• Hi all and welcome to TheWoodHaven2 brought into the 21st Century, kicking and screaming! We all have Alasdair to thank for the vast bulk of the heavy lifting to get us here, no more so than me because he's taken away a huge burden of responsibility from my shoulders and brought us to this new shiny home, with all your previous content (hopefully) still intact! Please peruse and feed back. There is still plenty to do, like changing the colour scheme, adding the banner graphic, tweaking the odd setting here and there so I have added a new thread in the 'Technical Issues, Bugs and Feature Requests' forum for you to add any issues you find, any missing settings or just anything you'd like to see added/removed from the feature set that Xenforo offers. We will get to everything over the coming weeks so please be patient, but add anything at all to the thread I mention above and we promise to get to them over the next few days/weeks/months. In the meantime, please enjoy!

Australian word of the year, 2024

Mike G

Mike G

Petrified Pine
Joined
Jul 30, 2014
Messages
12,480
Reaction score
3,033
Location
Suffolk
Name
Mike
Ensh*ttification: the deterioration over time of goods and services. For example "Nitromors was an effective paint stripper until it went through the process of ensh*ttification".

"Google was a useful search engine until ensh*ttification allowed businesses to distort the results".

"Deezer was a wonderful competitor to Spotify until ensh*ttification set in".

"Stanley made the world's best planes for more than half a century, but then bowed to the pressures of ensh*ttification in the 1970s".


Ensh*ttification's meaning can be adjusted with prefixes, and suffixes:

"Cascamite was ensh*ttified 20 odd years ago, but has since gone through a process of disensh*ttification and returned to it's previous quality". Presumably, the accountants within the organisation who resisted the process of disensh*ttification could be described as antidisensh*ttificationites.


Given the state of the world at the moment, I can see this word coming in very useful. For once in their lives Australians may have contributed something useful to the lexicon.
 
Well, they also invented the Sheila, g'day, and the thing that instantly makes me categorise people as an idiot "no worries".

What we need is a word for AI. As far as I can see AI trawls all available content and produces an average of it. Thus everything tends towards mediocrity over time.

Not heard of your word Mike except from you here, but can see the appeal.
 
Superb. I will use that.
As for AI, I like the word “hallucinating” to describe when AI just makes it up.
 
Enshittification is a useful concept. The major part of the contribution (the term, its meaning, and popularity) is Canadian though.
 
Well, who'd've thunk it?

I'm curious how you knew that, Andy.
 
Very interesting link Andy, thanks for posting that.

Not keen on his choice of font mind you. Hard to read. Seems fashionable these days for some reason as Hostinger use similar for their web default.
 
AJB Temple said:
What we need is a word for AI. As far as I can see AI trawls all available content and produces an average of it. Thus everything tends towards mediocrity over time.
Click to expand...

As someone with my name, the growth of AI has made me hate sans-serif fonts (which don't distinguish between I and l): they're really bad for my self-esteem 😜 🤣
 
The ensh*ttification of the acronym AI annoys me too. Every manipulated image you see is now described as AI. It almost never is. It is CGI of some description. People seem to think that AI means anything done on a computer.
 
I read somewhere that Google Maps (incredibly useful when on holiday) is actually AI, 'specially when you want to get from A to B and the girlie then tells you when to turn left, right or whatever - Rob
 
Mike G said:
The ensh*ttification of the acronym AI annoys me too. Every manipulated image you see is now described as AI. It almost never is. It is CGI of some description. People seem to think that AI means anything done on a computer.
Click to expand...
Agreed. I listen to a lot of podcasts and when I can't be bothered to skip through the adverts (usually when I'm half way through cutting a dovetail or something) it's amazing how many of them claim to be using the latest/cutting edge Ai (AL, does this help?) to do what they do. I'm sure it's nonsense advertising jargon.
 
I would venture this personal.viewpoint:

C.G.I. is an image of either a) an actual object, that has been digitally 'enjanced'(!) or, b) an actual object that has been embedded into a C.G.I. background. In both respects, it's emulating older photographic techniques.
A.I. appears, again, to be one of two things: Google et al seem to use it as a précis tool to summarise many references. Secondly, particularly photographically, A.I. seems to be able to synthesise images - probably via a mass sampling tool or algorithm akin to its Google précis - that have not existed before.
I realise this is a woolly, layman's viewpoint; I may be missing some subtle aspect. Pundits keep pushing that A.I. can 'learn', but all of the above is just clever(??) copying? Can someone please quote a reference where A.I. can clearly be seen to have "learned" in the sense and context that we all did in our educative years, and secondly, can anyone please point to a published paper detailing where a box of electronics produced something unique, rather than a derivative copy?
 
Last edited:
Without looking anything up, I'll try and offer an example of machine learning. (If this is wrong, I am the one to blame.)

In medicine, experienced radiologists look at thousands of digital x-rays to screen patients for cancer. They use their experience and knowledge of anatomy to sort the images into "clear" and "possibly not clear."

A collection of assessed images can be used to train an AI program to do the same task, cheaper and quicker. There's no need for someone to devise a set of descriptions or rules to sort out the cancerous cases. Somehow, the black box picks up on what looks different, without being taught. It has "learned" from its own "experience".
 
AndyT said:
Without looking anything up, I'll try and offer an example of machine learning. (If this is wrong, I am the one to blame.)

In medicine, experienced radiologists look at thousands of digital x-rays to screen patients for cancer. They use their experience and knowledge of anatomy to sort the images into "clear" and "possibly not clear."

A collection of assessed images can be used to train an AI program to do the same task, cheaper and quicker. There's no need for someone to devise a set of descriptions or rules to sort out the cancerous cases. Somehow, the black box picks up on what looks different, without being taught. It has "learned" from its own "experience".
Click to expand...
I agree that this is a great step forward, but what happens when the experienced radiologists have retired? The human ability to do this job will be lost.
We have all experienced “computer says no!” for more trivial things.
 
SamQ aka Ah! Q! said:
I would venture this personal.viewpoint:

C.G.I. is an image of either a) an actual object, that has been digitally 'enjanced'(!) or, b) an actual object that has been embedded into a C.G.I. background. In both respects, it's emulating older photographic techniques.
A.I. appears, again, to be one of two things: Google et al seem to use it as a précis tool to summarise many references. Secondly, particularly photographically, A.I. seems to be able to synthesise images - probably via a mass sampling tool or algorithm akin to its Google précis - that have not existed before.
I realise this is a woolly, layman's viewpoint; I may be missing some subtle aspect. Pundits keep pushing that A.I. can 'learn', but all of the above is just clever(??) copying? Can someone please quote a reference where A.I. can clearly be seen to have "learned" in the sense and context that we all did in our educative years, and secondly, can anyone please point to a published paper detailing where a box of electronics produced something unique, rather than a derivative copy?
Click to expand...
I may be incorrect, but as I understand it most AI is just an agglomeration machine. It synthesises a wide range of stuff. So can be very wrong. There is some inference going on, but it is superficial.

Mind you, of course AI could come up with something unique. Infinite number of monkeys..

A long time ago (in a galaxy far far away) I was involved with some ‘expert systems’. Investment analysis. I am reminded of this: ‘The first matrix I designed was quite naturally perfect, it was a work of art, flawless, sublime. A triumph equalled only by its monumental failure.’

Which was what happened. Heuristic and Bayesian is the way to go.

Oh, and students submitting plainly AI rubbish. They don’t even regard it as wrong. Although I believe some universities have tried slipping coursework that is AI generated past the assessors. Not detected. I think some are now returning to invigilated exams. Cue bleats from students.
 
I waste too much time on Facebook and recently on wood thingies pictures purporting to be somebody’s work have come up, obviously either Ai or what Mike said, but definitely not made from wood by a clever man.
They are fairly easy to spot when you “get your eye in” far too glossy the hardware doesn’t do the right thing or is in the wrong place, but it’s the wood grain that is incorrect that gives it away, I called it out until I realised I was just teaching the damn machine how to improve!
 
Lurker said:
I agree that this is a great step forward, but what happens when the experienced radiologists have retired? The human ability to do this job will be lost.
Click to expand...
Which is why the real benefit of this type of machine learning is on cases where humans don't know how to differentiate it - the radiology example is nice because it's easy for a layman to understand what's going on, but the really interesting cases are where we have all sorts of measurements available but just don't know how to use them to predict something. A classical machine learning system can be fed all of those historical data, along with after-the-fact knowledge of which of them were later found to have the thing we want to predict, and find ways that aren't visible to humans. The key, however, is that these systems are designed with knowledge of the problem they're analysing coded into them by humans who also understand the problem.

This is completely separate from the current wave of generative large language models which are frequently referred to as "AI" and which work purely by measuring similarity of words and which ones are often used together to synthesise realistic-sounding language. These LLMs (including ChatGPT, Gemini, Bard, and every other member of that class) have no concept of actual knowledge or information. The image generation is similar; it just looks at image elements that are associated with the words it's been given, and which other image elements go near them. That's why it's so bad at hands, for example - it knows that <this shape that goes with the word 'finger'> has another copy of itself next to it three times out of four, so it adds fingers until its random number generator says stop. It doesn't have a concept of human anatomy to know that the correct number is four fingers and one thumb.

Of course that's a gross oversimplification of what's going on internally, but it's a far more useful mental model than trying to ascribe any sort of actual intelligence to these things. They can certainly be useful for some specific tasks, but they are absolutely not intelligent.
 
Tiresias said:
Oh, and students submitting plainly AI rubbish. They don’t even regard it as wrong. Although I believe some universities have tried slipping coursework that is AI generated past the assessors. Not detected. I think some are now returning to invigilated exams. Cue bleats from students.
Click to expand...
It's been reported that Oxbridge and other top universities have been given the OK for open book exams and allowing students to take exam papers home with them to complete. It will give ethnic minority students a better chance of passing, apparently.
 
Infinite number of monkeys..
Click to expand...

that would be 4 Chan and 8 chan, stormer etc ,truth social ..twitter is rapidly heading that way too.
 
You must log in or register to reply here.
Back
Top