OpenAI rolls out Memory feature for ChatGPT
OpenAI has introduced a cool update for ChatGPT (rolling out to paid and free users – but not in the EU or Korea), enabling the AI to remember user-specific details across sessions. This memory feature enhances personalization and efficiency, making your interactions with ChatGPT more relevant and engaging.
.
Key Features
Automatic Memory Tracking
ChatGPT now automatically records information from your interactions such as preferences, interests, and plans. This allows the AI to refine its responses over time, making each conversation increasingly tailored to you.
Enhanced Personalization
The more you interact with ChatGPT, the better it understands your needs and adapts its responses accordingly. This personalization improves the relevance and efficiency of your interactions, whether you’re asking for daily tasks or discussing complex topics.
Memory Management Options
You have full control over this feature. You can view what information is stored, toggle the memory on or off, and delete specific data or all memory entries, ensuring your privacy and preferences are respected.
Memory is now available to all ChatGPT Plus users. Using Memory is easy: just start a new chat and tell ChatGPT anything you’d like it to remember.
Memory can be turned on or off in settings and is not currently available in Europe or Korea. Team, Enterprise, and GPTs to come. pic.twitter.com/mlt9vyYeMK
From DSC: The ability of AI-based applications to remember things about us will have major and positive ramifications for us when we think about learning-related applications of AI.
“Knowing is not enough; we must apply. Willing is not enough; we must do.”
Johann Wolfgang von Goethe
Learning Transfer’s ultimate outcome is behaviour change, so we must understand the conditions that trigger a behaviour to start.
…
According to Fogg, three specific elements must converge at the same moment for a specific behaviour to occur. Given that learning transfer is only successful when the learner starts behaving in the desired new ways, Fogg’s work is critical to understanding how to generate these new behaviours. The Fogg Behavioural Model [*1] states that B=MAP. That is, a specific behaviour will occur if at the same moment there is sufficient motivation, sufficient ability and sufficient prompt. If the behaviour does not occur, at least one of these three elements is missing or below the threshold required.
The prompt is, in effect, a call to action to do a specific behaviour. The prompt must be ‘loud’ enough for the target person to perceive it and be consciously aware of it. Once aware of a prompt, the target immediately, and largely unconsciously, assesses their ability to carry out the requested behaviour: how difficult would this be, how long will it take, who can help me, and so on. They base this on their perception of the difficulty of the requested behaviour, and their ability, as they see it, to achieve that behaviour.
Executive function is a set of mental skills that include working memory, flexible thinking, and self-control. We use these skills every day to learn, work, and manage daily life. Trouble with executive function can make it hard to focus, follow directions, and handle emotions, among other things.
Snapshot: What executive function is
Some people describe executive function as “the management system of the brain.” That’s because the skills involved let us set goals, plan, and get things done. When people struggle with executive function, it impacts them at home, in school, and in life.
There are three main areas of executive function. They are…
Using Drawing as a Powerful Learning Tool — from edutopia.org by Selim Tlili When students draw something they’re learning about, they’re more likely to remember key details.
One of my main goals as a science teacher is to open students up to seeing all of those beautiful and interesting details. I do that by having students draw things and clearly write what they observe. Drawing something requires students to look at their subject far longer than they are accustomed. Writing what they see forces them to consciously acknowledge it. I explain to students that just as every single human is unique, so is every coin, plant, and salt crystal.
For the first time, a physical neural network has successfully been shown to learn and remember “on the fly,” in a way inspired by and similar to how the brain’s neurons work.
The Rundown: University of Sydney researchers have created a “brain-like” nanowire network capable of learning and remembering in real-time, similar to that of human brain function.
The details:
The nanowire neural network self-organizes into patterns, functioning like the brain’s synapses by responding to electrical currents.
What we teachers desperately need, though, is an ocean of examples and training. We need to see and share examples of generative AI—any type of artificial intelligence that can be used to create new text, images, video, audio, code, or data—being used across the curriculum. We need catalogs of new lesson plans and new curriculum.
And we need training on theoretical and practical levels: training to understand what artificial intelligence actually is and where it stands in the development timeline and training about how to integrate it into our classes.
So, my advice to teachers is to use any and all the generative AI you can get your hands on. Then experience—for yourself—verification of the information. Track it back to the source because in doing so, you’ll land on the adjustments you need to make in your classes next year.
From DSC: Interesting.
Learners can now seamlessly transition between AI-powered assistance (AI Tutor) and Live Expert support to get access to instant support, whether through AI-guided learning or real-time interactions with a human expert.
ASSIGNMENT MAKEOVERS IN THE AI AGE WITH DEREK BRUFF — from teachinginhighered.com by Bonni Stachowiak Derek Bruff shares about assignment makeovers in the AI age on episode 481 of the Teaching in Higher Ed podcast
Comment on this per Derek Bruff:
Why not ask ChatGPT to write what King or X would say about a current debate and then have the students critique the ChatGPT output? That would meet the same learning goals while also teaching AI literacy.
(Be sure to read Asim’s contribution for a useful take.)
Here’s a closer look at the concurrent AI landscape in schools — and a prediction of what the future holds.
So far, high-profile ventures in the instruction realm, such asKyron Learning, have fused teacher-produced, recorded content with LLM-powered conversational UX. The micro-learning tool Nolej references internet material when generating tasks and tests, but always holds the language model closely to the ground truth provided by teachers. Both are intriguing takes on re-imagining how to deliver core instruction and avoid hallucinations (generated content that is nonsensical).
As a result, real-time 3D jobs are among the most in demand within the tech industry. According to Unity’s vice president of Education and Social Impact, Jessica Lindl, demand is 50% higher than traditional IT jobs—adding that salaries for real-time 3D jobs are 60% greater.
“We want to provide really simple on ramps and pathways that will lead you into entry level jobs so that at any point in your career, you can decide to transfer into the industry,” Lindl says.
University World News continues its exploration of generative AI in our new special report on ‘AI and Higher Education’. In commentaries and features, academics and our journalists around the world investigate issues and developments around AI that are impacting on universities. Generative AI tools are challenging and changing higher education systems and institutions — how they are run as well as ways of teaching and learning and conducting research.
My advice for you today is this: fill your LinkedIn-feed and/or inbox with ideas, inspirational writing and commentary on AI.
This will get you up to speed quickly and is a great way to stay informed on the newest movements you need to be aware of.
My personal recommendation for you is to check out these bright people who are all very active on LinkedIn and/or have a newsletter worth paying attention to.
I have kept the list fairly short – only 15 people – in order to make it as easy as possible for you to begin exploring.
It is crucial to recognize that the intrinsic value of higher education isn’t purely in its ability to adapt to market fluctuations or technological innovations. Its core strength lies in promoting critical thinking, nurturing creativity, and instilling a sense of purpose and belonging. As AI progresses, these traits will likely become even more crucial. The question then becomes if higher education institutions as we know them today are the ony ones, or indeed the best ones, equipped to convey those core strengths to students.
Higher education clearly finds itself caught in a whirlwind of transformation, both in its essence and execution. The juxtaposition of legacy structures and the evolving technological landscape paints a complex picture.
For institutional leaders, the dual challenge lies in proactively seeking and initiating change (not merely adapting to it) without losing sight of their foundational principles. Simultaneously, they must equip students with skills and perspectives that AI cannot replicate.
“They begged, bargained with, and berated their instructor in pursuit of better grades — not “because they like points,” but rather, “because the education system has told them that these points are the currency with which they can buy a successful future.””
Habit #2: Engage students in a brain dump or two things as an entry ticket or exit ticket. Spend one minute or less having students write down everything (or just two things) they remember from class. The key: Don’t grade it! Keep retrieval practice no-stakes to emphasize it’s a learning strategy, not an assessment strategy.
Teaching from the heart in 13 steps — from timeshighereducation.com by Beiting He Engaging your students through empathy requires teachers to share their own stories and vulnerabilities and foster a safe space for learning. Here, Beiting He offers 13 ways to create a caring classroom
In summary, “I wish” is about proposing positive changes and improvements, while “I wonder” is about asking thoughtful questions to gain insight and foster meaningful conversations within the team.
If we want students to remember – to lock new information or ideas into long-term memory – getting meaningful repetitions still is key. And the science of learning still backs that up.
So … if we want students to get repetitions to make new learning permanent, how can they do it? Here are 10 ways to help students get repetitions for practice – and how classroom technology can help.
In this episode, I share ten engaging activities that combine education, technology, and plenty of fun to make the first week of class super memorable. From digital scavenger hunts to virtual field trips, hear about a few of my favorite ways to create an interactive start to your school year.
Tips for First Week of School Activity Ideas
Establish routines in a fun way.
Provide opportunities for collaboration.
Introduce tech tools that will be used all year.
From DSC: Dr. Burns has a great list of tools/tips/resources in this posting.
Four directions for assessment redesign in the age of generative AI— from timeshighereducation.com by Julia Chen The rise of generative AI has led universities to rethink how learning is quantified. Julia Chen offers four options for assessment redesign that can be applied across disciplines
Direction 1: From written description to multimodal explanation and application
Direction 2: From literature review alone to referencing lectures
Direction 3: From presentation of ideas to defence of views
Direction 4: From working alone to student-staff partnership
With the latest national test results showing a dispiriting lack of progress in catching students up academically in the wake of the pandemic, one potential explanation stands out: stubbornly high rates of student absenteeism. Vast numbers of students haven’t returned to class regularly since schools reopened.
Cognitive Load Theory is an influential theory from educational psychology that describes how various factors affect our ability to use our working memory resources. We’ve done a digest about cognitive load theory here and talked about it here and here, but haven’t provided an overview of the theory so I want to give an overview here.
…
Cognitive load theory provides useful and dynamic model for how many different factors affect working memory and learning. Hopefully this post provides a useful overview of some of the main components of cognitive load!
Simplify the explanations of what you’re presenting as much as possible and break down complex tasks into smaller parts
Don’t place a large amount of text on a slide and then talk about it at the same time — doing so requires much more processing than most people can deal with.
Consider creating two versions of your PowerPoint files:
A text-light version that can be used for presenting that content to students
A text-heavy version — which can be posted to your LMS for the learners to go through at their own pace — and without trying to process so much information (voice and text, for example) at one time.
Design-wise:
Don’t use decorative graphics — everything on a slide should be there for a reason
Don’t use too many fonts or colors — this can be distracting
Don’t use background music when you are trying to explain something
Why it matters: The best AI assistants will be the ones that require the least prompting. They’ll get to know who you are, what you need, and your modus operandi. Profiles are a good starting point, but we believe the game-changer will be larger context windows (that’s nerd-speak for the amount of context ChatGPT can handle). .
From DSC: And how about taking this a step further and remembering — or being able to access — our constantly updated Cloud-Based Learning Profiles?
In particular, TutorND’s emphasis on applying principles of cognitive science – the science of our how minds work – in tutoring practice has not only bolstered the interest and confidence of some of its tutors to pursue teaching, but also strengthened their instructional skills and meaningfully contributed to PK-12 student growth.
Today, TutorND trains and supports 175 tutors in schools across the greater South Bend community and across the country. Given that these tutors are students, faculty, and staff interested in cognitive science research and its application to student learning, they’re able to bridge theory and practice, assess the effectiveness of instructional moves, and foster learning experiences for students that are rigorous, affirming, and equitable.
A recent analysis by researchers at NYU, Princeton and the Wharton School finds that many of the jobs that will be most “exposed” to generative AI such as ChatGPT are in the college teaching profession.
One of the first narratives to emerge from the sudden explosion in usage of ChatGPT is the risk of students cheating on writing assignments.
But use by college teachers is growing quickly too, and adoption by educators may be critical to making the case that AI will augment the jobs humans are doing rather than replace them.
Each SmartCard features a special marker that, when scanned with a tablet, unlocks informative virtual content students can interact with using basic hand gestures and buttons. According to its developers, Justin Nappi and Sudiksha Mallick, SmartCards can be especially useful for neurodivergent students, including those with attention-deficit/hyperactivity disorder (ADHD), autism, or dyslexia.
Last week, as I was interviewing Shaun Vecera about a new initiative he directs at the University of Iowa, he made a comment that stopped me in my tracks. The initiative, Learning at Iowa, is meant to create a common vocabulary, based on cognitive science, to support learning across the university. It focuses on “the three M’s for effective learning”: mind-set, metacognition, and memory.
“Not because those are the wrong ways of talking about that. But when you talk about learning, I think you can easily see how these skills transfer across not just courses, but also transfer from the university into a career.”
From DSC: This reminds me of what I was trying to get at here — i.e., let’s provide folks with more information on learning how to learn.
When the researchers asked their interview subjects how they first learned about peer instruction, many more cited informal discussions with colleagues than cited more formal channels like workshops. Even fewer pointed to a book or an article.
So even when there’s a really well-developed recipe, professors aren’t necessarily reading it.
…
In higher ed, teaching is often seen as something anyone who knows the content can automatically do. But the evidence suggests instead that teaching is an intellectual exercise that adds to subject-matter expertise.
This teaching-specific math knowledge, the researchers note, could be acquired in teacher preparation or professional development, however, it’s usually created on the job.
“Now, I’m much more apt to help them develop a deeper understanding of how people learn from a neuroscientific and cognitive-psychology perspective, and have them develop a model for how students learn.”
Erika Offerdahl, associate vp and director of the Transformational Change Initiative at WSU
From DSC: I love this part too:
There’s a role here, too, for education researchers. Not every evidence-based teaching practice has been broken into its critical components in the literature,
