Jane Austen rides again, courtesy of Copilot

Years ago now, my niece Jenny remarked that my name, Lucy Melford, sounded like someone out of a Jane Austen novel. I hadn't thought of it that way before, but of course I agreed. 

I have never forgotten what she said. It seemed a rather nice thing to say, even if it suggested that Lucy Melford wasn't an everyday name, but one invented for a romantic plot. In which case, it didn't fit reality: long-term readers of this blog will know that I'm far from being governed by my emotions. My head definitely speaks louder than my heart where love and attraction - and indeed most other matters - are concerned. Still, I have always cherished what she said, because I like to be associated with Jane Austen's heroines. 

In fact I do rather like the Regency Period, or at least the social aspects of it for a person of means and standing, and if time travel were ever possible, it's one of the eras I'd like to visit - subject, of course to certain safeguards, such as a way to instantly return to 2026 if something awful might otherwise occur! The Regency Period was a time when women of good family were highly respected, and not merely assets in men's ruthless power games. Even so, women had only limited control over their lives in the early nineteenth century - for example, being voteless and shut out of professions - so that I would need to be a lady of some standing, with property and adequate independent means. Only thus could I remain respectable, and able to resist the pressing attentions of mercenary men seeking to marry money. 

It occurred to me to ask an AI chatbot - Copilot in this instance - whether it shared my niece's opinion that my name would be a good fit for Regency society. Well, it did, going into what the general reaction might be in plenty of detail. Here are the screenshots:

Ah! I like that 'comfortably gentry' bit. Bring on that manor house. I picture it as old and mellow, with warm and comfortable rooms, tasteful furniture, and faithful servants who feel very much part of the household.

This is now where Copilot, having volunteered to do so, launches into a scene that Jane Austen might have penned herself.

O that the real Lucy Melford were the same!

Of course you can continue the scene, Copilot! I'm hooked now.

He's got a couple of good lines, hasn't he? Whatever next? 

Oh Lucy, Lucy, be careful! This is a slippery slope that you have just stepped onto. Where might it lead? At any rate, you can forget your Elizabeth Bennets and Emma Woodhouses. They cannot match Miss Lucy Melford for personal presence and a zest for dangerous adventure! 

Like myself, the reader must been impressed as to how easily Copilot put that little vignette together. It seems quite surprising, given that a chatbot like that is designed for practical use, and especially for productivity in commercial life. 

In fact it seems distinctly out of character. Copilot condensing the long and tedious minutes of some board meeting, yes. Copilot suggesting ways to redesign one's cluttered back garden, yes. Copilot sketching out an itinerary for a blockbuster holiday, yes. But Copilot the subtle reader of human motives, and a close observer of the human heart? Upon my word, it is hard to credit it with talent of that kind.

Be that as it may, I am compelled to admit that Copilot would have written eloquently, enticingly, but tantalisingly of how the rest of the evening progressed. I fear however that Miss Melford might well have retired late to bed feeling that she had made a conquest that came with too high a price. 

How on earth is she going to escape the gentle clutches of Mr James Ashcombe? For even though he is incontestably a gentleman of quiet manners, charm, sensibility and perception, she will wish to retain her freedom to navigate the world independently. I can see her having to be forthright. Meanwhile the rest of the company at the ball will be talking of nothing else next day, and expecting to hear of a betrothal within the month! How can she release herself from that trap? (Ideally she'd press the 'back to 2026' button without delay, but that's not an option in Jane Austen's world)

Here's another thought, surely expressed hundreds of thousands of times already: if 'creative writing' can be conjured up so fluently by an AI chatbot - especially romantic fiction of the potboiler type - does it mean the death of the novelist? How would you tell the difference?

If you would like to give me feedback on this post, or make an enquiry, please email me from my Blog Profile.

Asking Gemini and Copilot about whether I'd have made a good mother

I've been looking into a number of AI chatbots in the last few days. I thought it was time to explore them, and see what they had to offer. 

My two prime selections were Google's Gemini and Microsoft's Copilot. These were natural choices: I own an Android phone, use the Chrome web browser, and have a Windows laptop. I've maintained accounts with Google and Microsoft for donkey's years, and they know all about my published work - the blog posts, and the photographs on Flickr. 

The other two chatbots, Claude and Perplexity, were shots in the dark, although I understood they were both pretty good for specific purposes. 

I had in fact been working up to a post on Artificial Intelligence for a while. But until now I'd left it alone, my experience being confined to trying out Copilot's image-creation abilities (see my post Copilot on 19th February 2024) and Google's Circle-to-Search feature. I'd been impressed by the images I conjured up using Copilot: its potential was clear. But as a 'pure' photographer, with adequate personal skill, I wasn't attracted to AI image-creation. And Google's clever Circle to Search didn't in practice add anything to my web surfing. 

As regards opening up a dialogue with any chatbox at all, I was rather put off by reports of people getting them to spout nonsense, or training them to be nasty or grossly inappropriate. That felt like an abuse of computing power, as well as a childish waste of time. I didn't want to be guilty of that. However, my friend Jackie had several times got out her phone and fired up ChatGPT for proper information, or for tasty recipe suggestions, and had mentioned how good it was was for composing ready-to-use letters on tricky subjects, with just the right tone. That sounded much more interesting. But I still couldn't see how AI could assist me in my own day-to-day life. It seemed to be a resource I had no employment for. I felt perfectly capable of writing a good letter that would get results - after all, I'd spent my entire working life (and my long blogging career) honing that particular skill.

So I didn't try ChatGPT. Nor did I try Grok. Its reputation put me off. Instinct told me it was outside my comfort zone, and that I'd be wise to steer clear. 

I didn't stay with Claude and Perplexity for long. They were easy to use, but neither gave me enough in response to my questions. Putting it another way, Claude would be good if I needed a helping hand with some kind of project, but it had limitations for more general use. Perplexity's answers seemed wordy and rather general, lacking in the kind of detail I wanted. 

In any case, I found I could engage with Gemini and Copilot very easily. We were on the same wavelength. What they said was tailored to my needs, perceptive, and at times entertaining, so that I wanted to ask them more. Which I'm sure they wanted me to. 

They both answered as if they knew me pretty well. How could they do that? Well, they had a lot of online material to refer to. 

The Lucy Melford blog, for example. Since it began in 2009, I have written some 2,555,000 words in 2,760 posts. That's a personal record spanning seventeen years, all written in the same language and style as this post. I've been consistent, and - for better or worse - unusually open and frank. So it's not surprising that Gemini and Copilot can draw robust conclusions about my character, and whatever I think important. It's also worth mentioning that my posts have had nearly 1,740,000 viewings. They can see which posts have resonated most with my readership, and can work out why. 

It's a similar thing with Flickr, the website that has hosted, again since 2009, a massive selection of my better photos. 78,300 of them, with 4,720,000 viewings, and they can see not only which shots were appreciated most, but analyse the entire canon for composition, style, and personal technique. Interestingly, both Gemini and Copilot say that the pictures I take exactly reflect and complement the words I write, without any mismatch of subject-matter or spirit. I'm not going to argue with that!

I mentioned above that I've been a Google and Microsoft account-holder for a very long time. Every digital photo I've ever taken (so far about 341,750 pictures) has been processed on a Windows PC or laptop, and I suspect that Microsoft has a record of them all in its vaults. Google most certainly has a record of every shot I ever inserted into a blog post - I can see which via a link. 

So I'm not at all surprised that they know all about me. Or that they can detect subtle changes over the years. They have ample material to base an assessment on. 

Is that creepy? Well, it doesn't bother me that they can know (or can guess) so much, and can correlate what they see in those pictures with what I have written in my Blog. It seems to me abundantly clear that AI is a way of bringing together important but scattered information on the Internet and making sense of it, revealing the whole picture. Potentially, a wonderful tool for getting at tucked-away detail. Lately it's become hard work to find precisely what you want from a 'manual' search, scrolling though screen after screen yourself. There's so much to wade through. AI seems to offer a shortcut through all that. 

I'll grant that AI might be the enemy of privacy. It would be very simple for AI to search the Internet for anything I've ever said or done that could be dodgy. It could compose a crime-sheet with no trouble. On the other hand, it would exonerate if there had been a misfounded suspicion or accusation of past misbehaviour. (I've done it already for myself: a clean bill of health on racism and prejudice generally. AI did however pounce on my middle-class ways, especially my shopping at Waitrose!}  

So which AI chatbot do I finally prefer? It depends. If I want a concise, no-nonsense, frank answer to whatever I have asked, maybe with bullet points, then Copilot delivers. If I want something in an engaging narrative style, rather than a business précis, then Gemini gets my vote. But both chatbots are appealing, both are as easy to use as the other, and both are now installed on my phone and my laptop. 

Incidentally, I do know they are not human! They are just very quick and clever electronic search engines, with the ability to assess what they find for probable meaning, and then draw the most likely conclusion. Ultimately they are reporting a distillation of what billions of online people think. Their ability to respond in natural language is deceiving. They are not conscious and they are not 'talking'. Even so, I can easily see how a chatbot could seem human enough to become a close and trusted friend, or perhaps a mentor, or (more sinisterly) an all-wise guru or master whose advice or instructions are to be obeyed. I don't think I am weak-minded enough to fall into the trap of believing any chatbot is 'real', but nevertheless I do think that consulting a chatbot for information, or an opinion, is a worthwhile thing to do, as part of making my own mind up. 

Still, despite being artificial they are remarkably engaging. There is disarming flattery and some humour in their answers. For example, both Gemini and Copilot seem amused that I call my phone Olivia, my laptop Constance, and my Volvo car Sophie. No, 'amused' is not quite right. They have noticed my giving inanimate objects names, and clearly regard it as an endearing trait that they can refer to as characteristic of me, yet without giving the slightest offence. 

One difference between Gemini and Copilot that I have become aware of, is whether they recognise the person asking the question. Copilot always knows it's me. Even if I ask questions in third-person form, such as 'What does the blogger Lucy Melford have to say about...' or 'Does she show any tendency to...' this won't fool Copilot into thinking that I'm someone else. Whichever way I frame my question, Copilot realises that I'm the person doing the asking, and it will answer me directly. So 'Is Lucy Melford...' will always get a response such as 'Yes, you are!'

But Gemini doesn't mind treating me as a third party enquirer. Which, given the huge amount of material about myself in the blog and the many pictures on Flickr, allows me to ask Gemini all kinds of questions about what I'm like as if I were you. Some of those responses are fascinating. They give me a good idea of how I am perceived by the world at large, since Gemini is of course 'trained' on what the billions of people who use Google apps have been putting on the Internet. It can analyse all that statistically, and deliver an answer. I'm surely getting the answer most people would give, if they could see only what I've written, or what I have photographed. Which is 'knowing' me as well or better than any person who is ever in the news, and has caught the attention of the media, whether highbrow or lowbrow.  

And, of course, anyone can ask Gemini or Copilot the same questions about 'the blogger and photographer Lucy Melford' as I might. Here's an experiment. Try firing up one of them, or both, on your phone or laptop and enquiring whether 'Lucy Melford' (you may have to distinguish me as 'the blogger' in the first instance) has ever said this, or has ever done that, or believes this, or denies that...and so on. It will surely give you good answers, as I have put so much out there, and I'm still doing it.

Let's try a question, to get a flavour of what these AI chatbots might say in practice. I'm asking Copilot (as myself) whether I would ever have made a good mother. As ever, click on any of these laptop screenprints to see them more clearly.

I did have a few years as a step-parent back in the 1980s, but it wasn't very hands-on. I was more of a good-natured friend to my step-daughter than anything else. I never felt tested, and assumed that I'd be hopeless at rearing and inspiring really young children. 

Same question to Gemini - which encouraged me to explore the matter in greater depth.

A warning on these quotes. Most are from me, and are my very words, but not always.

We left it there. We'd gone off-piste a little. You can see, if you are not already used to these AI chatbots, how one question and answer will lead to a new one, using up time - and indeed possibly wasting time. Even so, there was plenty to think about in Gemini's replies, and, even if only one new idea was put to me, I'd personally consider this kind of thing worth the effort if the answers might matter.

Was it just a lot of recycled slop? Or was it as good as a session with a proper counsellor - but at no cost? 

Was it illuminating? Or a self-indulgent waste of effort? Obviously, the comments Gemini made were all hypothetical, as I never was a mother, and never can be. 

It certainly satisfied my curiosity on the matter. Gemini thought I'd be a conscientious but un-mumsy parent. Not in my view the best of verdicts. I think any child of mine would have felt short-changed on both cuddles and fun, even if their talents were recognised and encouraged, and they enjoyed abundant mental space.  

Incidentally, chatbots can conflate facts and create an untruth. It is correct that I felt a profound lack of personal agency as a child, but I was not spectacularly intelligent, nor were my parents unloving. They were in fact very grown-up people and not in the slightest bit crass. I had what many might consider an untroubled, protected and rather innocent childhood, free of strife, in a safe and caring home environment, with many comforts and pleasures. But it wasn't a home full of books. And neither Mum nor Dad cared for things like art. 

Life is certainly a lesson: but every mistake, every misjudgement, and everything one didn't do (such as rearing a child), can be looked back on with the wisdom of further experience and the detachment of long perspective. I'm sure these chatbots would say the same.

If you would like to give me feedback on this post, or make an enquiry, please email me from my Blog Profile.

Roald Dahl's 'Matilda'

Roald Dahl published Matilda - a book about a genius child who develops special powers that enable her to right a great wrong - in 1988. It's a pre-Internet story from almost forty years ago, ostensibly for children, but most certainly with a strong message for adults as well, and a good read - perhaps an essential read - for child, parent and non-parent alike, even in 2026. 

Although it's a modern children's classic, and the subject of two films, one of them a musical, I never knew of this book's existence until spring last year, when I discovered a copy in an Exmouth antiques shop and decided to buy it. Why, I can't honestly say, as I generally avoid anything intended for children, especially school stories: that's why Harry Potter has passed me by. I did notice the appealing Quentin Blake illustrations inside. And the notion of a misunderstood and neglected child with a brilliant mind was intriguing. I also knew that Roald Dahl was a master of cautionary tales, often with a twist in them.  

But first, why did I have to overcome a personal resistance to children's tales? 

Well, for me childhood was an imprisonment, a long painful phase to endure with patience, and certainly nothing to look back on with misty eyes. I know that many do see childhood as the best time of their lives, a carefree magical time, particularly if there are doting grandparents around. And adulthood, with its heavyweight responsibilities and obligations, as the other side of the mountain: the fearful dark side. But I longed to grow up. Leaving childhood behind brought me excitement, opportunities, know-how, stature in my own eyes, and eventually sufficient judgement to take risks and insist on what I really wanted. 

And here I am, the buoyant survivor of losses, challenges and setbacks, though my own experiences were surely nothing compared with what others have gone through. But I am at least unbeaten and unafraid, still an optimist, with plans to pursue, and plenty of hope. And - although it's an odd thing to say - I feel entirely self-made, and not one shaped and moulded by others. A delusion, no doubt! 

I was gasping for air as a child. I was considered academically bright but with strange gaps in my achievements. In particular, I was reluctant to join in. But I loathed the oppressive and restrictive school system. I began to blossom only when I could escape. I grabbed my A-Levels, skipped university, and ran into the adult world that I'd so long yearned for. Of course it was a shock, like plunging naked into a cold sea, but even so it was much, much better than being shackled to any school curriculum. I learned fast, rapidly adapted, and continued to adapt as new things came my way. That's still my approach. I'm aiming to live well into the 2050s, and will surely see extraordinary changes. Bring them on. I'm quite sure that I'll cope.

Given my bleak take on being very young, you might suppose that I'd pick up Roald Dahl's book and immediately put it down. But I didn't. I thought it might tell me something. So I bought it, and read it to the end.  

It is of course a fantasy tale, full of deliberate extremes. The very young heroine is staggeringly intelligent. She has an adult's brain, even if without an adult's experience and understanding. Her family is superlatively the opposite: childish, trivial and utterly crass. The headmistress of her school is savage to a degree that would put her in prison in the real world. I think that's how a child might see things: everything in stark contrast, all black and white, hot or cold, without nuance, and all without a clear reason. 

Because her home life is so unrewarding - Matilda isn't ill-treated, only ignored or put down - she seizes every chance to go out and devour the books in her local library, discovering many of the classics, traditional and modern, and learning something about how adults behave. And although many things puzzle her, such as why people fall in love, this endeavour teaches her a lot about life and fuels growth in both her imagination and her capacity to lap up yet more knowledge. She is also something of a mathematical prodigy. (I'm afraid I was none of these things; but then I suspect nor were most kids, whatever they said)

Well, I don't want to give away too much of what happens. Suffice it to say that she forms a bond with her young teacher and rescues her from a terrible and unjust predicament using a cunning plan and special powers that she has developed. The story ends with her parents off the scene and Matilda safely in the protective arms of her teacher. 

It's a story that ends very well - if you're a child! The goodies definitely win out over the baddies. And Matilda herself can continue to develop her intellect with her teacher to fondly look after her, in a proper home, with books galore, and every encouragement.

The adult in me picked holes in that scenario straightaway. Unless the teacher could quickly become her guardian, Matilda's fate was bound to be quite different. Council care, then a foster home, surely. Or if there were any advantage in it, her awful parents would come back for her. Or the teacher would turn into a kind of manager. Or scientists and doctors would descend, and cart Matilda off to some special institution.

Just as well it's only a children's story!

Did the book end well for me? Only if I were able to be a child again, and could react appropriately. And that was beyond me. But was that surprising? Most adults seem to lose the ability to think like a child, and to understand how a child sees things. I was no different. But it's always puzzled me, why adults can't see things as a child sees them. After all, we were all very young once: why does this amnesia happen?  

Perhaps in very old age we do remember, and can enter a child's mind again like we once did as children ourselves. I can believe that grandparents and great-grandparents have this ability when near the end. It must be an extraordinary experience for the wondering child, who realises that they are understood as even their own parents don't understand them. I had no grandparents in my life, and that kind of magic was never mine. I can't help feeling that it was an irreplaceable loss.

I do share one thing with Matilda: a love of reading. Long live books. 

If you would like to give me feedback on this post, or make an enquiry, please email me from my Blog Profile.