AI

the tricky part is that AI changes weekly. So how can we be concrete about something so fluid?

Here’s how I’ve started to think about it: Be flexible about tools, but concrete about values.

Students don’t need us to predict the future of AI. They need us to articulate the principles that guide our choices. That might be things like:

Transparency: Always disclose when AI is used. Integrity: Use AI to assist thinking, not replace it. Learning: Choose methods that strengthen your own skills. When students internalize these values, they can adapt them to whatever new tool emerges next semester: Claude, Gemini, Perplexity, or something we haven’t heard of yet.

A good AI policy, like a good syllabus, isn’t a list of prohibitions. It’s a shared framework for reasoning through change.

“AI should not be used to put words in anyone’s mouth,” Sen. Elizabeth Warren told NOTUS. “AI is creating something that does not exist, and when our politics head down that path, we’re in trouble.”

Democratic Sen. Andy Kim told NOTUS that adopting AI in political ads could lead politics “down a dark path.”

“We need to be very strong and clear from the outset that it would be wrong and really disastrous for our democracy when we start to see those types of attacks,” Kim said.

Amazon is trialing AI-powered smart glasses that give delivery drivers hands-free scanning, navigation, safety cues, and proof-of-delivery to speed up last-mile routes.

More Insights:

Glasses overlay hazards and tasks; scan packages, guide turn-by-turn on foot, and capture delivery proof.

Auto-activate when the van parks; help find the right parcel in-vehicle and navigate complex apartments/businesses.

Paired vest controller adds physical controls, a swappable battery, and an emergency button.

Works with prescription and light-adapting lenses; pilots underway in North America ahead of broader rollout.

Roadmap: wrong-address “defect” alerts, pet detection, and low-light adjustments; launched alongside “Blue Jay” warehouse arm and Eluna AI ops tool.

Why it matters: If AR meaningfully cuts seconds per stop and reduces errors, it could reshape the economics—and safety—of last-mile logistics, signaling a future where AI quietly augments every movement of frontline work.

Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory.

Key findings:

45% of all AI answers had at least one significant issue. 31% of responses showed serious sourcing problems – missing, misleading, or incorrect attributions. 20% contained major accuracy issues, including hallucinated details and outdated information. Gemini performed worst with significant issues in 76% of responses, more than double the other assistants, largely due to its poor sourcing performance. According to the Reuters Institute’s Digital News Report 2025, 7% of total online news consumers use AI assistants to get their news, rising to 15% of under-25s.