Hidden text can undermine your AI [Brainyacts #268] — from thebrainyacts.beehiiv.com by Josh Kubicki

What if one line of invisible text could change how your AI interprets an entire case file?

This isn’t hypothetical. It’s called prompt injection, and as more legal teams connect LLMs to client files, discovery docs, and contract repositories, the risk is growing, fast.

The AI, following instructions like a good little pattern matcher, now refers to the entire thing using the wrong legal frame.

You never saw the white text.

Your associate didn’t catch it.

The system happily processes it as truth.

And now you have slop in the system. Not because the AI hallucinated but because it was tricked.

 

From DSC:
One of my sisters is a Professor of Communication at a major university and she recently sent me the following article. I thought I’d pass it along to those of you who are also interested in this topic. My sister has written several books on media literacy and has created museum-based exhibits to address this topic.


How to find trustworthy news reporting in these times? — from ericdeggans.substack.com by Switching Codes w/Eric Deggans
I don’t have the perfect answer. But here’s a few ideas.

Here’s a few suggestions for how to navigate our new, challenging news media reality.

  • Find individual journalists to trust…
  • Don’t let the perfect be the enemy of the good…
  • Be skeptical of those telling you what you want to hear…
  • Learn what outlets are good at covering…
  • …and more

The above article links to:

My take on why Americans don’t trust journalists anymore — from ericdeggans.substack.com by Switching Codes w/Eric Deggans
It’s more complicated than journalists failing in their mission. And reversing the numbers are key to preserving democracy.
.


On a somewhat-related note, see:

AllSides Media Bias Chart


 

Eye implant and high-tech glasses restore vision lost to age — from newscientist.com by Chris Simms
Age-related macular degeneration is a common cause of vision loss, with existing treatments only able to slow its progression. But now an implant in the back of the eye and a pair of high-tech glasses have enabled people with the condition to read again

People with severe vision loss have been able to read again, thanks to a tiny wireless chip implanted in one of their eyes and a pair of high-tech glasses.

“This is an exciting and significant study,” says Francesca Cordeiro at Imperial College London. “It gives hope for providing vision in patients for whom this was more science fiction than reality.”

 

 

Amazon’s AWS outage on October 20 knocked services like Alexa, Snapchat, Fortnite, Venmo and more offline for hours — from engadget.com by Kris Holt
A massive outage highlights why relying on a few companies to power much of the internet is far from ideal.

It felt like half of the internet was dealing with a severe hangover on October 20. A severe Amazon Web Services outage took out many, many websites, apps, games and other services that rely on Amazon’s cloud division to stay up and running.

Sites and services that were affected by the AWS outage include:

  • Amazon
  • Amazon Alexa
  • Bank of America
  • Snapchat
  • Reddit
  • Lyft
  • Apple Music
  • Apple TV
  • Pinterest
  • Fortnite
  • Roblox
  • The New York Times
  • Disney+
  • Venmo
  • Doordash
  • Hulu
  • Grubhub
  • PlayStation
  • Zoom

From DSC:
Hmmm…doesn’t this put a bit of alarm in your mind? I can’t help but wonder…if another government wants to wreak havoc on another country — or even the world — that is an increasingly possible situation these days. In fact, its already happened with social media and with cybersecurity-related issues. But taking down banking, commerce, exchanges, utilities, and more is increasingly possible. Or at least that’s my mental image of the state of cyberwarfare. 

 

Goldman economists on the Gen Z hiring nightmare: ‘Jobless growth’ is probably the new normal — from fortune.com by Nick Lichtenberg (this article is behind a paywall)

The challenging U.S. labor market is entering a new normal, according to Goldman Sachs economists David Mericle and Pierfrancesco Mei, who tackled the phenomenon of “jobless growth” in an Oct. 13 note. It resonates with what Federal Reserve Chair Jerome Powell memorably described in September as a “low-hire, low-fire” labor market, in which, for some reason, “kids coming out of college and younger people, minorities, are having a hard time finding jobs.”

Some analysts blame the downturn in entry-level hiring on the impact of AI on the economy, others on macroeconomic uncertainty, especially the seesawing tariffs regime from the Trump administration. The takeaway is clear, though, that getting hired is really hard in the mid-2020s.

This shift is clear in data collated by the investment bank. Payroll growth by industry shows almost all sectors outside health care posting weak, zero, or even negative net job creation, despite otherwise solid macroeconomic indicators. Meanwhile, the share of executives who mention both AI and employment in the same context on earnings calls has reached historic highs.?

For now, Mericle’s “low-hire, low-fire” diagnosis serves as both warning and guide: Jobless growth may not mean mass layoffs, but it does mean fewer opportunities for job seekers and slower rebounds from economic shocks in the years to come.?

 

From siloed tools to intelligent journeys: Reimagining learning experience in the age of ‘Experience AI’ — from linkedin.com by Lev Gonick

Experience AI: A new architecture of learning
Experience AI represents a new architecture for learning — one that prioritizes continuity, agency and deep personalization. It fuses three dimensions into a new category of co-intelligent systems:

  • Agentic AI that evolves with the learner, not just serves them
  • Persona-based AI that adapts to individual goals, identities and motivations
  • Multimodal AI that engages across text, voice, video, simulation and interaction

Experience AI brings learning into context. It powers personalized, problem-based journeys where students explore ideas, reflect on progress and co-create meaning — with both human and machine collaborators.

 

The above posting on LinkedIn then links to this document


Designing Microsoft 365 Copilot to empower educators, students, and staff — from microsoft.com by Deirdre Quarnstrom

While over 80% of respondents in the 2025 AI in Education Report have already used AI for school, we believe there are significant opportunities to design AI that can better serve each of their needs and broaden access to the latest innovation.1

That’s why today [10/15/25], we’re announcing AI-powered experiences built for teaching and learning at no additional cost, new integrations in Microsoft 365 apps and Learning Management Systems, and an academic offering for Microsoft 365 Copilot.

Introducing AI-powered teaching and learning
Empowering educators with Teach

We’re introducing Teach to help streamline class prep and adapt AI to support educators’ teaching expertise with intuitive and customizable features. In one place, educators can easily access AI-powered teaching tools to create lesson plans, draft materials like quizzes and rubrics, and quickly make modifications to language, reading level, length, difficulty, alignment to relevant standards, and more.

 

 

Why Co-Teaching Will Be A Hot New Trend In Higher Education — from forbes.com by Brandon Busteed

When it comes to innovation in higher education, most bets are being placed on technology platforms and AI. But the innovation students, faculty and industry need most can be found in a much more human dimension: co-teaching. And specifically, a certain kind of co-teaching – between industry experts and educators.

While higher education has largely embraced the value of interdisciplinary teaching across different majors or fields of study, it has yet to embrace the value of co-teaching between industry and academia. Examples of co-teaching through industry-education collaborations are rare and underutilized across today’s higher ed landscape. But they may be the most valuable and relevant way to prepare students for success. And leveraging these collaborations can help institutions struggling to satisfy unfulfilled student demand for immersive work experiences such as internships.


From DSC:
It’s along these lines that I think that ADJUNCT faculty members should be highly sought after and paid much better — as the up-to-date knowledge and experience they bring into the classroom is very valuable. They should have equal say in terms of curriculum/programs and in the way a college or university is run.

 

When Heads Butt — from kathleendelaski.substack.com by Kathlee deLaski
An actual debate at LinkedIn Headquarters on the value of the college degree

Coincidentally published the same day, our books do take opposing views in parts. Scott is very strong on the value of the traditional degree, but wants to help students see beyond “the major,” telling them not to rely on what essentially amounts to a minority of the courses you’ll take in college to find your path to passion and employment. He urges them to find a “field of study,” to build in a broader range of self-directed experiences and classes to find your purpose and profession. (I love this and recommend his book.)

WhiIe I don’t recommend against college, “Who Needs College Anymore?” points to the growing number of employers and colleges that are offering skills-based career paths, that can be achieved with shorter term programs…that can be a stepladder to a degree, but don’t leave you credential-less if life gets in the way. I call on colleges to embrace all the market share they are leaving on the table, the 60% plus of Americans who are not getting a four-year degree, and to consider providing more modular professional pathway opportunities in addition to degrees.


Also re: higher education, see:

After years of quietly falling, college tuition is on the rise again — from hechingerreport.org
As colleges also pare back services, many students are paying more and getting less

Students nationwide are facing increases in tuition this fall of as high as 10 percent, along with new fees and rising costs for dorms and dining. And as in Pennsylvania, it’s an abrupt change from a period during which something happened that most Americans probably didn’t notice: Tuition had actually been falling, when adjusted for inflation, after decades of outpacing the cost of almost everything else.

That’s among the conclusions of The Hechinger Report’s update of its Tuition Tracker tool, which shows what students pay to go to individual colleges and universities based on their families’ incomes.

Considering the growing skepticism that college is “the magic ticket to the American dream,” said Gillen, raising tuition, for many higher education institutions, “definitely has the potential to be penny-wise and pound-foolish.” 

But universities and colleges are confronting unprecedented problems on the funding side. 

 

The State of AI Report 2025 — from nathanbenaich.substack.com by Nathan Benaich

In short, it’s been a monumental 12 months for AI. Our eighth annual report is the most comprehensive it’s ever been, covering what you need to know about research, industry, politics, and safety – along with our first State of AI Usage Survey of 1,200 practitioners.

stateof.ai
.

 


 

10 Tips from Smart Teaching Stronger Learning — from Pooja K. Agarwal, Ph.D.

Per Dr. Pooja Agarwal:

Combining two strategies—spacing and retrieval practice—is key to success in learning, says Shana Carpenter.


On a somewhat related note (i.e., for Instructional Designers, teachers, faculty members, T&L staff members), also see:

 

“A new L&D operating system for the AI Era?” [Hardman] + other items re: AI in our learning ecosystems

From 70/20/10 to 90/10 — from drphilippahardman.substack.com by Dr Philippa Hardman
A new L&D operating system for the AI Era?

This week I want to share a hypothesis I’m increasingly convinced of: that we are entering an age of the 90/10 model of L&D.

90/10 is a model where roughly 90% of “training” is delivered by AI coaches as daily performance support, and 10% of training is dedicated to developing complex and critical skills via high-touch, human-led learning experiences.

Proponents of 90/10 argue that the model isn’t about learning less, but about learning smarter by defining all jobs to be done as one of the following:

  • Delegate (the dead skills): Tasks that can be offloaded to AI.
  • Co-Create (the 90%): Tasks which well-defined AI agents can augment and help humans to perform optimally.
  • Facilitate (the 10%): Tasks which require high-touch, human-led learning to develop.

So if AI at work is now both real and material, the natural question for L&D is: how do we design for it? The short answer is to stop treating learning as an event and start treating it as a system.



My daughter’s generation expects to learn with AI, not pretend it doesn’t exist, because they know employers expect AI fluency and because AI will be ever-present in their adult lives.

— Jenny Maxell

The above quote was taken from this posting.


Unlocking Young Minds: How Gamified AI Learning Tools Inspire Fun, Personalized, and Powerful Education for Children in 2025 — from techgenyz.com by Sreyashi Bhattacharya

Table of Contents

Highlight

  • Gamified AI Learning Tools personalize education by adapting the difficulty and content to each child’s pace, fostering confidence and mastery.
  • Engaging & Fun: Gamified elements like quests, badges, and stories keep children motivated and enthusiastic.
  • Safe & Inclusive: Attention to equity, privacy, and cultural context ensures responsible and accessible learning.

How to test GenAI’s impact on learning — from timeshighereducation.com by Thibault Schrepel
Rather than speculate on GenAI’s promise or peril, Thibault Schrepel suggests simple teaching experiments to uncover its actual effects

Generative AI in higher education is a source of both fear and hype. Some predict the end of memory, others a revolution in personalised learning. My two-year classroom experiment points to a more modest reality: Artificial intelligence (AI) changes some skills, leaves others untouched and forces us to rethink the balance.

This indicates that the way forward is to test, not speculate. My results may not match yours, and that is precisely the point. Here are simple activities any teacher can use to see what AI really does in their own classroom.

4. Turn AI into a Socratic partner
Instead of being the sole interrogator, let AI play the role of tutor, client or judge. Have students use AI to question them, simulate cross-examination or push back on weak arguments. New “study modes” now built into several foundation models make this kind of tutoring easy to set up. Professors with more technical skills can go further, design their own GPTs or fine-tuned models trained on course content and let students interact directly with them. The point is the practice it creates. Students learn that questioning a machine is part of learning to think like a professional.


Assessment tasks that support human skills — from timeshighereducation.com by Amir Ghapanchi and Afrooz Purarjomandlangrudi
Assignments that focus on exploration, analysis and authenticity offer a road map for university assessment that incorporates AI while retaining its rigour and human elements

Rethinking traditional formats

1. From essay to exploration 
When ChatGPT can generate competent academic essays in seconds, the traditional format’s dominance looks less secure as an assessment task. The future lies in moving from essays as knowledge reproduction to assessments that emphasise exploration and curation. Instead of asking students to write about a topic, challenge them to use artificial intelligence to explore multiple perspectives, compare outputs and critically evaluate what emerges.

Example: A management student asks an AI tool to generate several risk plans, then critiques the AI’s assumptions and identifies missing risks.


What your students are thinking about artificial intelligence — from timeshighereducation.com by Florencia Moore and Agostina Arbia
GenAI has been quickly adopted by students, but the consequences of using it as a shortcut could be grave. A study into how students think about and use GenAI offers insights into how teaching might adapt

However, when asked how AI negatively impacts their academic development, 29 per cent noted a “weakening or deterioration of intellectual abilities due to AI overuse”. The main concern cited was the loss of “mental exercise” and soft skills such as writing, creativity and reasoning.

The boundary between the human and the artificial does not seem so easy to draw, but as the poet Antonio Machado once said: “Traveller, there is no path; the path is made by walking.”


Jelly Beans for Grapes: How AI Can Erode Students’ Creativity — from edsurge.com by Thomas David Moore

There is nothing new about students trying to get one over on their teachers — there are probably cuneiform tablets about it — but when students use AI to generate what Shannon Vallor, philosopher of technology at the University of Edinburgh, calls a “truth-shaped word collage,” they are not only gaslighting the people trying to teach them, they are gaslighting themselves. In the words of Tulane professor Stan Oklobdzija, asking a computer to write an essay for you is the equivalent of “going to the gym and having robots lift the weights for you.”


Deloitte will make Claude available to 470,000 people across its global network — from anthropic.com

As part of the collaboration, Deloitte will establish a Claude Center of Excellence with trained specialists who will develop implementation frameworks, share leading practices across deployments, and provide ongoing technical support to create the systems needed to move AI pilots to production at scale. The collaboration represents Anthropic’s largest enterprise AI deployment to date, available to more than 470,000 Deloitte people.

Deloitte and Anthropic are co-creating a formal certification program to train and certify 15,000 of its professionals on Claude. These practitioners will help support Claude implementations across Deloitte’s network and Deloitte’s internal AI transformation efforts.


How AI Agents are finally delivering on the promise of Everboarding: driving retention when it counts most — from premierconstructionnews.com

Everboarding flips this model. Rather than ending after orientation, everboarding provides ongoing, role-specific training and support throughout the employee journey. It adapts to evolving responsibilities, reinforces standards, and helps workers grow into new roles. For high-turnover, high-pressure environments like retail, it’s a practical solution to a persistent challenge.

AI agents will be instrumental in the success of everboarding initiatives; they can provide a much more tailored training and development process for each individual employee, keeping track of which training modules may need to be completed, or where staff members need or want to develop further. This personalisation helps staff to feel not only more satisfied with their current role, but also guides them on the right path to progress in their individual careers.

Digital frontline apps are also ideal for everboarding. They offer bite-sized training that staff can complete anytime, whether during quiet moments on shift or in real time on the job, all accessible from their mobile devices.


TeachLM: insights from a new LLM fine-tuned for teaching & learning — from drphilippahardman.substack.com by Dr Philippa Hardman
Six key takeaways, including what the research tells us about how well AI performs as an instructional designer

As I and many others have pointed out in recent months, LLMs are great assistants but very ineffective teachers. Despite the rise of “educational LLMs” with specialised modes (e.g. Anthropic’s Learning Mode, OpenAI’s Study Mode, Google’s Guided Learning) AI typically eliminates the productive struggle, open exploration and natural dialogue that are fundamental to learning.

This week, Polygence, in collaboration with Stanford University researcher Prof Dora Demszky. published a first-of-its-kind research on a new model — TeachLM — built to address this gap.

In this week’s blog post, I deep dive what the research found and share the six key findings — including reflections on how well TeachLM performs on instructional design.


The Dangers of using AI to Grade — from marcwatkins.substack.com by Marc Watkins
Nobody Learns, Nobody Gains

AI as an assessment tool represents an existential threat to education because no matter how you try and establish guardrails or best practices around how it is employed, using the technology in place of an educator ultimately cedes human judgment to a machine-based process. It also devalues the entire enterprise of education and creates a situation where the only way universities can add value to education is by further eliminating costly human labor.

For me, the purpose of higher education is about human development, critical thinking, and the transformative experience of having your ideas taken seriously by another human being. That’s not something we should be in a rush to outsource to a machine.

 

Sam Altman kicks off DevDay 2025 with a keynote to explore ideas that will challenge how you think about building. Join us for announcements, live demos, and a vision of how developers are reshaping the future with AI.

Commentary from The Rundown AI:

Why it matters: OpenAI is turning ChatGPT into a do-it-all platform that might eventually act like a browser in itself, with users simply calling on the website/app they need and interacting directly within a conversation instead of navigating manually. The AgentKit will also compete and disrupt competitors like Zapier, n8n, Lindy, and others.


AMD and OpenAI announce strategic partnership to deploy 6 gigawatts of AMD GPUs — from openai.com

  • OpenAI to deploy 6 gigawatts of AMD GPUs based on a multi-year, multi-generation agreement
  • Initial 1 gigawatt OpenAI deployment of AMD Instinct™ MI450 Series GPUs starting in 2H 2026

Thoughts from OpenAI DevDay — from bensbites.com by Ben Tossell
When everyone becomes a developer

The event itself was phenomenal, great organisation. In terms of releases, there were two big themes:

  1. Add your apps to ChatGPT
  2. Add ChatGPT to your apps

Everything OpenAI announced at DevDay 2025 — from theaivalley.com by Barsee
PLUS: OpenAI has signed $1T in compute deals

Today’s climb through the Valley reveals:

  • Everything OpenAI announced at DevDay 2025
  • OpenAI has signed $1T in compute deals
  • Plus trending AI tools, posts, and resources

Also relevant/see:



 

 

Making Retrieval Practice a Classroom Routine — from edutopia.org
By regularly working in activities that get students to recall content they’ve learned in the past and apply it, teachers can ensure deeper understanding.

Also see:

  • The Teaching Tips section out at RetrievalPractice.org
    Retrieval practice is a simple research-based teaching strategy that dramatically raises students’ grades. When students retrieve and bring information to mind, this mental challenge produces durable long-term learning. Easy learning leads to easy forgetting. Stop cramming, reviewing, and re-teaching. Instead, simply ask students what they remember. No prep, no grading, just powerful teaching. The science of learning exists. It’s time to unleash it.
 
© 2025 | Daniel Christian