Click here to watch a short video with Katherine Rundell, who writes amazing books for children and young (and older) adults.
The grandchild reads during breakfast, when sitting down with a snack after a long day at school, before and after dinner, before bedtime, in bed and basically whenever possible. Several times a week, the grandchild reads to me over social media, we then discuss the books. Currently it's Harry Potter of which I know shamefully little, book four, halfway through. The grandchild's mother, my daughter, would never leave the house without a book, never ever. One of my most vivid parenting memories is the exact moment she stopped moving her lips while reading, when I watched her, book in hand, her body slipping down along the wall until she sat on the floor, totally absoberd. The book was James and the Giant Peach by Roald Dahl.
In hindsight, I retired at the right time - three years ago now. Not that it felt like the wrong time, not ever. But as my work mostly involved translating and editing science manuscripts, the arrival of AI translation and editing apps meant that almost all study groups and researchers started to use them, happy to be able to cut costs and time. It took all involved maybe six months to realise that the software was just a tool and not always up to the task. Which made my job harder and more time consuming because editing AI generated stuff most of the time meant I had to go all the way back to the source, i.e. the actual human author, to figure out what the heck they wanted to say. Not always, mind you, sometimes the software produces excellent, succinct texts. Anyway, it's all in the past and I understand that most science publishers, especially the big ones, now use their own AI software to detect AI generated texts which they then reject as not original work. Go figure.
I haven't wasted much time on AI but the recent apocalyptic warnings made me laugh.
The performance of these systems involves computing power, training data and energy resources.
Software that fails to deliver the desired results or even causes damage is a not a superpower taking over the world, it is actually a faulty product, not a masterpiece. The big AI firms are producing malware on a massive scale and are attempting to shift the blame for their poor production standards onto others by warning of a fictional future apocalypse when in reality they are trying to force others to take action.
But AI does pose a very real threat to the survival of humanity. Large language models are extremely inefficient, and their now ubiquitous use is exacerbating the climate crisis. Their land use is vast, their energy and water consumption is enormous. Add to that the mass theft of copyright-protected works feeding the software, the even more rapid spread of disinformation and the financial overextension, which is beyond my understanding but I figure people are now speculating on profits that are mathematically hardly possible, risking a global economic crisis.
So there are plenty of good reasons to join forces globally to stand up to the tech giants. But none of them is a super-intelligent robot seeking to seize world domination.
Yes. Plenty of good reasons. As always, thank you.
ReplyDeleteWhat I truly know about AI could be written on an index card with room to spare. I have one child who will not touch it. I have others who use it frequently. I have a son-in-law who uses it in helping with medical questions and possible treatments. My husband asks google AI (I think) questions all the time.
ReplyDeleteAnd here I am. Floundering as to what to think, what to do.
I'd rather think about my grandson who reads like your granddaughter. The way I read, the way you read. My head is firmly implanted in the sand.
I work for Elsevier as a freelance medical editor. They are slowly phasing out all of the editors, writers, etc.; they use an AI company in India and also have their own proprietary AI editing program. I had another year or so before I was going to retire; I won't make it. My contacts have kept me in work as much as they can, but at some point it will all be done through AI. I think the ripple effects of companies using AI are devastating; so many jobs lost. And the fact these researchers are not writing their own stuff. Ugh. I had a guy submit a one page Letter to the Editor, and he acknowledged he had used an AI program to write it. Dear lord.
ReplyDeleteYes, Elle, that's the shape of things to come.
DeleteAI continues the dumbing down of humanity. I red some list of what teachers think of today's students and it is really horrifying. They can't read, they can't write, they have no idea how to do research or formulate thoughts into words. You give them a task and they look at you blankly with no idea how to go about it. They must be spoon fed and these are older students. Children have been coddled and over protected, everything done for them and now AI thinks for them. In thirty years when these students should be in positions of leadership, running their lives or businesses or governments they will be incapable. It reminds me of a Star Trek episode where they visit a planet and all the people are grown children, ignorant and incapable of anything, their lives run by an AI that does everything, thinks for them, provides everything. That's where humanity is headed.
ReplyDeleteI read, always have a book, would piss off my teachers in school because I would read in class instead of paying attention. I go to one doctor appointment with a book because I know I'll be waiting an hour. I'm the only one with a book, everyone else is on their phones.
I retired from the library at a good time, and I also got out of journalism at a good time. Both of these careers are being upended by AI. (Anything involving information, in fact.)
ReplyDeleteI somehow missed this post. AI, like so many things that humans develop, has both good points and bad points. It reminds me of Alfred Nobel and his invention of dynamite and so many other inventions that had good intentions. AI is being used in medicine to help diagnose tumours to small for humans to see on CT scans which is good. It's helping diagnose pancreatic cancer at an earlier stage.
ReplyDeleteAnd then AI is being used for so many other less well intentioned things. I think the cat is out of the bag as far as AI is concerned but what I do not want is data centres taking over land, using resources, and destroying what's good. That we can do something about.
That young woman in the video gives me hope that there are young people who still think.
I was writing a not-so-serious email to my daughter when I was suddenly interrupted by an AI source posing as a literary critic. Pointing out, inter alia, that I was failing to finish off some of my sentences, a prosodic device of which I was entirely aware. I wondered if I could torture this importunate bastard and I replied in a deliberately obscure - possibly pre-Shakespearean - manner and he (it had to be a he, even if electronic) was disposed to reply in kind. So I turned up the wick of obscurity and - after a few false starts - I had him by his intimate parts, reducing him to gushing greeting-card compliments. A fig for him.
ReplyDeleteThis is not to say I'm pooh-poohing your problems as a translator, God forbid, or the wider implications of hacking big important systems. But there is one area where AI does a genuinely helpful - if modest - job.
I have two digital-display clocks both with fiendishly diffiicult set-up procedures. Years ago I Googled the makers' names with the model numbers and was gratified to receive free instructional sheets for both. The language was wilfully difficult but I was able to struggle successfully. Both these sheets then got lost. It being clocks-back time, I resorted to Google again only to find out that such material has now been taken over by commercial concerns who charge for the service. I might have paid but I remembered the painfully techno language. Google has now launched an AI service called Gemini which I tried; immediate response and in much simpler language.
Even so, I have been taking French lessons since 1973 and using what I learned to engage notably Languedocians in their home talk (Whoops: I mean French not Oc), It's imperfect but it delights me. I hate the thought of what I've achieved being de-skilled.
Thanks for this anecdote, I had a similar AI contact yesterday and so I tried your approach, it was great fun, not with the gushing result you had, but the thing veered completely into deep waters and ended up with houseplant advice.
DeleteA while ago, I bought online a small thermos-like gadget to transport heat sensitive medicine on a long distance flight. Of course it came from China and of course the miniscule instruction leaflet was in Chinese. I considered calling to a house down the road where a somewhat Chinese looking person had recently moved in but in the end we tried AI and received instructions on how to defrost a fridge. I figured it out by trial and error and it was pretty useless in the end.
And by the way, I sent you an email, what happened to the promised 665 words on my writing style?????
DeleteComment as Roderick Robinson, it says; how could I be any other? More to the point who'd want to be me?
ReplyDeleteWhen did you send the email? Did it appear? As far as I remember I stopped temporarily at 665 words through sheer fatigue, realising there was so much left to say. Also because I was under the impression that you'd said neither yea nor nay. Never mind, you are one of my most treasured echo chambers and I will resume once this communicatio reaches full stop.
To my horror I cannot find those 665 words. I'm going to have to start from scratch. I fear you'll have to wait a bit.
ReplyDeleteWow! I had a bright idea and there it was. You won't have to wait as long.
ReplyDeleteI wrote offline, copied, tried to paste but I've been told my piece is too long for a comment. Sounds ominous, doesn't it? Make a suggestion or - if you prefer - let's forget the whole thing.
ReplyDeletesend it to dsabine@gmx.de
Delete