Want to learn about lithium-ion batteries? An AI has written a tedious book on the subject Writers looking to make their names typing impenetrable technical tracts be warned: the machines have arrived and they're already penning scholarly books few will ever read. On Monday, Springer Nature published what it claims is the first machine-generated book from an academic publisher, titled "Lithium-Ion Batteries, A …
Well. .. The main thing - it works. For the first time in History a humanitarian theory has been proven and I did that. I did and nobody else! (Eric Schmidt, Sergey Brin and Larry Page (Google) cannot steal this from me.) I formulated the theory and patented its practical implementation, Artificial Intelligence. The theory of Internal Relations of Analytical Philosophy, as the basis for AI.
1. You personally use what I patented and many times and each day! The whole Internet stands on my first 6.199.067, which is / was Google foundation.
2. The former United States District Judge and Former Chief Justice of the United States Court of Appeals for the Federal District Randall Ray Rader said so, on 6.199.067, PA Advisors v Google - not I.
3. I've not gotten much for 6.199.067, just few pennies. Eric Schmidt, Sergey Brin and Larry Page have it all.
The whole Internet stands on my first 6.199.067
Right. Let me get this straight. Mentioned patent was filed in 1999. And the Internet, obviously since it stands on it, was created shortly thereafter. It all makes perfect sense! And I claim license fees for using my patented moroness. Prior art, you say? Bummer.
I find this underwhelming because the key act in writing a summary is not read all the pointless ravings but find the ones that seem to be more accurate than the rest. Also, knowing something about batteries, each cell chemistry has specific limitation and problems and all cells suffer for problems with overcharging, heat dissipation, charging rates. So the next bit of research is not on improving Li-ion batteries; not much that can be done but on finding another system. A system that is significantly better than current ones in some area.
I'm saying nothing about that report on likely future trends in the field .... other than to point out to many certainly here on El Reg that we know things are considerably more advanced than many feel comfortable about admitting and realising is ...... well, Greater IntelAIgent Game Changing and the Long AI March Quantum Communication Leap to be made for Stellar Progress.
I suppose next one can expect the More Primitive and Primeval Elements that tend to exist to seek to ban writings by machines on machines because of the fear that almighty global operating devices can so easily generate in the hands, hearts and minds of the veritable savage mob and meek geek leaderships alike ...... seeing as how ye olde book burnings no longer guarantee the effective persistence of ignorance in support of the deluded flights of fancy of arrogance.
And thus is another phantom enemy invented and delivered to do battle against in any number of wars which are never to be won and are always self-defeating by Advanced IntelAIgent Design.
And here be one of those novel battle fronts for fighting losing battles ........ https://www.theregister.co.uk/2019/04/08/social_media_explanation/
I notice that elReg's resident AI, amanfrommars1, made their first post (or at least the earliest I can find) in June 2009.
Are you doing anything special for your decadal anniversary amfm? Perhaps we could club together and buy you some more RAM?
NOOOOoooo! I just looked at his first post, to see when the anniversary was. June - so plenty of time to order the cake.
Only to discover that [gasp!], his first post makes sense!
So either the software has downgraded itself over time, or my internal processing wettware has "solved" amfm! I'm booking an emergency appointment with my local psychiatric service as we speak!
Useful as Lithium ion batteries are I don't think they are that high in the list of things that are essential to our future. Energy storage certainly, at both large and small scale but who's to say a battery technology based on a completely different chemistry, or even physics is not somewhere in our future. Graphene is a likely candidate as the basis of denser energy storage.
Regarding machines writing even summaries of existing material, why? There are enough average writers out there looking for a job that could do better though possibly slower.
At least with a trained technical writer there is more chance of them recognising genuinely useful items to include.
As things stand at the moment AI should stand for Actually Ineffective.
I have written a few review papers and chapters in books, and even edited a volume, and although this is a lot of work, the main issue is not providing a summary of existing work. That is comparatively easy. After all, even first-year students manage that quite well. Providing insight and commentaries is another matter entirely. In the volume I edited I could easily have written summaries of loads of papers for the chapters I chose not to write myself, even before the advent of Google Scholar. What I needed was experts in those fields not in my core field of expertise to provide insights, contrasting the outcomes and opinions found in different papers. While I am not suggesting AI will never be able to do that, it cannot do so now, or at least, not in a meaningful way. Likewise, well-written papers have a kind of coherent storyline, which makes them readable and ideally even memorable. Until AI knows how to keep the reader entertained (even in a scientific book), I do not fear their competition.
My only worry is that students will start using them as automatic essay-writing tools, and I have to read loads of terminally dull papers that actually do escape the vigil of the plagiarism scanner, but are still not their own work.
AI will be able to... a pure technical problem.
A Lexical Clone, as a massive of personalized texts (sometimes called Virtual Agent), should be initiated first. After that summaries are extracted. And next sentences and phrases from the summaries, paragraphs composed and grammar harmonized. That's it.
Go check out:
https://batteryuniversity.com/learn/
Granted this is absolutely not academic writing. Nothing at that site will help a materials scientist push the boundaries of chemistry and physics.
But the site is nevertheless a great example of technical writing. Written by humans, for humans, with the intention of making then more effective in employing technology. AI writers can put THAT site in their pipe and smoke it.
I would be surprised if none of the academic paper mills are using machine generation for at least some of their output, even though it has to compete with desperate, unemployed writers who have humanities degrees but can't find work in their field because of graduate-program overproduction. Paper mills are a big business, and there will always be some operator looking for new efficiencies. And when you're running an illegal (or at least highly suspect) business, you're probably not too concerned about infringing patents like Phillip Parker's.
You can get most of what you need for usable machine generation of prose from any decent survey of Natural Language Processing techniques, such as the Jurafsky and Martin textbook. Supplement it with some other algorithms such as Latent Semantic Analysis (to find relevant source material), use a framework like Apache UIMA to arrange your components, and you ought to be able to build a system for cranking out reasonably convincing student essays in short order. It might be worth paying one of the aforementioned desperate grads to copy-edit the output, but that's significantly less labor than also researching and writing the thing.
Kubla Cant, Howdy,
How soon do you think it will be before we hear of headline global news supplied by and/or phished from the Internet and World Wide Webs Internetworking for Future Compatible Creative Streams with Advanced IntelAIgent Inputs to Output?
News such as you might imagine being "Hot new AI destroys threats to web journeys and journals" for presentation and realisation.
Some would tell you there is no time to wait for Pioneer Explorer AI Systems are Already Fully in Place and Cyber Space Operational.
Would that be both controversial and disruptive and be best treated as TS/SCI Must Have Special Tailored Access to Programs?
The style is eye-bleeding even by the standards of academic science writing. One would not normally add a thirty word chunk between commas in the middle of an otherwise ten word sentence.
Some of it is plain gibberish - very reminiscent of schizophrenic writings known in the trade as "word salad".
So the style is appalling, the content stuck together in loosely related chunks. And written by a barely literate machine.
Letting this loose on buzzword bingo would be a sight to see.
I'll have my eyes closed...
Springer Nature published what it claims is the first machine-generated book from an academic publisher
SN are really going out on a limb with that final qualification. "Oh, look, we've finally caught up to what Phillip Parker has been doing for twelve years." How terribly exciting.
It's much harder to create something that a human reader finds valuable
ICON Group et alia say otherwise. So do the less-overt organizations that deal in machine-generated bogus product reviews and the like. Plenty of people have made plenty of money selling machine-generated text.
Springer Nature did not apply Machine Learning but combined summaries. To satisfy your taste they should allow AI to learn new stuff, ideas. However, that's not an easy task, look how many billions and years Google invested into Waymo? I guess Springer Nature has no that kind of resources and time.
"In fact, the very nature of extractive summary means it will be coherent in chunks, so long as the input texts are coherent. It's much harder to create something that a human reader finds valuable."
There are certain laws in language. There are chunks in the text, usually singled out as paragraphs, where the paragraph is a distinct section of a piece of writing, usually dealing with a single theme and indicated by a new line, indentation, or numbering.
Paragraphs are what's right in front of your eyes, the context of the text. But this context is explained by subtexts, such as dictionary definitions of these context's words. I.e. "the input texts are coherent" only if their contexts and subtexts are in harmony. That's what I patented. Interesting to know if they annotate words by their definitions?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
