Sunday, September 29, 2024

Realistic dialogue is here....

Advanced Voice Mode on the OpenAI app is now avaialble in the UK. I’m getting pretty used to advanced releases and was hugely impressed by ChatGPT o1, with its open thinking and reasoning capabilities. Then got pretty excited by NotebookLM with its automatic podcast feature and other functionality. But this was a user experience like no other.

Fluid

Once again, I feel as though I’m being swept up and away, as AI trailblazes forward at a blistering speed. Once again I’d emphasise that you really do have to ‘try it to get it’. An interesting paper just published showed that much of the scepticism on GenAI comes from those who have not tried it. Lesson - for any real critical thinking or analysis, some use is necessary.

Immediate impression – conversations are just so much more fluid, natural and therefore real. This is Alexa and Siri on steroids. Much better on interruptions, which matters, as as voice dialogue can be a little odd when there’s clumsy turn taking. Here, you feel you are in control and not getting dragged along by the ‘machine’. It feels more like talking to a human. Nass & Reeves were right, the more you eliminate non-human cues, such as odd pauses, clumsy interruptions and misfires, the more you are fooled into thinking the machine is a person. I keep saying this but interfaces matter in tech.

It copes superbly well with pauses or interruptions and pick up where you left off with ease. It also recognises when you are finished speaking. No need to tap button when you stop – that was badly needed. Also better on filtering out background noise.

Accents 

It also picks up on conversational cues better – recognises intonation, speed of speech, even if it is a question. A variety of voices are available – different genders, accents, British American or Scottish, different speaking styles are also available. Choosing a voice just makes things feel better. I can express myself more freely as it has no problem with my strong Scottish accent. I tossed an entire caber at it on this one. Yet it coped with my odd vernacular and very fast speech. Believe me you wouldn’t have understood what I said  - it did. Asking it to speak back to me in a Scottish accent was fun.

Memory

As it remembers coversations, subsequent conversations are more meaningful over time. It keeps track of preferences, interests or details about you to make the chats more relevant. It then took me by surprise. More of that later. Memory matters and if I had a disability or specific preferences it will remember and tailor the conversion to my needs. This is a boon for accessibility and inclusion.

Accessibility

On that point, I’ve been doing some work in DEI and accessibility in AI, where people have truly underestimated, not only hoiw much has already been achieved but also its potetial to give agency to people, who find education and the workplace difficult. This is of immediate use for the blind, dyslexics and to be honest – anyone. We all have learning difficulties as learning is difficult. This gives access and agancy to so many who find education and learning difficult.

Learning

Tried it for learning a second language – astonishing on pronunciation, like talking to a good teacher. We ar egtting ever closer to this being a tutor, trainer, coach, mentor, counsellor and so on. I think we can see how any training that requires role playing – managerial training, sales training, customer care, dealing with patients and so on. You can prompt it to play roles – that’s very cool. When the dialogue and pedagogic paths are tighter it will be a UNIVERSAL teacher, on demand, anytime, anyplace on anything, at any level, for anyone in any language. 

What’s coming next?

My guess is that future enhancements are likely to be even better at understanding and emotion detection, as emotional cues allow more nuanced reactions. More customisation is inevitable. But what we’ll also see is is more contextual awareness, to improve relevance and make converstions more intuitive. Integratoipn with other applications and devices will come in time.Eventually we will be speaking fluently to avatars, images of people htat look like real people. This is inevitable. The integration of this realistic dialogue into robots also seems certain.

Conclusion

Honestly. This was amazing. Maybe the first time I’ve ever felt that I was speaking to another human but actually a machine. This is going places – whether it is help with learning, getting things done in the workplace or getting some help if you’re feeling down. 

One moment stood out. She dropped the fact that I had a dog into the converstion, not just that, she knew it was a Schnauzer. I love my dog. He was snuggling up to me at that very moment on the sofa. I stoked his back, he looked over his shoulder at me. All three of us had a moment.

 

Saturday, September 28, 2024

Is SaaS in learning dead?

Our LMS seem antiquated, a bare learning landscape of frustration, overloaded with outdated content a stockpile of cartoonish modules and tedious slideshows dotted with MCQs.

The market seems weary of simply renewing licences for software that has low use figures, often just slabbing out compliance content with no actual measurement of its productivity or impact. Measuring bums, even on virtual seats is to measure the wrong end of the learner. 

These systems get locked in as a matter of habit not impact. But dreams don’t last forever and pricing on LMS software, content libraries and other learning services has got fiercely competitive. Many are seeing revenues plateau or fall. Management is being swapped out, share-prices falling, revenues no longer predictable, profits still showing overhang on 3 year licence fees. It has all become fiercely competitive.

AI is disruptor

Having been involved in a lot of AI projects for over a decade, since early 2014, the assumption has always been that people need to build some sort of AI SaaS product, sit back and watch the money pour in. A recent Tweet questioning this, by the much admired John Rush, went viral on this point.

Yet AI is already bringing services directly to your users, be it learners or employees in any sector at any level. With ChatGPTo1 reason has now appeared. AI now gives massive agancy to the individual separate from your corporate systems. Like a supercar at saloon prices it has overtaken many services on the fast lane. SaaS services get locked into fixed depreciating code. These inflexible code sets and fixed products make progress impossible, whereas AI seems to release new astounding updates on a weekly basis, so quick, it is hard to keep up.

AI is not really a business change, it is a change in mindset. You have to think of of power being plugged directoy into actual learners and employees. It releases agancy and motivation. Once you realise that everyone is using it, you need to see that its direct impact bypassing your existing systems. 

AI on the SLY

The studies on productivity and impact are clear, that’s why most are using AI on the SLY. Individuals are increasing their productivity, making their lives easier, while organisations, locked into legacy systems are trying to ignore it or fighting it wit negativity. That’s because it needs a shift in mindset, to accept the probability game, understanding that everything has flaws, everyone is biased, especially people and processes. Sure AI may occasionally come up with the wrong answer but so do people and at least AI is fast! In fact, for most education, learning and business uses, hallucination is not a problem at all, as it has been pushed away out to the periphery. The appearance of reasoned ‘thinking’ has meant a whole load of new educational and organisational uses have been folded in to the omelette. And you don’t make an omelette without breaking eggs.

Productivity

Traditional Saas product in learning may be struggling but it also means that straight to user services in real time are alive and kicking ass.

One group of companies that are doing well are those that use AI to simply increase internal productivity. This gives them edge and the financial room to increase productivity or respond to pricing pressure. In one case we’ve seen astounding increases in productivity on process, both massive savings in time but also increases in quality. This gives you time to focus on customers and quality. How do you make your learning company more valuable? Get lean. Get agile. AI allows you build software fast. AI allows you to fast-track productivity increases. AI gives you quality increases fast. AI unleashes life in an organisation.

Performance support

Performance support, which is what 99% of AI is, is now in the hands of users. There’s tons of rooms for good delivery here, as organisations need to deliver performance support based on their own motherlode of content. In steps AI with RAG and other techniques to do precisely that - superbly well. Suddenly, you’re not relying on generic content, that everyone’s using, giving your organisation no edge at all. You can focus and lever what you already own. 

Conclusion

We’re still locked into the lumbering LMS, often a dark dungeon of disappointment, also over-engineered content, often out of date and hopelessly cliched and generic, with its cartoon libraries and slideshows peppered with MCQs. There is a sense that the old era is dying in the vine, coming to an end.

 

Sunday, September 22, 2024

Good survey (1045 parents & teens 13-18) showing high use of GenAI in schools (with short podcast)

Yet another good survey (1045 parents and teens 13-18) showing high use of GenAI in US schools. Now the predictable story of learners using AI on the SLY.

Use

First, we have to recognise that they’re using it and that their use is across a number of tools for different purposes. But homework and assignments are the most common.

  • 7/10 teens, 13 to 18 used at least one type of AI
  • Using AI for a variety of purposes
  • Help with homework most common 
  • 40% report using AI for school assignments 
  • 63% of AI users used chatbots or text generators
  • 57% used search engines with AI generated results

Permission

It would appear that about half have had no guidance, or don’t know of such guidance from schools. This is worrying and shows how institutions lagging in their professional duties, to create the right context for learning and behaviour. Parents, of course, are usually kept in the dark

  • Even split on using AI with/without teacher’s permission 
  • 6/10 teens say school has no rules/don’t know for AI use

Parents

Most parents are in dark about their child’s generative AI use. They say schools have not communicated about AI policies and only 37% of parents whose teen had used GenAI platform thought their child had used it.

Accuracy

With around half using other sources to check accuracy, this is to be expected from teens. However, as hallucination errors rapidly decline, and capabilities in maths and science increase, this is a diminishing problem. This is less of a problem with high school content, compared to say Universities.

  • 49% checked other sources to verify accuracy of AI  for assignments
  • 39% using AI detected problems and inaccuracies

Discussions

Teens who’ve had class discussions about AI more likely to have nuanced views about its usefulness and challenges. More often they say it has changed the skills, educational path, or job they plan to pursue. That’s fascinating. The Harvard review at College levels said exactly the same thing, that these experiences are shaping hteir view of their own futures, even resulting in rethinks on study paths and careers.

Conclusion

Use is normalised, with a range of tools for a range of purposes but still mostly for homework and assignments. Schools tardy in having policies and not much use within schools. Parents in the dark about all of this.

Here's an AI generated podcast about the report:



There’s lots of other interesting detail in the report. Remember this is a US study.
https://www.commonsensemedia.org/research/teen-and-young-adult-perspectives-on-generative-ai-patterns-of-use-excitements-and-concerns


Friday, September 20, 2024

Another great teaching and learning tool. NotebookLM's automated podcasts will blow your mind


NotebookLM shows how AI can be used right now in learning. The automatic podcasts will blow your mind.

This is a monumental achievement, as anyone can now take a motherlode of documents and get AI to extract summaries, answer questions, give a timeline, even create a podcast.

We have worked on note taking software adding AI features, and seen it work in practice. This really is an area that is ripe for application, as note taking is an established practice, the bridge between the teacher and learner, the active learning that allows the learner to move forward.

With NotebookLM, I uploaded a large book on learning theorists, at 300,000 words, saved as a 700+ page PDF. It covers 2500 years, with 300 learning theorists and has taken me years to write. Here's the list of Learning theorists and topics covered:


Chat

It gives a pretty good summary, briefing and timeline, century by century, across 2500 years. 

Then, I started to ask it questions, using the chjat function, engaging in dialogue with my own work. It is superb. If every teacher, department, course had its curated content in this context, this would be a boon for students. If you want to know some specifics about a particular learning theorists or more geneal view of a group, it does a credible job.

Learners can ask general questions, then, when engaging in a class or assignment, specific questions, getting carefully considered answers. I asked it some general questions about historical trends, why so few women in learning and so on. Pretty good…

What is impressive are the provenance numbers that link to the source text in hte document for most components in the answer.

Podcast

But the big hit comes at the end. I asked it to do an ‘Audio overview’ of all 300 learning theorists in 700 pages. This was way beyond my expectations. A dialogue podcast, at around 11.5 minutes, was amazingly realistic, with a male and female commentator bounding off each other. The language was informal and sounded like real dialogue. I have since asked people to listen – they ALL thought these were real people.

It covered the material well, pulling out relevant theorists and making insightful observations as they took us through 2500 years of learning theory. 

For novice learners who want a friendly introduction, those with poor literacy skills, English as a second language, but also dyslexic learners, this is a wondrous feature. I can see this being a great introduction to any topic, subject, course or assignment. 

Fronting a learning journey with an introductory podcasts seems like a fine idea to me. A brief and accessible introduction to a subject, super quick to produce and prepares the learners for deeper learning.

I can see this getting real traction with learners at school and college.

Take notes

You can also create and manage notes. You select the box in the upper-right corner of one or more notes, and NotebookLM displays text-buttons with actions to take on the notes. These include:

Summarise the selected notes.
Suggest related ideas.
Create a study guide.
Create an outline.
Combine the selected notes into a single note.

I am always astonished at how many learning professionals do not take notes at conference talks I give. It's almost endemic. Yet, notes, especially in school and college are the bedrock for subsequent learning. It is not the notes in themselves but what you do with the notes afterwards that matter. That's why the debate over handwriting v typing is irrelevant - the learning takes place largely in what follows the act of writing - the retrieval process, reflection, manipulation, subsequent analysis and expansion of those notes. Notes launch good learning journeys. This encourages a more structured approach to notes, with more tools and help.

Sharing notebooks

You can share notebooks by opening a notebook and clicking the Share icon in the top-right corner of the screen. You can grant either Viewer or Editor access to another user by adding their email address.

A viewer will have read-only access to all the source documents and notes you shared with them in the shared notebook. An editor will be able to view, add, or remove sources and notes in your shared notebook as well as share it further with other users. Personal gmail accounts can share a notebook with up to 50 other individual users

PS

If you want REAL podcasts and deep dives into specific topics in this this content, there are 36 1 hour podcasts in the 'Great Minds on Learning' series here…
https://greatmindsonlearning.libsyn.com/

Further experiment

Experimenting further, here's the NotebookLM AI-generate summary and 10 min podcast from four large articles I wrote on Heidegger, Foucault, Lyotard and Derrida. Just one small section from the huge 700 pager.



I loved how it created jokes, even analogies. Absolutely brilliant. Who said AI couldn't be innovative!

And here's the AI generated summary

Summary

This text provides an overview of the philosophical and educational theories of four prominent figures in the 20th century: Heidegger, Foucault, Lyotard, and Derrida. While exploring their individual contributions to areas such as ontology, power relations, and language, the text also criticizes their theories, particularly their tendencies towards relativism, cynicism, and obscurity, arguing that these thinkers often detached their ideas from the real world. The text highlights the influence these figures have had on critical pedagogy, postmodernism, and contemporary educational practices.

Now here's the real long-form podcast by myself and John Helmer...
https://greatmindsonlearning.libsyn.com/gmols6e36-continental-theorists-with-donald-clark




Thursday, September 19, 2024

GenAI students learn more than twice as much in less time, with more engagement motivation enjoyment & growth mindset

Evidence really mounting on GenAI tutors performing as well as or better than other teaching methods. In this Harvard study AI Tutoring outperformed even active learning (Kestin et al. 2024).

Recent breakthroughs in generative AI (GAI) are set to transform education as we know it. But we're still figuring out the best ways to use this new technology and how effective it really is. This study is particularly interesting, as it compares University students learning with an AI-powered tutor to those in a traditional active learning class. They made sure the AI tutor used the same teaching methods as the regular lectures. And guess what? 

Results 


Compared to those in the regular class, students using the AI tutor learned:

  • more than twice as much 
  • in less time 

If this continues, these findings show that AI tutoring could massively boost learning, making a strong case for adopting it widely in education.

The authors note that passive lectures are the least effective, being passive, going at the wrong pace for many learners, with no feedback and little engagement. Even active learning is never one-on-one and therefore suffers from the same problems as passive lectures. 

Motivation


Improved levels were found on other qualitative measures:

  • Engagement
  • Motivation
  • Enjoyment
  • Growth mindset

This matters, especially ‘motivation’, where I have seen significant impact in many contexts. This, I some ways is more important the younger the learner. 

Conclusion

Little, other than the flawed Bloom 2 Sigma paper, is being seen in presentations on efficacy. I have been presenting this and other papers, to show that we are quickly gathering evidence of GenAIs effectiveness in learning. The hypothesis, that GenAI is a powerful learning tool seems obvious when one looks at the astonishing use by so many in so many contexts, school, University and workplace. Academic publishing lags long behind reality. This is a welcome addition to the evidence.

Paper 
https://www.researchsquare.com/article/rs-4243877/v1


Monday, September 16, 2024

ChatGPT o1 brilliant 'worked example' teaching & learning tool


We have in ChatGPT o1 a brilliant teaching tool, because it shows its thinking. 
Yes I know, It is not 'thinking' as we do". But it is mightily impressive, even on oblique problems. Showing your 'working' is useful. You can do this for any question - past papers, teacher written and do so on. 

I tried it first on a few, very different logic, language and lateral thinking problems. It got them ALL right but what was more impressive was the exposure of its thinking. It not only got the right answers for all, it explained in detail its process of coming to the answer. This 'explaining' lies at the heart of good teaching and learning. 

Worked examples

The fact that it does 'explain' its thinking means it is a perfect 'worked example'. Effective worked examples are associated with early stages of knowledge and skill development. They unpack a problem and allow learners to see the solution in detail. 

By providing an expert's problem solution for a learner to study. Any bottlenecks in the steps can be unlocked. You can further query bits you don't fully understand. It worked well for all of of the short problems and a full GCSE school exam question from a past paper in English Literature, on Jekyll & Hyde, where it did a brilliant job at answering the question. More than this, it unpacked its thinking, game me a superb mnemonic and produced Flashcards for practice.

We have also gone on to turn this into a question-generated spaced practice tool.

Starter problems

Note that some are pure logic, some linguistic, some quite oblique.

  • Logical Reasoning Problems
  • Mathematical Puzzles
  • Word Riddles and Linguistic Puzzles
  • Lateral Thinking Problems

Q I have cities, but no houses. I have mountains, but no trees. I have water, but no fish. What am I?

Q What English word has three consecutive double letters?

Q A woman shoots her husband, then holds him underwater for five minutes. Next, she hangs him. Right after, they enjoy a lovely dinner. Explain.

Q I come from a mine and get surrounded by wood always. Everyone uses me. What am I?:

Q A girl has as many brothers as sisters, but each brother has only half as many brothers as sisters. How many brothers and sisters are there in the family?

Q What disappears as soon as you say its name?

Q How can the number four be half of five?

Q I have keys, but no locks and space, and no rooms. You can enter, but you can’t go outside. What am I?

Q  I have keys, but no locks?

Q What gets wet while drying?

Q A is the brother of B. B is the brother of C. C is the father of D. So how is D related to A

Q What are the next three letters in this combination? OTTFFSS

Q This belongs to you, but everyone else uses it.

Q How is seven different from the rest of the numbers between one and ten?

ChatGPT 01 got all of these rights, each one took only a few seconds. But what was impressive was the unpacking of its 'thinking'. That's where the learning takes place.

Let me give one example:

A is the brother of B. B is the brother of C. C is the father of D. So how is D related to A?

Thought for 13 seconds

Answer:

D is A's niece or nephew.


Explanation:

Let's break down the relationships step by step to understand how D is related to A.

  1. A is the brother of B:

    • This means A and B are siblings, and A is male.
  2. B is the brother of C:

    • B and C are siblings, and B is male.
    • Since B is the common sibling between A and C, all three (A, B, and C) are siblings.
    • Given that A and B are brothers, and B and C are siblings with B being male, it's reasonable to infer that C is also male (since B is referred to as brother to both A and C).
  3. C is the father of D:

    • C is male and is the parent of D.
    • This means D is the child of C.
  4. Determining the relationship between A and D:

    • A and C are brothers.
    • C is the father of D.
    • Therefore, A is the uncle of D.
    • Consequently, D is the niece or nephew of A.

Conclusion:

  • D is related to A as a niece or nephew.
  • A is D's uncle.

Summary:

  • A, B, and C are brothers.
  • C is the father of D.
  • Therefore, D is A's niece or nephew.

Riddle Solved!


I then took a standard question from a past exam paper in English Literature.


GCSE ENGLISH LITERATURE EXAM QUESTION


Robert Louis Stevenson: The Strange Case of Dr. Jekyll and Mr. Hyde      

Read the following extract from Chapter 1 (Story of the Door) of The Strange Case of Dr Jekyll and Mr Hyde and then answer the question that follows.  In this extract, Mr Enfield tells Mr Utterson about his encounter with Mr Hyde.    

“Well, it was this way,” returned Mr. Enfield:
“I was coming home from some place at the end of the world, about three o’clock of a black winter morning, and my way lay through a part of town where there was literally nothing to be seen but lamps.  Street after street, and all the folks asleep – street after street, all lighted up as if for a procession and all as empty as a church – till at last I got into that state of mind when a man listens and listens and begins to long for the sight of a policeman.  All at once I saw two figures: one a little man who was stumping along eastward at a good walk, and the other a girl of maybe eight or ten, who was running as hard as she was able down a cross street.  Well, sir, the two ran into one another naturally enough at the corner; and then came the horrible part of the thing; for the man trampled calmly over the child’s body and left her screaming on the ground.  It sounds nothing to hear, but it was hellish to see.  It wasn’t like a man; it was like some damned Juggernaut.  I gave a viewholloa, took to my heels, collared my gentleman, and brought him back to where there was already quite a group about the screaming child.  He was perfectly cool, and made no resistance, but gave me one look so ugly that it brought out the sweat on me like running.  The people who had turned out were the girl’s own family; and pretty soon, the doctor, for whom she had been sent, put in his appearance.  Well, the child was not much the worse, more frightened, according to the Sawbones; and there you might have supposed would be an end to it.  But there was one curious circumstance.  I had taken a loathing to my gentleman at first sight.  So had the child’s family, which was only natural.  But the doctor’s case was what struck me.  He was the usual cut-and-dry apothecary, of no particular age and colour, with a strong Edinburgh accent, and about as emotional as a bagpipe.  Well, sir, he was like the rest of us; every time he looked at my prisoner, I saw that Sawbones turn sick and white with the desire to kill him.  I knew what was in his mind, just as he knew what was in mine; and killing being out of the question, we did the next best.  We told the man we could and would make such a scandal out of this as should make his name stink from one end of London to the other.  If he had any friends or any credit, we undertook that he should lose them.  And all the time, as we were pitching it in red-hot, we were keeping the women off him as best we could, for they were as wild as harpies.  I never saw a circle of such hateful faces; and there was the man in the middle, with a kind of black, sneering coolness – frightened, too, I could see that – but carrying it off, sir, really like Satan.”    15 0 7 Starting with this extract, explore how far Stevenson presents Mr Hyde as a threatening and dangerous character.

Write about:
• how Stevenson presents Mr Hyde in this extract
• how far Stevenson presents Mr Hyde as threatening and dangerous in the novel as a whole.

D - Dark Atmosphere

A - Atrocious Actions

R - Reactions of Others

K - Known Yet Unidentifiable Deformity

S - Symbolism of Evil

I - Impact on Dr. Jekyll

D - Dehumanization Through Animal Imagery

E - Evocative Language and Imagery

You said:

Flashcard Set: Understanding Mr. Hyde's Threatening and Dangerous Nature

Tips for Using These Flashcards