. During a recent employee town hall, Suleyman also highlighted that the model achieved significant efficiency gains compared to offerings from competitors like Chinese model DeepSeek.

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.

ElevenLabs’ new Instant Soundtracks takes one of the most tedious parts of video editing off your plate – upload a video, and it scores it with original music and adds sound effects timed to the action, like a whoosh when your logo appears. It can also shift the mood between scenes and match the style of audio already on your clip, and it puts every effect on its own layer so you can tweak it without starting over. You can try it free, but the free plan won’t let you download anything, so plan on the $6/month Starter plan, which covers commercial use online.

My favorite part of this prompt (one of the example prompts supplied by ChatGPT) is the last line.

Back in 2025, I (Professor Snider here) wore a $50 Bee on my wrist for two weeks. The idea was great – an AI that listens to your day and turns it into recaps and to-dos. But it was early. It knew my voice and nobody else’s, and it kept mistaking podcasts and TV shows for real conversations.

Now Bee is part of Amazon, and it just launched two new devices. Bee Frontier ($74.99, available now) is the next-gen wearable you can put on your wrist or clip to your shirt. Bee Eclipse ($99.99, coming soon) is a credit-card-sized device that sticks magnetically to the back of your phone and lasts 21+ days per charge. Bee now plugs into Alexa+, so you can ask your Echo for a meeting summary or your to-do list. It also links to Amazon.com to surface products you mentioned in conversation.

AI researcher Andrej Karpathy posted on X that the shared understanding of large language model capabilities is rapidly widening. He describes the dynamic as a sharp funnel, noting via napkin math that roughly 75 percent of the global population has barely interacted with advanced frontier models. And just around 5,000 people have truly seen what AI can do.

Artificial General Intelligence (AGI): An AI system that can perform a broad range of thinking tasks at a level comparable to or beyond humans, including adapting to unfamiliar problems. Today’s tools already handle many different tasks, but their abilities are uneven. A model can produce an impressive draft and still struggle with reasoning or carrying a complicated project through to completion. AGI is the idea of much broader, more adaptable intelligence: a system that can learn new tasks and apply what it knows across different situations. That would not necessarily make it flawless. Humans make mistakes, and general intelligence does not mean being good at everything.

The catch is that there is no universally accepted finish line. OpenAI’s charter describes AGI as highly autonomous systems that outperform humans at most economically valuable work. Google DeepMind researchers have proposed levels ranging from “emerging” to “superhuman,” based on the breadth and strength of a system’s abilities. Others, including Meta’s former chief AI scientist Yann LeCun, argue that the term itself is misleading because even human intelligence is specialized.

Miss out on any previous AI terms of the week? Check out our Glossary of AI terms, where we’re collecting all entries in the series. And if you want us to define a term in an upcoming newsletter, let us know!

The Innovation Profs are Drake University Professors Chris Porter and Chris Snider. Porter leads Drake’s Artificial Intelligence major and Snider is a multimedia professor in the School of Journalism and Mass Communication.