AI + Education Weekly Update

Week of May 26th, 2025

To catch up on another busy week in AI + Education news, check out our weekly updates on the key highlights here or follow our CEO on Linkedin.


Week of May 26th, 2025

Google’s Veo3 Video Generator Released

The text to video generator is getting a lot of buzz around its prompt accuracy and audio/ video combinations with ~900,000 visitors last week. Veo has potential classroom applications like generating video book trailers, visualizing complex concepts, and real-world simulations. Advances in video generation like this also highlights the increasing ease of creating hyper-realistic deepfakes.

Day of AI + Common Sense Media’s AI Toolkit for Schools & Families

The toolkit is designed for schools to intentionally communicate with families about AI literacy through avenues like social media, newsletters, and parent teacher conferences, and community gatherings. It includes introductory videos, customizable templates, and activity guides for schools to share.

LLMs Continue to Have Alignment Issues

As new models with increasing capabilities are being released, alignment of the systems continue to be an issue. The goal of alignment is to build AI systems that not only perform tasks effectively, but also do so in a way that is aligned with human intentions, values, and ethics.

Two recent examples of misalignment are o3 refused to be shutdown when explicitly told (OpenAI) and Claude threatening to contact authorities if it perceived wrongdoing.

Anthropic Releases AI Fluency Course and Voice Mode for Claude

The four hour course leads users through eight modules. Each module includes a video, an activity, and a reflection and aims to educate users to “collaborate with AI effectively, efficiently, ethically, and safely”.

Voice mode allows mobile app users to have spoken conversations with the model, allowing the user to switch between text and voice within the same conversation.

Anthropic CEO Warns of Rapid Job Loss Due To AI

In a recent interview, Anthropic CEO Dario Amodei warned that within the next 1–2 years, AI could begin replacing significant numbers of knowledge workers, including roles in customer service, writing, and programming.

He says “We, as the producers of this technology, have a duty and an obligation to be honest about what is coming. I don't think this is on people's radar." Amodei proposes solution such as increasing awareness, a focus on AI literacy, informing public officials, and creating government policy solutions.

Zochi Becomes the First AI to Receive Research Publication Agentically

Zochi, an AI created by Intology, completed the entire research process, including designing experiments, analyzing papers, and writing the reports independently with the exception of humans formatting fixes when necessary. The conference is highly selective, with an acceptance rate of 20%, and often takes PhD researchers years to gain acceptance. This publication represents a milestone in AI-generated and assisted research and could speed up scientific discovery.

Previous
Previous

New RAND Research Reveals Growing AI Training Gap

Next
Next

OpenAI Academy & Claude for Education Debut