John Murray
@Murray_Je
Followers
2K
Following
5K
Media
434
Statuses
6K
#HigherEd Learning Experience Designer, MA EdTech (Concordia) | Drafting Comm member for 1st crowdsourced @UN strategy | Tweet on many facets of #onlinelearning
Montréal, Québec
Joined January 2012
This study shows: many top LLMs routinely overgeneralise scientific results (e.g. omitting details limiting the scope), even when asked for accuracy. Newer LLMs often performed worse. This bias risks widespread misinterpretation of research, highlighting the need for mitigation.
0
2
0
As advanced as AI seems, there's something anti-intellectual about it. It treats thinking as an obstacle on the path to "the answer." But for the intellectual, thinking isn't a means to an end. It's the thing itself. No wonder so many thoughtful people hate it.
273
689
4K
"Intention–behaviour gap in self-regulated learning…is often treated as motivational or volitional failure. https://t.co/Jtz3WlTvkf shows even well-informed students may default to ineffective strategies because they are automatically cued & resistant to change" via @C_Hendrick
link.springer.com
Educational Psychology Review - Students frequently rely on ineffective learning strategies instead of those that promote long-term retention. This is not simply a matter of lacking metacognitive...
0
0
0
Neuro-myths such as Learning Styles are the ultimate litmus test of a profession and it's one that education keeps failing again and again.
@C_Hendrick My university's Academic Resource Center offers learning style assessment to our students. I have made literally dozens of attempts to get them to stop, to no avail.
8
29
136
"Generative AI ...is everything that higher education should not be: exploitative, non-consensual, inaccurate, corrosive of cognitive growth and human relationships, destructive of people and the environment, and effectively unregulated." Blog link in replies.
1
2
2
A Student’s Right to Refuse Generative AI «I came to realize what I valued most in my education: discussion-rich classes, reciprocity with my peers and professors, and to be challenged and work through those challenges with the guidance of my professors.»
refusal.blog
Elizabeth Palumbo, Syracuse University Photo by cottonbro studio on Pexels.com I’m sitting at a desk in a college classroom in Upstate New York. It’s nearing the end of the fall semester of my seni…
0
2
18
We are delighted to announce the launch of the AI Accountability Lab. Led by Dr @Abebab and housed in the @AdaptCentre in @tcddublinscss, the AIAL will play a leading role in advancing AI accountability research. Read more at: https://t.co/H8GL7Bi1ML
8
29
171
So many presentations withn"of course Gen AI has ethical issues" and then they go on to use it anyway and recommend that students use it , because apparently the ethical concerns are resolved by simply saying 'this has ethical issues' once before starting to use it.
0
0
1
New review says ineffective ‘learning styles’ theory persists in education around the world
0
0
1
"In a survey across the University of Michigan’s three campuses..., respondents in all roles [faculty, students, & staff] described more concerns than potential uses with generative #AI and said they wanted more information and training": https://t.co/fSuqHqFj6m
#ethics #tech
bridgemi.com
In September, the school gave personal AI assistants to the more than 50,000 students at the University of Michigan.
0
2
0
I am going through my angry phase of AI (it comes in waves, but I'm at a particularly deep phase of anger right now). I can't stand how much people diminish the reproduction of bias and occasional hallucinations. Why? Why does this not matter to some people?
7
5
46
Typical Silicon Valley (pseudo-) logic: let's just pretend this tech is magic, and that its costs don't matter. Yes, there are many meaningful climate-positive applications of AI, but we urgently need to address the environmental impacts of (superfluous) generative AI.
Former Google CEO Eric Schmidt says energy demand for AI is infinite and we are never going to meet our climate goals anyway, so we may as well bet on building AI to solve the problem
25
59
321
"When incorrect answers were flagged, getting better at giving correct answers was one way to optimize things. [But] getting better at hiding incompetence worked just as well. Human supervisors simply didn’t flag wrong answers that appeared good and coherent enough to them." #AI
The more sophisticated AI models get, the more likely they are to lie
0
8
6
"It seems rash…to promise 2 sigma effects from AI when human tutoring has rarely produced such large effects…. Over-promising can lead to disappointment, & reaching for impossible goals can breed questionable educational practices" https://t.co/gLwqTa8OIL
#AI #Chatbot #HigherEd
educationnext.org
An experimental intervention in the 1980s raised certain test scores by two standard deviations. It wasn’t just tutoring, and it’s never been replicated, but it continues to inspire.
0
0
0
Stimulating, holistic keynote by @jenni26cg at @LHSymposiumAI that provided overview of what currently guides conversations about #ethics and #AI. We are currently narrowly focused on IT concerns and not on the broader questions related to #inclusion. #HigherEd
0
0
0
Important reminder: "The training that we do to make LLMs better conversationalists often misrepresents their ability." -paraphrased quote from Dr. Michael Townsen Hicks, @LHSymposiumAI #glasgowAIethics #GenAI #HigherEd
0
1
3
“[Don’t] assume LLMs will continually improve—the better approach is to engage w/ this technology as it exists now 2 explore its strengths & limitations. Engaging w/ AI does not necessarily mean integrating it into instructional practice.” https://t.co/0KbeMlgoJK
#GenAI #HigherEd
0
0
0
How can cognitive science and good "old" pedagogy help us crack the challenge of integrating Generative AI in (higher) education? My attempt to answer this question, inspired by many discussions and questions... https://t.co/J1OYEoNw3z
1
0
1
Hoping to see this gif cropping up in every AI seminar on assessment from now on.
Leaving aside everything else, it's a pretty dumb use of time.
0
2
2
"The most essential ingredient is to be transparent & talk to students…[they] may share concerns about #academicintegrity & they want to know how the technology will impact career readiness. They aren’t chatbots…. They’re there to learn" https://t.co/zMGgjKNJ5Y
#GenAI #HigherEd
everylearnereverywhere.org
Instructors must explain their requirements for student use of AI and discuss how they plan to use AI in the classroom.
0
0
0